ACTA PHYTOGEOGRAPHICA SUECICA 68 EDIDIT SVENSKA V AXTGEOGRAFISKA SALLSKAPET

Studies • m Plant Ecology

dedicated to Hugo Sjors

Edited by Erik Sjiigren

Almqvist Wiksell International, Stockholm &

UPPSALA 1980 Suggested citation: Studies in plant ecology. Dedicated to Hugo Sjors. Erik Sj ogren. Ed. Acta Phytogeogr. Suec. 68. Uppsala.

ISBN 91-7210-068-0 (paperback) ISBN 91-7210-468-6 (cloth) ISSN 0084-5914

© The respective authors 1980

Svenska Vaxtgeografiska Sallskapet Box 559, S-75 1 22 Uppsala

Editor: Erik Sjogren Technical editor: Gunnel Sjors

Phototypesetting by Textgruppen i Uppsala AB Printed in Sweden 1980 by Borgstroms Tryckeri AB, Motala

Preface

The Swedish Phytogeographical Society is publishing the present volume in honour of Professor Hugo Sjors, Head of the Institute of Ecological Botany from 1962 to 1980 and holder of the chair of Professor of Ecological Botany during the same period. Hugo Sjors has been and still is the chairman of the Swedish Phytogeographical Society. This combination of activities has b;;en particularly fruitful not only for the Society but also for the Institute and its students. Manuscripts submitted to Acta Phytogeographica Suecica are prepared fo r publication largely as a result of efforts made by members of the Institute. In the world of science Hugo Sjors is fo remost known as an ecologist of mires. His knowledge and results in this respect are not restricted to Scandinavian mires but cover such vegetation within the entire northern hemisphere. Research on mire vegetation at the Institute was earlier firmly anchored in the work of Rutger Sernander and G.E. Du Rietz but for the last two decades it has been led by Hugo Sj ors, with a rich variety of new methods and approaches. During the second half of the 20th century research in ecological botany has been conducted on an increasingly wide base. Research and education at the Institute of Ecological botany has been adapted to this international trend largely as a result of efforts made by Hugo Sjors. Thus, he has initiated and encouraged a large number of research projects within the sector dealing with the reaction of different types of vegetation to cultural influence, and environmental management in connection with this kind of influence. Papers published in this volume are by pupils of Hugo Sjors and by botanists who at one time or another have conducted research at the Institute of Ecological Botany. It has not been possible to collect these contributions under a common title. Instead, they may be considered to illustrate the multi-facetted research programme at the Institute. Nevertheless, it may be observed that many of the contributions contain aspects associated with changes in vegetation. They emphasize to some extent how ecological botany today has a considerably greater interest in vegetation dynamics than earlier, when a large sector concerned plant sociological limitations of plant communities that often came to be regarded as fairly static. The work involved in preparing this volume has been done in a relatively short space of time thanks to the intensive and willing efforts of many of the staffof the Institute of Ecological Botany : Miirta Ekdahl and Stina W rern have typed the manuscripts and all the tables; Agneta Nordgren drew most of the diagrams and maps; Folke Hellstrom did all the photographic work as well as the reductions of diagrams and other illustrations; Nigel Rollison helped the editor with language revision and some translations; Kuno Thomas son checked the lists of references submitted with each contribution; Gunnel Sj ors edited and prepared the material from a technical point of view; Ake Sjodin and Salme Sed man kindly helped with library work; Henrik Mosegaard drew the cover illustration of the bound edition. The Society wishes to warmly thank all people concerned.

Uppsala in December, 1980 The Swedish Phytogeographical Society Erik Sjiigren Contents

Two Mexican conifers - Cupressus lusitanica and Pinus patula - in a juvenile development observation trial in Tanzania. By Tore Arnborg 7 Vegetation changes after fe rtilization on drained peatlands in Central Sweden. By Ingvar Backeus 17 The history of woodland in Fnjoskadalur. By Agust H. Bja rnason 31 Strandverschiebung und Strandvegetation auf Havero Prastang. Von Wolfgang Cramer 43 The Sibirian fern Athyrium crenatum (Somf.) Rupr. found in Sweden at K vikkjokk, SW Lule Lappmark. By Sten Dahlskog 51 The downward migration of plants on a rising Bothnian sea-shore. By Lars Ericson 61 Ragvaldsmossen. A pollen-analytical study of the Postglacial forest his­ tory at the "limes norrlandicus", South-Central Sweden. By Magnus Fries 73 A comparison between the alpine plant communities of Alaska and Scan- dinavia. By Olav Gj rerevoll 83 Attached algal vegetation in some streams from the Narssaq area, South Greenland. By Catarina Johansson 89 Three new subspecies of the apomictic Ranunculus auricomus L. (s. str.) from the Flora Upsaliensis area in Uppland, Sweden. By Erik Julin & John Axe/ Nannfeldt 97 Fine-root dynamics in a Scots pine stand with and without near-optimum nutrient and water regimes. By Hans Persson 101 Vegetationsentwicklung in einem Weidegebiet auf Siid-Oland, Schweden. Von Lars Rodenborg 111 Plant cover regeneration in a mechanically disturbed limestone heath on Oland, Sweden. By Ej vind Rosen 123 Observations on two bushfires in the Western Kalahari, Botswana. By Christina Skarpe 131 Continued investigations of epiphytic lichen flora around K varntorp in Narke. By Erik Skye 141 Dynamics and production of semi-natural grassland vegetation in Fenno- scandia in relation to grazing management. By Eliel Steen 153 Plankton of Lake Kariba re-examined. By Kuno Thomasson 157 Rhodomonas minuta and Rhodomonas lens (Cryptophyceae). Aspects on fo rm-variation and ecology in Lakes Malaren and Vattern, Central Sweden. By Eva Willen, Mats Oke & Frank Gonzcilez 163 Phytoplankton from lakes and ponds on Vestspitsbergen. By Torbjiirn Willen 173

Acta Phytogeogr. Suec. 68 Two Mexican conifers Cupressus lusitanica and Pinus patula 1n a juvenile development observation trial in Tanzania Tore Arnborg

Aim of the experiment traduced to Tanzania in about 1905 and proved out­ The experiment was planned as an observation trial standingly successful at high altitudes on mountain to get some information about individual differences sites with deep volcanic soils and annual rainfalls ex­ in plant populations of two exotic conifers, Cupressus ceeding 800 mm. It also did well on poorer soil and lusitanica and Pinus patula, during the ftrst year after since 1930 it has been planted for timber production planting. The experiment started with plantation of in East Africa. seedlings on 22 April, 1977 and ended on 15 June, 1978 but some specific notes were taken until Pinus patula Schiede and Deppe. This pine has a fair­ December of the same year. The main fe atures ly narrow distribution range within Mexico, around observed and measured were the development of the Lat. 20°N and only at high altitude, the highest shoot (the crown with top and branches) and the recorded being about 3000 m above m.s.l. There is root-system. no specific provenance in the home country that is The overall intention with the trial was to compare known to be remarkably better than others. Some a few planting methods in common use or easy to characters, however, show marked clinal variation. apply, and to study the variation in development The monography by \Vormald(1 975) gives all kinds within each species and between the two. In spite of of information about this pine, in its home country the limited number of seedlings, only 180 of each and as introduced fo r man-made fo rests in tropical species, some fm dings give distinct proof of and subtropical regions in Africa and Asia. differences which obviously are of interest to The species belongs to subsection Oocarpae, like silvicultural management practice. It must, however, P. oocarpa and P. radiata. The Spanish name is be stressed that this is an observation trial and that "pino patula" or "pino shino", and in English it is the purpose of this paper is to provide some ideas called the "Mexican weeping pine" or "patula pine", about ways and means to plan fo r fu rther, more the latter being the most accepted name and the only sophisticated, experiments. In addition to the main used here. trial, some notes and photographs from other ex­ The patula pine was introduced to South Africa in periments have been included, even one from Kenya. the earliest years of this century and was very soon used fo r commercial plantations. In East Africa it was only occasionally planted as a single ornamental The species tree and not until around 1930 was it given attention Cupressus lusitanica Miller. This cypress is widely as a very fast growing exotic. Then the patula pine distributed in Mexico, from Lat. 26°N southwards. It was planted in stands, and so was P. radiata, which was introduced to Portugal around the year 1600 and on some sites was even more rapid growing. (The when first named and described, in 1778, it was latter, however, was attacked by Dothistroma blight called "lusitanica" which means "from Portugal". and in 1956 banned for industrial plantations.) Later it was introduced to India, where it was often In Tanzania, as well as in Kenya and Uganda, the called "the cedar of Goa", and to Africa. The most patula pine grows on a wide range of soils. It does common name now is "Mexican cypress". For very well on young volcanic ash but also on mature, further information, reference may be made to Wolf fairly infertile laterite. The rainfall is 900 mm or Wagener (1948) and Dyson & Raunio (1977). more. The pine has a strong root-system and general­ & The Mexican cypress was, as far as is known, in- ly a pronounced tap root.

A eta Phytogeogr. Suec. 68 8 To re Arnborg

Ra1nfall Mean monthly mm temp C 0 250 25

• 20 200 • • • •

• • 150 • • 15

100 10

50

F M A M A S 0 N 0

Fig. 2. Mean monthly rainfall (dotted bars) and temperature (large dots) in Lushoto, 52 years of record (Lundgren 1978). Added here are the recorded amounts (solid bars) of importance to planting and establishment of the experiment, 1977.

with raw material from man-made forests ofthe same species. The ecology of the two species, especially con­ cerning nutrient cycling and risk of soil deterioration, and the importance of soil management in this Fig. 1. Cupressus lusitanica of good appearance and Pinus respect, has been studied extensively by Lundgren patula of a common type with a heavy, branched crown. (1978). The root systems are well developed. The planting method was the usual for commercial plantations, i.e. removal of the polythene tube, in which the seedlings were raised in The observation trial the nursery.-About 14 months after planting, 26 months The experiment was laid out at the Forestry Research from seed. Experiment 889, 30 June 1978. Station, Lushoto, in the Usambara mountains in northeast Tanzania, and numbered 889 in the list of silvicultural trials handled by the station. The site was These two conifers, Cupressus lusitanica and Pinus in a garden placed at the disposal of the author during patula, are the most common forest trees planted in the years 1977 and 1978. The area, a slope with deep, the highlands of Tanzania, in the upper midland and humus-rich laterite soil, was terraced before planting upland forest zones, from around 1200 to about 3000 to prevent erosion. In addition to the conifers under m above m.s.l. Several thousand hectares have now trial, the terraces-about four metres wide-were used been planted since the 1950's which is much more for an observation study (888) of some 25 indigenous than the area of plantations of indigenous trees like tree species, planted in 1977, and for bananas. The Juniperus procera and Podocarpus-species. The rota­ altitude here is about 1470 m above m.s.l., the rain­ tion, the years from planting to clearfelling, is fall is stated to be 1060 mm with the long rainy generally 20 to 30 years, the running production season in March - May (Fig. 2). around 20 m2 per ha and year, in some cases much The first planting was on 22 April 1977, the se­ higher. To some extent a second generation has cond on 13 May, the third on 3 June. Spacing 3 x 3 already been established on land set aside for these dm. Four methods were applied (see below). Number government plantations. A cellulose industry has of seedlings each time of each species: 60. been planned in Tanzania which will use mostly Replications: 3 (4 x 5 seedlings). Maps and marks patula pine. In Kenya a papermill is already running made it possible to fo llow each individual. In addition

Acta Phytogeogr. Suec. 68 Two conifers in a juvenile development observation trial in Ta nzania 9

to the measurements of the plants in this trial, they were kept under more or less continuous observation and were also photographed. The photographs published here were all taken by the author.

The plant material The seed of Cupressus lusitanica came from the seed orchard at Mukussu, near Lushoto, that of Pinus patula from non-registered planted stands in the country. In both cases the seeds were those common­ ly used for commercial plantations. The seedlings were raised in the research nursery, Lushoto, where well-established practice is fo llowed (EAC/NORAD 1979). After sowing in April 1976 and transplanting into polythene tubes (15 x 6 1/2 cm) a few weeks later, the seedlings were kept in the nursery for about a year. Every second month they were lifted up and root-pruned, which prevented them from growing too much. The repeated pruning from below, and splashing out of soil, during watering from the top, often causes loss of too much soil in the tube and imbalance between shoot and root. However, the seedlings used were normal and in good condition. The average data of ten seedlings of each species were found to be (the length of the shoot and the root cylinder in cm, the weight of the shoot and naked root, fresh and air-dried, in grams):

C. lusitanica shoot: 56 cm-16 g (f resh), 5 g (dry) root: 11 cm-4 g (fresh), 1 g (dry) P.patula shoot: 62 cm-28 g (fresh), 10 g (dry) root : 11 cm-9 g (fresh), 4 g (dry) The longest single root was 16 and 21 cm respec­ tively. The roots were not coiled. Thus, the sizes of the seedlings were about the same in the two species, and the weight of pine about twice that of cypress. The pine seedlings had well-developed mycorrhiza; the roots of the cypress were not examined (en­ dotroph). Mycorrhiza is inoculated with forest soil added to the nursery soil mixture. Different types of seedlings are shown in Figs. 3, 4 and 5.

Fig. 3. Cupressus lusitanica in experiment 890, three Planting methods weeks after planting. The tallest plant is 7 5 cm. June 1977. The fo ur methods used were: (A) digging a hole, planting the seedling with the polythene tube on, fill­ ing soil around, lightly and carefully pressing it around the cylinder; (B) removal of the tube, otherwise exactly the same as (A); (C) digging a little wider hole, loosening up the soil cylinder, spreading out the root system, filling soil carefully around and above the roots. Method (D) was the same as (C), only that a few litres of water were given directly after planting.

Acta Phytogeogr. Suec. 68 10 To re A rnborg

Fig. 5. Two years old plants of patula pine showing diffe rences in branching and top development but all with good root systems. Raised in Swaziland beds, overleft from the planting season (Fig. 14). Sao Hill, 9 June, 1977.

Fig. 4. Pinus radiata (left) and P. patula seedlings in Results, conclusions and assumptions polythene tubes, ready for planting. Shume nursery, 8 Oc­ Growth of the shoot and crown tober, 1977. The root strengthening period is assumed to last from a few weeks to one or two months depending on temperature and access to water. When the plants in late April were planted out, the root-system was ap­ parently largely inactive, with few "white tips". The picture was about the same at a later planting time. The seedlings had been "kept short" in the nursery. In all methods attempts were made to get the up­ In the new environment, perhaps after a shock period per part of the root system two inches below the sur­ of some days only, they started to develop. rounding surface, but not in a deeper hole, as often The growth of the shoot started later, although a practised in the field. Method ,(B) with careful few individuals were a slightly ahead of the average, removal of the plastic tube, is the most common and more among cypresses than in the pine population. In most likely the best covered-root planting method so general, the plants showed practically no growth dur­ far used. The covered-root method (A) with the ing the months of May until August. The pines had plastic tube remaining in place is not used. Bare-root few if any branches when planted out as seedlings. planting of seedlings with naked roots direct from the The cypress plants had short branches only, although nursery is recommended in regions with very high it was possible at an early stage to separate the rainfall during the planting season. Methods (C) and "narrow-crowned" from the "wide-crowned", as il­ (D) can be compared with bare-root planting. lustrated in Fig. 3. When the more rapid growth of

Acta Phytogeogr. Suec. 68 Two conifers in a juvenile development observation trial in Tanzania 11

Table Measurements of plants/trees in the juvenile develop- en average tallest I. (open) (fd led) 29) ment observation tria 1 (889), Lushoto, 1977-1978. 6 A CUPRESSUS LUSITANICA • PINUS PATULA Cupressus 1us i tan i ea Pinus �atula 27) 0 23/6 3/lO 28!12 15/6 cm 23/6 3/10 28!12 15/6 0 • PODOCARPUS USAI'IDARENSlS () JUN!PERUS PROCERA 22) • 45 8 42-49 44 ( 1) 112 28 50-59 121 47 3( 5) 16 97 7 60-79 92 32(48) 2( 5) 20) 21 21 80-99 64(65) 3( 7) 47 100-119 37 (20) 12(22) 18) 44 8 120-139 4( 1) 28( 30) 23 20 140-159 45(36) 16) 21 lG0-179 2':(22) 27 180-199 15( 9) 14) 31 200-219 6( 4) 19 220-259 3 ( 3) 120 13 260-295 1( 1) 49 69 118 189 average 52 62 91 (84) 153(143) cm 100 7 26 31 34 dead,n 12 37 40 41 180 180 180 180 tota1 ,n 180 180 180 180 80

60 the shoot started in October the branches also started 40 to develop and the individual differences between the crown system became very pronounced (see below). 20�,---���----��----�--r-.-.--.� The fastest growth occurred during the months of 1977 A M J J A S 0 N D J F M A M J J 1978 March/April/Ma y (the main rainy season). Fig. 6 gives the average plant/tree height and single tallest Fi,. The growth of the plants/trees in observation trial 6. individual height at the fo ur times of measurements. 8 9 and (triangle with dot) in Two in digenous species 890. Table I shows the numbers in groups of height­ for comparison. classes. There is a remarkable variation already in December and in June, one year after planting-there is a very wide spread. The smallest, but healthy, cypress was 76 cm (one year earlier 49 cm), the tallest 295 cm (64 cm). The smallest pine was 68 cm (50 cm), the tallest 275 cm (44 cm).

Fig. 7. Cupressus !usitanica from the main experiment, three groups, representing the planting methods

(A), (8) and (C), from left to right. June 16, 1978.

eta Phytogeogr. Suec. rJ 68 12 Tore A rnborg

Mukussu seed orchard. Some of the clones in this orchard are very wide-crowned and were not in­ cluded in the new orchard, planted in 1978. There l were also a fe w clones suspected to have traces of C. macrocarpa, a cypress from California, similar in characteristics to lusitanica but rather more coarsely branched. Being more susceptible to cancer, it is no longer planted in East Africa. The variation means variation in wood quality (amount and size of knots). It often has been stated that with plus-trees outstanding in growth, as parent trees in an orchard, tree breeding will very soon give faster growing forests, yielding perhaps much more than the already high-producing cypress fo rests of to­ day. In addition, the branches must be considered and not only the length-more wood in the stem, less in the branches-but also the angle between branch and stem. There are some very pronounced fastigiata-types as well as extreme pendula-types, both considered poor quality trees (risk of bark enclosement in wood), and, of course, never selected as plus-trees. In the material in the trials there were a few dwarf trees of condensata-type. So-called "idiot-trees", like those in Fig. 8, could be one in a hundred planted out. The poorest are rejected already in the nursery. It has been proposed that a breeding programme should be started using cuttings from outstanding young trees of C. lusitanica and considering growth and quality. The tallest tree-in trial 890-grew from seed to 3.4 m in 27 months and thus after planting this individual added nearly 3 metres to its height in 14 months and had a crown-width of 1 1/2 metres only.

8. C. !usitanica, about 15 months after planting, Fig. Pinus patula. The individual differencesin branching some remarkable types, from very poor to representing and crown system became very clear already during good. Kisumu, Kenya, June 2, 1974. very the first six months after planting. At the end of the year of observation the most common type was similar to those in Fig. 9. Some had wide, very bushy crowns like the one in Fig. 1, with the top hardly overgrowing the longest branches. About 40% had a Va riation in branching and crown development more or less retarded top. A few had a top which ob­ C. /usitanica. Fig. 7 shows some differences in viously had ceased growing, although it was alive, branching and crown width in the main trial. In the others had a strong but more slowly growing top observation experiment 890, established especially which most likely should have exceeded the branches fo r a study of this variation, the ratio height:width in later on. In some cases only one branch grows fast the whole collection (171 plants alive of 180 planted) and the tree might get a double stem in such a way 13 months after planting was 188:11 2 cm or 60%. that it could be called a "false forking". Fig. 10 il­ The tallest of all had the measures 340:155 cm. The lustrates differenttypes of branches and top develop­ most narrow-crowned had a width 20% of the height ment, indicating later differences in the crown ( 194:41 cm). The widest crowns were up to 90% of system, which is an important consideration when the height (258:235). The seed for this experiment plus-trees for the breeding programme are chosen. was the same as fo r the main trial (889), viz. the Table I shows (in brackets) the numbers of plants in

A eta Phytogeogr. Suec. 68 Tw o conifers in a juvenile development observation trial in Ta nzania 13

Fig. 9. Pinus patula from the main experiment, three groups, repre­ senting the planting methods (A), (B) and (C), from leftto right. June 16, 1978.

height classes counting the stem top only, not in­ cluding exceeding branches. There was no "uninodal type" (no branches between the yearly whorls, like in Pinus sylvestris), no indication so far of development of "foxtail type". Some of all these "multinodal types" had, however, few and short branches. The tallest tree of all was like

Fig. 11. The top shoot of an extremely nice and fast­ growing pine in trial 889 divides into two tops (left). Within few months the two stems in this true forking get several Fiig. 10. Different types of crown development in experi­ upright branches and the tree seems to be a poor fastigiata m1ent 889. (Cfr. the text.) June 30, 1978. pine. June 30 and December 7, 1978.

A eta Phytogeogr. Suec. 68 14 Tore Arnborg this, adding 2.3 metres to its height, during 13 months after planting. This pine was the only one showing "true forking" (an initially normal top bud divides into two buds, growing from the beginning into two equal stems). Fig. 11, (left) shows this fo rking. Very soon this tree started to develop a number of upright branches, fo rming a multistem tree, most likely to be a typical "fastigiata pine" (same Fig., right).

Variation in root-system development After the main growing season in 1978 the observa­ tion trial could not be continued any longer, although signs of slower growth due to competition had still not been noted. During the year the roots had not been studied but now, on June 15th, some represen­ A B tative trees were dug up and photographed. The c different methods of planting show diffe rent in­ Fig. 12. Root studie of C. lusitanica (upper) and P. patula fluences on the development of the root-systems. In from trial 889, representing three different planting method . See Fig . 7 and 9. June 16, 1 978. short the fm dings were: Method (A). The polythene tube, which was still in place and generally unbroken, had fo rced the roots to grow out horizontally below the tube. A few cypress roots had penetrated the plastic. The pines studied had no roots going through the tubes but all had started to develop a taproot. No coiling (spiralling roots) was observed. Method (B). The cypress root had more upper roots going outwards, but some of the main roots were still vertically pressed together as they had been before planting. The growth thus initiated was even more pronounced in some of the pines. It was as if the upper part of the root had been enclosed in the soil cylinder, although the plastic cover had been removed. The taproot was similar to the one in (A). Methods (C) and (D). Most of the spread-out roots had developed more "naturally", although some roots were too bent and "one-side" going, most likely due to the way they were placed at planting; this is known to be common in bare-root plantings, sometimes causing the trees to have less stability. Fig. 12 shows different types of root-system development in the methods (from left) (A), (B) and (C). On the pine root (A) the plastic tube was removed after the tree was dug up.

Planting methods and survival A total of 77 plants, or 20 %, died during the year. Method (A) had the highest survival with 5 dead, and Method (C) the lowest with 28 dead, which means that one-third of the bare-root planted trees with no watering died. However, Method (D) had nearly the same death rate. Method (B) was better for pine (3 Fig. 13. Conelets on young trees, the total age (from seed) dead), than fo r cypress (8 dead). The planting con- being about 32 months. December 7, 1978.

A eta Phytogeogr. Suec. 68 Two conifers in a juvenile development observation trial in Tanzania 15

ditions were best in April ( 44 mm rain the day before J J �nd on the day of planting)-the poorest in May (several dry days). In general, nothing can be con­ cluded from this observation trial, as the dead plants could not be dug up to find out the cause of death. However, there were indications that the covered­ root planting was better than the bare-root planting; the highest mortality occurred during the first months after planting. See also Table I.

M 0 Early flowering In November 1978 a few young trees in the trial started to develop inflorescences, both fe male and male. Fig. 13 shows cones at their first year stage on a patula pine, 32 months of age from seed.

J DEC

rainfall (total 936 mm) SUMMARY evapotranspiration (total 1231 mm) It must be repeated that from the beginning these rlanting (Dec. - March) preparation trials were planned to be observation experiments on­ first statement tHjf±ti±H± for next planting and weeding ly, emphasizing the individual differences more than beating up, where counting survival '�"-' ' the average ones . "-" needed (2nd year) • and weeding Seasonal differences in growth of Cupressus lusitanica and Pinus patula were registered. It is im­ Fig. 14. Planning for plantations and timing of planting must be in accordance to the climatic conditions. Thi is an portant to learn about the growth rhythm through attempt to illustrate the round-the-year life saving more intensive registrations (weekly fo r two years programme for establishment and checking, taking into after planting) for timing of measurements of con ideration the rainfall and evapotranspiration as at Sao progenies and other research experiments. Hill. The ba ic figure with values from meteorological The variations in development of the crown from record i taken from Brown & Cocheme ( 1973). seedling to young tree were noted, indicating hereditary characters of importance fo r the tree breeding programme and fo rest improvement.

Fig. 15. A view over plantations of patula pine at an altitude of about 2200 m, in South Tanzania, on previous grassland with patches of b mbo grove Kiwira, January

14. 1978.

Acta Phytogeogr. Suec. 68 16 Tore A rnborg

The root-systems developed differently as a result REFERENCES of different planting methods. The method most com­ Brown, L. H. & Cocheme, J. 1973. A study of the monly used in practice seems to be the safest, agroclimatology of the highlands of Eastern although some deformations of the root-s ystem were Africa.-FAO/U NESCO/ WMO. 198 pp. noted. More root studies are needed. The indications Dyson, W. G. & Raunio, A.-L. 1977. Revised herita­ of the differencesbetween cypress and pine should be bility estimates for Cupressus lusitanica in East checked. If there are pronounced differences atten­ Africa.-Silvae Genetica 26. tion should be paid to those of importance for selec­ Lundgren, B. 1978. Soil conditions and nutrient cycling tion of sites. under natural and plantation forests in Tanzanian The climate is an important factor in the plantation highlands.-Reports in forest ecology and forest soils. 31.-Dept. of Forest Soils. Swed. Univ. of Agric. of all tree species. If data on rainfall and Sciences. Uppsala (Mimeographed). 426 pp. evapotranspiration are available, a planting pro­ FAO, 1976. Savanna afforestation in Africa.-TF-RAF 14 gramme as illustrated in Fig. may be useful. 95 (DENM.), FAO. -Rome. 185 pp. Average rainfall figures are of limited importance Wormald, T. J. 1975. Pinus patula.-Tropical fo restry without knowledge of time, intensity and probability papers, No. 7. Department of Forestry, Com­ as well as vaporization and soil conditions. This is, monwealth Forestry Institute, University of Oxford, however, another part of the programme for es­ Oxford (Mimeographed). tablishment of man-made fo rests in Tanzania, using EAC/NORAD. 1979. Lowland Afforestation Project. - in some areas exotic species like the two observed Summary report by Department of Silviculture. Oslo. here, and in other areas indigenous species, some of 96 pp. Wolf, C. B. Wagener, W. E. 1948. The New World them fairly fast-growi ng. In Fig. 6 the juvenile growth & Cypresses.-EL ALISO 1:1-144. of two trees from the Usambara mountains is il­ lustrated, both species producing valuable wood, the cedar of Juniperus procera and the podo of the Podocarpus usambarensis.

Tore Arnborg, Institute of Ecological Botany, Uppsala University, Box 559, S-75 1 22 Uppsala, Sweden

Acta Phytogeogr. Suec. 68 Vegetation Changes after Fertilization on Drained Peatlands in Central Sweden lngvar Backeus

Forest drainage on a significant level has been under­ Sweden is increasing, although it still takes place to a taken in Sweden since the late 19th century (Holmen surprisingly small degree (Holmen 1979). L64). At that time it was a common view that more Investigations on vegetation changes caused by or less all kinds of peatlands could be transformed fertilization on peatlands are few. No studies have in:o productive forest lands, provided the draining been published where the development has been was sufficient. Opposition against this view was ex­ fo llowed from the time prior to the fe rtilization. pressed at the beginning of this century, i.a., in Ger­ Malmstrom (1935) described the contemporary many by von Tubeuf ( 1908). In Sweden, Tiberg vegetation in detail and deduced earlier stages (1908) expressed the same view. A cultivation experi­ through various sources on a few mires in Vaster­ ment on a drained bog was published by Osvald in botten, N. Sweden. Information from Finland can be 1917. He demonstrated a considerable enhancement fo und in Paivanen & Seppala (1968) and in the development of Scots pine (Pinus sylvestris) Heikurainen & Laine (1976). Reinikainen (1965) and Norwegian spruce (Picea abies) afterfe rtilization studied the colonization after fe rtilization on naked with basic slag, potassium salt and/or Chile salpetre, peat at an abandoned peatery in Finland. whereas timing alone resulted in only very limited reaction. Melin (1917), however, maintained the view that Material and methods insufficient draining was the cause of the often ex­ perienced poor tree growth. He fo unded his opinion In this study vegetation analyses have been carried upon a number of vegetation analyses where he out on fertilized experimental plots on eleven sites in claimed to have fo und good tree growth on "Fuscum­ the southern part of central Sweden. The plots were bogs", "Vaginatum-bogs" and on other poor sites. analysed in the summer of 1978. In the light of present-day knowledge it can be said that Melin's field observations were not always very thorough (cf. Holmen 1964). It was not until 1935, I however, that Malmstrom clearly showed the I dependence of tree growth on the nutritional status of I • 59 the peat. He reported the results of some experiments '-\,) .52 carried out by Forest Officer V. Ahlund, who had spread woodash on some drained but treeless -- --� peatlands in the province of V asterbotten, N. Ocketbo 0 Sweden. Ahlund's fertilization experiments had a dramatic '• .. , effect on tree growth and they inspired other ex­ • Heby periments in different areas. Some early experiments 40 0 in Sweden were reported by Malmstrom (1952). Fer­ tilization experiments from the 1950s have been reported by C.O. Tamm (1956, 1965), Johansson & Ahgren (1959, 1962), Lindberg (1961) and Berg­ �5- ' strom (1965). Today there are a considerable number of ex­ perimental fertilization plots in Sweden, run by \'j'fi);} various forest companies and scientific institutes. Fig. I. Map of the southern part of central Sweden show­ Fertilization on a commercial scale on peatlands in ing the locations of the experiments.

Acta Phytogeogr. Suec. 68 18 Ingva r B ackeus

Tab 1 e I. Experiments on pi ne bogs run by Tab 1 e I I. L i ndberg' s ferti 1 i zat ion experiments . Oa ta concerning the fert i 1 i zed p 1 ots the Insti tute for Forest Imorovement. (from L indberg 1961). Amounts of ferti 1 i zers ( kg/na) .

peat- year of amount of ferti 1 i zer kg/ha plot no . m above land fertili- site no . M.S.L. type zati on basic slag pot.sulph . salpetre

20 40 60 x) 48 Skogs torpsmossen 140 1955 3750 500 500 40 80 60 Rm ) 46 Klotens Stormosse 290 1953 3750 500 l25oY 80 160 60 Rm 51 Saxbergs Kallmosse 370 Hm 1957 1875 250 375 40 80 59 FAgelsjo 410 Hm 1953 3750 500 80 160 180 52 Bodfl jot 550 Hm 1953 3750 500 85 49 LAngabron 125 Hm 1958 1875 500 44 Jl.tervandsmossen 110 Hm 1953 3750 500 40 Bjurforsangarna lOO Hm 1953 3750 500 58 Stora Hastmuren 70 Hm 1953 3750 500

x) applied in 1955 Hm = Sedge bog y) applied in 1957 Rm = Dwarf shrub bog

The trials on two of the sites were initiated in 1971 In some plots I did not analyse small quadrats but and 1972 by the Institute for Forest Improvement, only made a rough estimate of abundance in the Uppsala. Both sites are ombrotrophic pine bogs on following scale: which plots me�.suring 30 x 30 m were fertilized (see + one single specimen or tussock Table I). There are two series of plots with the same I sparse treatment on each site. Phosphorus was applied as 11 abundant basic slag, potassium as potassium salt and nitrogen Ill dominant as urea. I have also studied nine trials initiated by K. Persson 961) and Malmer (1962) introduced in A. (I Lindberg, the Swedish Forest Service (Doman­ Swedish phytosociological literature the concept 1952-1959. verket), Stockholm, in the period The "characteristic degree of cover", by H. Persson (1975) simplified to "characteristic cover", which since then has treatment of the plots is presented in Table 11. Lindberg published a presentation of the trials with been in common use at the University of Lund and to a preliminary results (Lindberg 1961) and the Forest minor degree at other Swedish universities. As the con­ Service kindly fu rnished fu rther information about siderable limitations in the usefulness of this concept have not been discussed previously, I will take this opportunity the trials. The plots measure 20 x 20 m. On each site I to explain shortly why I have avoided it : have also analysed a non-fertilized control plot. 1. The characteristic cover only gives qualitative infor­ I have not had the opportunity to fo llow any of the mation about the community, as a species occurring in a above-mentioned plots from the start of the trials. low number of squares may get the same characteristic Thus I can only draw conclusions from comparisons cover as a species occurring in many squares. It has been between the fe rtilized and the control plots. The two common practice to combine in tables the characteristic first-mentioned sites give some indications of the cover with the frequency, whereas other information is changes during the first years after fertilization, and omitted. None of these are quantitative concepts. The the other nine sites after a longer period. reader most probably also wants quantitative information In most plots I have analysed nine sample about a community. In that case he is obliged to calculate this information himself. This can be done by multiplying quadrats sized 1 m2, which were laid out in a regular the characteristic cover by the frequency (expressed as a grid. Shoot cover percentage was estimated within fraction rather than a percentage). In so doing he obtains 1, 1-2%, the following classes (class means a cover the mean cover (Etter 1949). of �1 and <2 %; 2 means �2 and <3%, etc.): 2. The (statistical) population, a property of which is supposed to be expressed through the characteristic cover, + < 1 % 20 20-30 % 1-2 % 30 30-40 % is the area in which a species occurs in a community. This l means that when a species occurs in one or a few squares 2 2-3 % 40 40-50 % 3 3-4 % only the population is poorly sampled and the significance, 4 4-5 % 90 90-95 % although impossible to calculate, probably low. (Some 5 5-10% 95 95-100 % authors do not include species with low frequency in their 10-20 % 100 100% tables.) lO 3. The definition of the population given above is only a The mean cover and the frequency have been es­ theoretical one. The area "in which a species occurs" is in timated from the data so obtained. In the tree layer practice dependent on the square size. Therefore the the degree of density is given. characteristic cover is also dependent on the square size. It

Acta Phytogeogr. Suec. 68 Vegetation changes after fe rtilization on drained peatlands 19

is disquieting that the population size is dependent on the study is comparatively homogeneous and the squares sampling technique. larger than those used by Persson, the number of squares When the square size is reduced the variation in the should be high enough to give a reasonably good material will be greater and the relative number of squares significance for the present purpose. where a certain species is not represented will be higher. As The characteristic cover was originally (Etter 1949) the latter squares are excluded from the calculation of the meant to be calculated from values obtained from such characteristic cover, it is not correct when Andersson releves as are used by central European phytosociologists. ( 1970) states that the small size of his squares is compen­ The interpretation of the figures will then be very different sated for by analysing a large number of squares. It would and much of the criticism of the concept will not be ade­ have been correct if he had calculated a quantitative quate. When used on small squares the characteristic measure such as the mean cover. cover can sometimes give useful information about 4. There is a systematic error because of which the sociability. High characteristic cover combined with low characteristic cover in several cases becomes too low. frequency indicates high sociability and low characteristic When a square fa lls on the limit between areas where a cover combined with high frequency low sociability. The species occurs and areas where it is absent, a cover value is figures must, however, be interpreted with great care for obtained varying from zero to a figure which may be the reasons given above. It should also be noted that a fix­ typical of the area inside the limit. The latter figure is the ed square size gives a very different meaning to the "true" characteristic cover but the value obtained will be characteristic cover and to sociability for small and large lower. This error can be quite considerable, e.g., for mat­ species, respectively. It is therefore possible that an estima­ fo rming species with abrupt delimitations such as tion of sociability independent of small squares would be Empetrum spp. and Arctostaphylos uva-ursi. preferable (Poore 1955; Braun-Blanquet 1964; Malmgren 5. The arithmetic mean is a "characteristic" value only 1979). when the property studied has a normal distribution within the population. This is most often not the case as regards cover. For instance, one commonly finds a bimodal dis­ tribution. In such cases the species occurs with high cover Vegetation changes after fertilization in certain parts of the community, whereas it is represented by scattered shoots in other parts. The latter will get too Short-term vegetation changes much influence on the characteristic cover. If there is one Vegetation changes during the first years after fe r­ sq uare with one specimen covering 2 and another % tilization on treeless peatland have been studied square with 50 specimens covering 100 the cha­ %, by Malmstrom (1935), Thurmann-Moe (1956), Pai­ racteristic cover will be 5 1 This means that the single %. vanen & Seppilla (1968) and Heikurainen & Laine sp·ecimen in the first square will get much more influence (1976). o the characteristic cover than any one of the 50 They all fo und a decrease in the coverage of sp·ecimens in the other square. This exaggeration of the im­ Sphagna. An increase of Pleurozium schreberi, p >rtance of the single specimen is still more clearly seen Pohlia nutans and other mosses (excl. Sphagnum) when considering what will happen if it is not sampled. and of Eriophorum vaginatum also seems to be a Then the square is not considered at all and the common feature, as is also the invasion of plants such ch.aracteristic cover thus suddenly rises from 51 to Chamaenerion angustifolium, Agrostis tenuis, % as lOlO %. Marchantia polymorpha, Ceratodon purpureus and In my opinion a quantitative measure must be included in Leptobryum pyriforme. Other changes are difficultto thee tables. In this study I have used the mean cover. It generalise as they are due to peatland type and other sh.ould be noted that the mean cover is not dependent on circumstances. In the investigation of Paivanen & sq uare size, provided an adequate number of squares are Seppala (1968) for example, the vegetation changes an.alysed. If the squares are from the same association in­ caused by fertilization are mixed with changes caused di �idual, some of the often emphasized disadvantages of by the draining undertaken only one year before. usring squares smaller than the minimum area can therefore In this study only ombrotrophic pine bogs have ea�sily be overcome in this way. been investigated. It was not considered meaningful Certainly the mean cover also has drawbacks. A species wi h very low cover can very well be the phytosociological­ to investigate all plots at the two studied sites as ly most important and the most characteristic species of vegetation differences were quite small. thre community. On the other hand a few specimens of a Some notes on the sites are communicated below. large species can reach a high mean cover without being verry important from a differential point of view. Heby: The plots were laid out in 1972 on the pine bog H. Persson (1975, p. 40 ff.) has pointed out that a high Savnemossen, N. of Heby, prov. Uppland. The first series nu:mber of squares is necessary to provide a "reasonable of plots is situated near the eastern limits of the bog and in stcntistic significance" of the mean cover. A much smaller some parts the peat is somewhat minerotrophic. This es­ arr.10untof squares is needed, however, if one only wants to pecially applies to plot 6, which was therefore omitted. The as

A eta Phytogeogr. Suec. 68 20 Ingvar Backeus series is found in the central part of the pine bog. It is Polytrichum juniperinum, P. commune, A ulacom­ regarded as more or less undrained. There is no influence nium palustre and, on hummocks only, Pleurozium of mineral soil water and the vegetation in the area is schreberi. homogeneous. Lindberg's experiments, some of which have been revisited in this study, are located on sites with poor Ockelbo: The plots were laid out in 1971 on Jonsmuren, S. of Ockelbo, prov. Gastrikland. This is a homogeneous om­ or no tree growth. brotrophic pine bog with trees considered to be productive even before fertilization. There are no plots with the Notes on the sites and brief remarks about the vegeta­ highest amount of fertilizers (no. 6) on this site. tion (see also Table FP = fertilized plot, OP = non­ 11; fe rtilized plot) All differences between the differently treated plots are evidently very small (Tables Ill and IV). A Plot no. 51 Saxbergs Kiillmosse (NW. of Grangesberg, decrease in the total cover of Sphagnum spp. seems prov. Dalarna). A sloping soligenous fen traversed by to have occurred, whereas this decrease is difficult to some ditches. Year of ditching unknown but not earlier trace if the species are treated separately. It was also than 1941. The plot is situated near one of the ditches on evident in the field that the vitality of the Sphagna in the lower (drained) side. On the upper, less drained side of the fertilized plots was poor. A corresponding in­ the ditch is a Molinia fen with Sphagnum subnitens (on crease in the cover of other Musci cannot be found in more sloping areas) and a poorer Trichophorum caes­ the Table. pitosum fe n (on less sloping areas). The slope of the plots is similar to the Molinia-dominated areas. In the field layer some species occur in the first 1957 Species list from FP: Betula nana, Calluna vulgaris, series of plots at Heby which are normally not seen Carex sp., Vaccinium uliginosum, A ndromeda polifolia, on ombrotrophic sites. Some of them also grow in the Rubus chamaemorus, Oxycoccus quadripetalus, Cladonia non-fertilized plots, however, and their occurrence rangiferina (in the SE. corner), Sphagnum sp. Designated may be attributed to a small influx of mineral soil as a well-drained sedge mire with spots of dwarf shrubs. water as well as to the fertilization. They do not occur On FP 16 pine and 6 spruce plants, 0. 1-0.3 m. in the purely ombrotrophic second series at He by, 1960 Rather good development of the trees. not even after heavy fertilization. 1978 FP only thinly populated by trees. A fe w high pines The tendency is rather clear that the number of and spruces up to 12 m. One big birch and a number of species gets smaller in the fertilized than in the non­ smaller birches, spruces and pines. Rather a large number of small plants of birch, a few spruces and pines. Calluna, fe rtilized plots. This especially applies to the bottom Eriophorum vaginatum and Polytrichum commune layer. Invasion of species not earlier present has oc­ dominate. The latter species occurs only very little outside curred to a very small degree. The occurrence of such FP. Cladonia rangiferina in the SE corner has completely species (Chamaenerion angustifolium, Calamagros­ disappeared, but it appears sparsely along the northern tis arundinacea and Leptobryum pyriforme) is more limit. In OP the trees are still more thinly distributed. or less occasional. The results are in agreement with Biggest pine about 10 m. Birches, spruces and most pines Malmstrom's (1952) opinion that the vegetation only reach 3-4 m. Small scattered plants of pine (more changes in dwarf shrub communities are small com­ than in FP) and birch, a few spruce plants. Calluna and pared to those in other communities. Sphagnum spp. dominate. It is evident that the short-term cover changes in Plot no. Klotens Stormosse (Kloten, prov. Vast­ this community are so small that they are difficult to 46 manland). Ditched several times 187 5-1925. In 1925 it detect by comparisons between differently treated was very effectively drained (see Lundqvist 1930). plots. Permanent plots which are followed fr om the 1953 Calluna bog with scattered stunted pines, ea. 50-75 time prior to the fe rtilizer treatment are necessary. years old and up to 6 m tall. The results, of course, are only attributable to the 1958 FP fenced against moose. mire type studied. On a tree-less peat land with a less 1960 Rather good development of the trees. close layer of dwarf shrubs the changes would 1978 The fence removed. In FP the pines are up to 13 m, probably have been greater. outside the plot generally somewhat smaller. Small spruces and a few small birches in FP, the former also outside the plot. Sphagna less common and other mosses more com­ Long-term vegetation changes mon in FP than in OP.

The vegetation 15 years after fertilization on a Plot no. Skogstorpsmossen (S. of Villingsberg, prov. peatland in V asterbotten, N. Sweden, was studied by 48 N arke) . Drained 1925-1 930 through a road ditch Malmstrom ( 193 5). The original tree-less, poor (probably suboptimal for tree growth). vegetation had been replaced by a birch fo rest with a 1955 Pine bog with ± continuous Sphagnum - Calluna sparse field layer and a bottom layer with hummocks. Sparse Eriophorum vaginatum, Rubus

A eta Phytogeogr. Suec. 68 Vegetation changes af ter fe rtilization on drained peatlands 21

Table Ill. Vegetation analyses in experimental plots on pine bogs fertilized in 1 971 and 1972 respectively. Small quadrat frequency (quadrat size 1 m2 ) is given in percent with the mean cover (sic!) as the exponent. 0° denotes species occurring within the plot but not within the small quadrats . In the tree layer an esti- mated cover for the whole plot is given.

site Ockelbo Heb series 1 1 plot 85 85 85 4 85

Tree layer Be tu 1 a pubescens + + + Picea abies + Pi nus syl vestris 70 80 60 60 70 Ill Ill 80 Ill 50 40 Ill total 70 80 60 60 70 80 50 40

Shrub layer 5 o o Be tu 1 a pubescens 67 ·r o o Picea abies 11+ oo 11+ oo oo + o + + Pinus syl vestri s 33 o 11 11 II

Field layer + + + Andromeda pol ifolia 11 33+ 11 22 + Betula pubescens 33+ + 11 + Be tu 1 a pubescens (seedlings) 22 1 2 3 + 1 1 4 Cal luna vulgaris 56 44 56 56 44 221 100 100 II 0 Empetrum nigr um 671 1005 li 5 Empetrum sp. 44+ 67+ 1 1 + 67 44+ 5 5 0 5 Le dum pa 1 us tre 100 1002 1005 1002 67 II 671 II 891 89 li

Oxycoccus microcarpus 22+ 22+ + + + 0. quadri petalus 44 + + II 100 78 + Picea abies 11 + + + + Pinus sylvestri s 11 11 11 33+ + + P. syl vesttis (seedl ings) 22 11 Sal ix cinerea 4 1 0 3 2 5 1 Vacci ni urn my rti 11us 78 100 78 100 100 II II 78 22+ 20 20 1 0 20 10 5 + V. uliginosum 100 100 100 100 44 Ill Ill 44 225 11 Ill 4 5 4 4 V. vitis-idaea 1002 893 100 100 100 II II 100 II 11+ 1003 II

o Chamaeneri on angus ti fo 1 i urn o Drosera rotundi fo 1 i a 22+ Me 1 ampyrum pratense 11+ 11+ + 2 4 1 0 3 L 1 0 Rubus chamaemorus 22 33+ 67 89 89 II II 78 III 89 100 II 0 Cal amagrostis arundinacea 0 Carex nigra 671 II 11+ 4 + + + + 92 1 1 Eriophorum vaqi natum 22 22 33 11 8 II II 100 II 100 100 II Luzula pilosa 22+ Poa palustris +

Bottom 1 ayer + + + Aul acomnium pa lustre 33+ 33+ 11 44 33+ + 33+ 44 + Di cranum fuscescens 22+ 22+ 11+ + 33+ 22+ + D. maj us 11+ X 50 20 2 4 5 4 + D. polysetum 100 100 100 100 100 III III 100 33+ 22 II D. scopari um 11+ + D. undulatum Brid. 22. 67 + ci Dicranum sp. o Drepanoc 1 ad us unci natus 11+ + + + + Hyl ocomi um splendens 11 11 22+ 67 44

A eta Phytogeogr. Suec. 68 Ingvar Backeus 22 Table Ill (contin.) site Ockelbo Heb series 1 2 1 plot 85 85 85 4 85 + Leptobryum pyri forme 33 + + Plaaiotheci um cf. denticulatum 11 10 10 920 20 10 1 + 2 Pleurozium schreberi 100 100 8 100 100 11 Ill 100 11 22 67 11 + + + + Poh 1 i a nutans 33 I 56 + 78 56 + Pohl ia sp. cfr. 11 + Po 1 ytri chum co11111une 11 + + 3 + + P. juniperi num s.l. 11 22 22 + 67 78 + o Tetraphis pel lucida 22 a + 30 40 20 20 30 5 2 3 total Musci ( exc 1 . Sphagna 1 es l 100 100 100 100 100 100 100 100 + 0 30 0 Sphagnum fl exuosum v.angusti fol . 11 9 01 0 8 10 II + s. flexuosum s.l. 22 Ill + + 1 s. fuscum 11 + 56 33 + s. gi rgensohni i 22 + + + o 1 + S. ,mage1 1anicum 11 11 33 o 89 33 + 5 + S. nemoreum 332 44 44 11 + 5 + 3 0 total Sphagnum 442 44 44 11 441 ao 89 o 1001

+ Calypogeia sphagnicola 11

Lophoco 1 ea heterophy 11a + + + Myl ia anomala + 33 22 + Ptilidium ci liare 11

0 + + + Cl a doni a deformi s col l. 0 ao 11 oo 11 11 + + + + + 7 1 c. cf. pyxidata 11 22 22 22 + 22 8 II o o o + + 2 + c. rangi feri na a a a 11 67 78 44 2 1 + + + 1 c. syl vatica co1 1. 44 56 44 22 II 44 100 c. coenotea - + + c. corn uta 67 1 � c. grayi } + + Cladonia sp. 11 11 } + Stereocaulon sp. 11

Table IV. Number of species in the same plots as in Table Ill. In the pai red figures the left figure denotes the total number of species and the right figure the number of species which are in co11111on with the control plot (plot 85). site Ockelbo Heb series 1 1 2 �lot 85 85 85 2 4 85 6 field layer 10 9 : 7 9 9 :8 14 12:12 13:11 14:11 13:11 11 13:10 11:9 bottom layer 12 7 :6 13 8:7 17 11:10 13:11 20 :13 13:11 1 7 15:12 18:14 total 22 16:13 22 1 7 :15 31 23:22 26 :22 34 :24 26 :22 28 28:22 29 :23

Acta Phytogeogr. Suec. 68 Vegetation changes af ter fe rtilization on drained peatlands 23

m ig. 2. Plot no. 59 Fagel jo. (a) The fertilized plot in 1953. IPhoto: The Swedish Forest Service. (b) The fertilized plot im 1978. Po�rtrichum commune dominating.

chamaemorus and Vaccinium uliginosum. Scattered s;tunted pines, 0.5-4 m, a fe w dwarfed birch shrubs and swme spruce plants. 11 960 The trees have reacted poorly to the fertilization. 11 978 The pines both within and outside FP are now ea. 10 rrn.Th ey have reacted very little to the fe rtilization. Small s;pecimens of spruce and birch are numerous in FP, the batter .n uch less so outside the plot. There is very little §phagnum in FP but plentiful Pleurozium schreberi. The liirnit is very sharp and easily visible. b !Plot no. 58 Stora Hd.stmuren (Valbo, prov. Gastrikland). Wear of ditching unknown. Original mire type also un­ k nown, but obviously some kind of fe n. 1953 Poor tree growth. Scattered Calluna hummocks 11 953 No notes on the vegetation. Spruces chlorotic before with Sphagnum, Polytrichum, some Betula nana, Vac­ fc'ertilization. Chlorosis disappeared after fertilization. cinium uliginosum and small Juniperus. Between them 11957 Thinning. Before thinning there were 49 pines, 22 mostly Scirpus sp. Miserable pine shrubs, 0.5-3 m, s;pruces and 67 birches in FP. After thinning there probably 50-60 years old. rcemained 23, 20 and 23 respectively. 1960 Rather good development of the trees. Calluna 11960 Very good development of the trees. 1978 In FP a mosaic of low hummocks with 11 967 First report on chlorosis within FP. Betula nana and a net of depressions with Polytrichum 11 978 In FP pine and birch up to 20 m. Under them, dense commune. Trees predominantly on the fo rmer, mostly s;pruce up to 10 m. The spruces are severely chlorotic, pines of various height (up to 10 m), in parts rather thin. rmost of them with many dry twigs, fe w needles and often Small shrubs of pine and scattered birch (both species con­ ctlead apexes. Sparse small plants of spruce. In the fieldand siderably browsed). OP is not quite homogeneous. Trees bJottom layer cover is very low. OP is heterogeneous. It is in only on hummocks and concentrated to the W. part of the pJarts similar to FP but with smaller trees (pine, spruce and plot. A fe w pines reach 6 m. Scattered small birches. In the bJirch), and in parts only scattered pines with no under­ E. part drainage is somewhat less efficient and here are vwood. In the latter parts, plentiful Cladonia spp. On the fo und small tussocks of Tr ichophorum caespitosum and vwhole, field and bottom layers are sparse. Molinia surrounded by naked peat. Sphagna dominate in the bottom layer.

!Plot no. 59 Ftigelsjd (NE. part of prov. Dalarna). A Originally a flark mire with high-hummocky strings. The Plot no. 44 tervd.ndsmossen (Gronbo, prov. Vast­ fllarks were drained in about 1910 and then more efTective­ manland). Drained in 1950. l)Y drained in 1946 when a road was built through the area. 1953 Plenty of sedges and grasses, some fo rbs, scattered Tihe plot is situated in a rather narrow strip in the mire Phragmites, some Calluna hummocks, Myrica, Betula \Where large flarks do not occur but only smaller mudbot­ nana and Scirpus sp. No Ledum. Scattered, small poor tcoms. pine shrubs and birches, 0.5-2 m. A few small Juniperus,

Acta Phytogeogr. Suec. 68 24 Ingvar Backeus

a Fig. 3. Plot no. 40 Bjurforsangarna. (a) The fertilized plot in 1953. Photo : The Swedish Forest Service. (b) The fe r· tilized plot in 1978.

b

one Salix caprea and a couple of miserable spruces. Molinia is not mentioned but probably dominant. 1957 A moose fence put up. 1960 Very good development of the trees. 1969 Chlorosis on spruce noted. 1978 Tree layer in the fertilized plot 10-12 m high with small glades in the E. part. Birch dominating, but also spruce and pine occur. Small plants of spruce and birch. Spruces chlorotic. Molinia still with high frequency. Trien­ talis also common. Bottom layer sparse. In OP a few pines have reached 2-3 m, others still only small shrubs. Also birches up to 2 m. A few, small spruce shrubs. Betula nana, My rica and Molinia dominating. A part of OP bordering FP is excluded, as it is evidently influenced by the fertilizers.

Plot no. 40 Bjurforsiingarna (Bjurfors, prov. Vast­ manland). Ditched 1925-30. Another ditch added in 1953. 1953 Plenty of Molinia and My rica, some Cal/una and Scirpus sp. Scattered weak pine shrubs, up to 4 m, weak birch shrubs, also up to 4 m (somewhat browsed by Fig. 4. Plot no. 40 Bjurforsangarna in 1978. In the fo re­ moose) and 4-5 small, miserable spruce plants. ground the control plot with Molinia caerulea, in the 1958 Deformed pines and numerous birch shrubs background the fertilized plot. removed.

A eta Phytogeogr. Suec. 68 Vegetation changes after fe rtilization on drained peatlands 25

L60 Very good development of the trees. described by Malmstrom (1935), where the fertilized 1578 In FP there is now a dense fo rest with pine, spruce area is densely forested and the non-fertilized com­ ar:d birch up to 11-12 m. Lower trees are mostly spruce pletely devoid of trees, the second cause should be the and some birches. Small plants (spruce) are sparse. most important one. Srruces chlorotic. Field and bottom layer sparse except in In those of the studied plots where the trees have small glades where Molinia is dominating (hence the high not reacted significantly to the fertilization, the mean cover of Molinia). In OP scattered pines and birches (u;>to 7 m) are fo und, as well as a few small spruce shrubs. vegetation changes must be attributed to the change M linia is dominating. Also Myrica and Calluna. in nutritional conditions only. This is the case on plot no. 48 Skogstorpsmossen. Only small or moderate

Pl t no. 49 Langabron (near Laxa, Narke). Original mire reaction occurred on no. 46 Klotens Stormosse, no. tyj)e not known, presumably a Molinia fen. Ditched 51 Saxbergs Kallmosse and no. 59 Fagelsjo. ar und 1920. Few differences in the field layer between the fer­ 1958 Clearing undertaken before fertilization. Well­ tilized and non-fertilized plots are common to these dr ined according to Lindberg. Vegetation : Calluna sites. Dwarf shrubs, especially Ca/luna vulgaris and vulgaris, Juniperus communis, Myrica gale, Sphagnum Va ccinium uliginosum, seem to be less common in sp., Oxycoccus quadripetalus, Carex sp. (probably the fertilized plots. On Klotens Stormosse and Sax­ Mo/inia), A ndromeda polifolia, Erica tetralix. Before the clearing the area was presumably similar to what is now bergs Kallmosse, Eriophorum vaginatum is con­ found around the plots, i.e. a thin birch fo rest with small siderably more prevalent on the fe rtilized areas. In spruces, some junipers and a field layer completely the bottom layer it is evident that Sphagnum spp. are dominated by Molinia. Erica tetralix occurs. less common and other Musci more common on the 1960 Rather good development of the trees. fe rtilized plots. On Saxbergs Krulmosse and at 1961 FP fe nced against moose. Fagelsjo the fertilized plots have been invaded by 1978 In FP a dense tree layer of birch and a few pines. Polytrichum commune. Tree height ea. 10 m. Scattered small spruces. Field layer On the other sites (no. 49 Langabron, no. 44 sparse, predominantly Molinia, especially in small glades. Atervandsmossen, no. 40 Bj urforsangarna, no. 52 Bottom layer sparse. Plenty of dead Molinia tussocks. In Bodfljot and no. 58 Stora Hastmuren) a dense fo rest OP no trees. Rather small birches (up to 2 m) and some small pines. Molinia dominating. Drainage probably not or thicket has grown up causing entirely different quite sufficient. conditions for the ground vegetation. A considerable portion of the original species in the fe rtilized plots on

Plot no. 52 Bodjljot (Bunkris, NW. part of prov. Dalar­ these sites has disappeared and has been replaced by na). Drained about 1910. Originally a poor fl ark mire. other species. It must be assumed that the direct 1953 Dominance of Scirpus sp. Some hummocks with effect of the fertilization on the fieldand bottom layer Calluna, Vaccinium uliginosum and Betula nana, a fe w in these cases is insignificant compared to the role poorly developed pine shrubs, 0.5-4 m, and scattered played by the diminishing light. weak birch and pine plants. Some Polytrichum and lichens Small specimens of Pinus, Betula pubescens and in on some of the hummocks. some cases Picea are fewer in the fertilized plots. An­ 1957 Moose fe nce raised. Very good development of the trees. dromeda polifolia, Betula nana, Ca/luna vulgaris, 1956-1960 Empetrum spp., Erica tetralix, Myrica gale, Oxycoc­ 1964-1966 Serious damage to the pines (most of them killed) by snow pressure fo llowed by attacks of the fungi cus quadripetalus, Vaccinium uliginosum, Erio­ Scleroderris lagerbergii and Cenangium abietinum. Birch phorum angustifolium, E. vaginatum, Molinia shrubs and willow take over. caerulea and Trichophorum caespitosum are 1978 Moose fe nce partly ineffective. In FP some rather drastically less common or entirely absent in the fer­ small pines and a few larger ones (up to 9 m). One single tilized plots. These species are mire plants which were large spruce. A dense thicket of birch (ea. 5 m, a few up to able to survive the draining. Some of them are even 8 m) and Salix cf. starkeana. Field layer sparse. In the bot­ enhanced by the disappearance of more hygrophilous tom layer only Polytrichum commune common. In OP one mire plants. Molinia caerulea, in particular, is never birch and a few pines. Several birch shrubs, probably due ' to the fencing (fewer outside the fence). Also some seen in such quantities on virgin mires as it is on scattered pine plants. Dwarf shrubs, Polytrichum spp. and drained ones. The conclusion that can be drawn here Sphagnum nemoreum dominating. Plentiful Cladonia spp. is that many less hygrophilous mire plants can with­ stand draining quite well but are severely damaged by When discussing the results (Tables V and VI) two the establishment of a tree layer. different causes for the long-term vegetation changes Concerning the bottom layer in these plots, the must be considered, namely the improved nutritional diminishing of Sphagna is evident and easily ex­ conditions and the influence caused by the growing plained. Other Musci are more plentiful in the fe r­ trees. In such an extreme example as the one tilized than in the non-fertilized plots, as regards both

A eta Phytogeogr. Suec. 68 Ingvar Backeus 26

9 ertilized plots . OP = non- Table V. Vegetation analyses in experimental plots fertrl ized in the 1 50s (Table IIl_, FP f -ferti lized plots . For other explanations , see Table Ill.

site no . 51 46 48 58 59 44 40 49 52__ _ _ Saxbergs Kl otens Skogs- Stora Ater­ Bjur- site name Ka1 1- Stor- torps- Hast- FAgel - vanc!s­ fors- LAnga- mosse mosse mossen muren mossen angarna bron Bodfljot � plot OP FP OP FP OP FP OP FP OP FP OP FP OP FP OP FP OP FP

Tree layer 7 9 Be tu 1 a pubescens + 20 + 30 60 + + 10 80 20 70 + 0 10 0 Picea abies + + + + + 50 + 20 30 + + Pinus syl vestris 20 30 30 40 40 50 50 50 10 50 10 10 + 30 + 20 10 + total 20 50 30 40 40 50 70 90 10 50 20 90 30 90 + BO 20 90

Shrub layer Betula nana x pubescens 1 2 + B. pubescens 56 44 11 10 5 Juni perus co11111uni s 44 11 1 1 1 1 1 5 10 1 2 • 5 • Picea abies 22 11 22 11 11 67 67 33 11 11 56 11 1 1 4 2 4 2 1 + 1 1 Pinus s'yl vestris 22 22 44 22 67 44 78 44 44 33 Rh'amnus frangu 1 a Sal ix lapponum S. phyl icifolia S. cf. starkeana

Field layer + + + + • + + + + Andromeda po 1 i fo 1 i a 56 44 22 33 11 67 56 89 44 5 3 1 1 10 10 10 10 + Betula nana 1 oo 56 67 56 - 100 100 100 - - 100 33 + B. nana x pubescens 11 + + + B. pubescens 78 22 11 + + + + B. pubescens (seedl ings) 22 33 11 44 60 10 50 10 0 0 1 30 4 10 + 5 Ca11una vulgaris 1 oo 10o 89 89 1006 896 44 - 1 oo 10o 89 - 782 11 100 + 3 Empetrum hermaphroditum 22 33 Empetrum sp. 1 Erica tetralix - 100 Juni perus co11111uni s Ledum pa lustre Lycopodium clavatum L. selago Myr'iea ga 1 e Oxycoccus mi crocarpus • • 3 0. quadri petalus 22 11 1oo + • • • Picea abies 22 33 33 oo 33 + P. abies (seedl ings) 11 • • • Pinus sy lvestris 22 11 11 + + P. sylvestris (seedl ings ) 11 22 Rhamnus frangula Salix cinerea S. phyl icifolia S. repens S. cf. starkeana o Sorbus aucupa ri a o 10 0 + + Vaccinium myrti llus 67 o 56 11 2 + 10 4 10 1 4 + 1 1 V. uliginosum 33 11 89 56 44 33 22 11 44 33 V. vitis-idaea

Acta Phytogeogr. Suec. 68 Vegetation changes af ter fe rtilization on drained peatlands 27

Table V (contin.)

site no . 51 46 48 Sl3 59 44 40 49 52 Saxbergs Klotens Skogs- Stora Ater- Bjur- site name Ka 11- Star- torps- Hast- FAgel- vands- fors- LAnga- mosse mosse mossen muren mossen an2arna Bodfljot � plot OP FP OP FP OP FP OP FP �OP FP OP FP OP FP OP FP OP FP + Chamaenerion angustifolium 11 0 Drosera rotundifol ia 0 0 0 Dryopteri s s pi nu 1 os a 0 0 0 Me 1 ampyrum pra tense 44+ 11+ 0 22+ 33+ 0 Potenti lla erecta 0 78+ 33+ 22+ 22+ 22+ 11+ Ramischia secunda 11+ 5 5 5 0 0 Rubus chamaemorus 67 89 100 1001 1001 - 331 22+ 0 Sol i dago virgaurea 0 Triental is europaea 22+ 33+ 67+ 22+ 89+ 22+ 44+ 0 Viola palustris 33+ 0 0 V. riviniana 0

Agrosti s canina 22+ 0 Ca rex canescens 0 + + c. globularis 33 11 + c. lasiocarpa 22 + + c. nigra 11 44 0 + c. paucifl ora 0 11 + c. rostrata 11 Des champs i a flexuosa 22+ Equi setum fluviati le 33+ E. palustre 11+ E. pratense 22+ E. syl vaticum 33+ Eriophorum angusti folium 44+ 11+ 56+ 11+ 0 E. vagi.natum 1001 1001 11+ 1003 78+ 89+ 56+ 11+ 89+ 67+ 33+ 11+ 44+ 22+ 56+ 11+ 56+ Luzula pilosa 11+ 5 4 0 5 80 50 0 Mo l inia caeru lea 44 56 - 1002 78+ 1003 100 100 783 100 1001 - Phragmi tes cofl11lunis 11+ 44+ 78+ Tri chophorum a 1 pi num 22+ 0 T. caespitosum 562 0 561 33+ 44+ 893 671

Bottom layer Amblystegium serpens 11+ Aulacomnium palus tre 33+ 891 33+ 22+ 33+ 11+ 11+ 22+ 22+ 22+ 33+ 44+ 89+ 33+ 56+ Brachythecium populeum 22+ Cal liergon sarmentosum 22+ C. strami neum 44+ 22+ 332 Cirri phy11um pi liferum 11+ Di cranum fuscescens 22+ 22+ 11+ 56+ 67+ 22+ D. majus 22+ 22+ 5 D. polysetum 11+ 671 781 56+ 100 441 33+ 44+ 441 22+ o D. scopari urn 56+ 44+ 44+ 33+ 11+ 22+ 33+ a D. undulatum Bri d. 11+ 56+ Hylocomium splendens - 22+ 33+ 11+ 22+ o Plagiomnium affine a

A eta Phytogeogr. Suec. 68 Ingvar Backeus 28

Table V (contin.)

site no . 51 46 48 58 59 44 40 49 52 Saxbergs Klotens Skogs- Stora Ater- Bjur- site name Kal l- Stor- torps- Hast- FAgel- vands- fors- LAnga- mosse mosse moss en muren moss en angarna bron Bodf)jot � plot OP FP OP FP OP FP OP FP OP FP OP FP OP FP OP FP OP FP

Plaqiothecium denticulatum ll+ ll+ ll+ 33+ 44+ 891 - 1001 56+ 44+ s 0 0 s Pleurozium schreberi 56 673 892 892 892 1001 1001 78+' 6 73 78 44+ 56+ 782 44+ 561 Pohlia nutans s.l. 67+ 33+ 22+ ll+ 22+ 33+ 33+ ll+ ll+ 33+ 67+ 89+ Polytrichastrum longisetum ll+ o o 0 0 Po lytri chum commune ll+ 89 1 - o 7830 - 22+ 22+ 221 100 1 0 0 0 0 4 P. juniperinum s.l. 44+ 44+ 33+ 0 0 22+ 0 22+ 563 0 78 44+ protonema 22+ s 0 4 0 0 0 4 40 4 4 7 0 0 total Musci (excl .Sphagnales) lOO 1002 100 1003 89 2 l00 1 loo 1 loo3 89 100 89 100 22+ 100 44+ 1002 100 1 1002

Sphagnum compactum ..+ 44 1 0 2 1 0 + s 3 0 S s. plexuosum s.l. 56 22+ 56 22+ 89 33 22 33 0 - lOO ll+ ll+ 1 + s s. fuscum 22 ll 33 22 112 ll+ ll+ gi rgensohni i 1 �- 22 22+ 0 0 s. imbricatum 892 - 0 + + 0 0 s. magel lanicum ll 44 0 0 33+ ll+ 33+ 33+ 1 0 s. magel lanicum +papi l losum 78 0 0 + + s s. nemoreum 0 0 ll ll 78 s. palustre 22+ ll+ 33+ s. papi l losum ll+ ll+ s. rubellum ll+ 22+ s 0 s. russowi i 67 0 1 + s. subni tens 67 44 673 s. subsecundum 33+ s. sp. Sect. Palustria 22+ Sphagnum sp. ll+ 30 3 0 0 0 0 0 0 s total Sphagnum 100 67+ 56 331 891 33+ 781 ll+ 1002 22+ 892 - 0 ll+ 1001 33+ 89 ll+

Cephalozia bicuspidata 22+ ll+ ll+ C. cf. connivens ll+ ll+ Cephaloziella spp. 22+ 22+ 33+ Gymnoco lea i nfl a ta 22+ Kurzia paucifl ora ll+ Lophoco lea heterophy lla ll+ ll+ ll+ 33+ 671 Lophozia sp. ll+ 0 0 Myl ia anomala 0 0 Pti lidium pul cherrimum 22+

0 Cetraria isl andica 0

Cladonia crispata 22+ 0 c. deformi s coll. ll+ 0 c. digitata 33+ o 0 c. rangi feri na o 22+ ll+ ll+ 0 22+ o o o 0 c. syl vatica col l. o o o 22+ 0 67+ 22+ 673 o c. subfurca ta o + + c. cf. pyxidata ll ll ll+ 22+ ll+ 44+ ll+ ll+ Cladonia spp. ll+ ll+ ll+ oo

A eta Phytogeogr. Suec. 68 Vegetation changes aft er fe rtilization on drained peatlands 29

Table VI. Number of species in the same plots as in Table V. For explanations see Table IV.

Saxbergs Kl a tens Skogs- Stara Ater- Bjur- site Ka l l- Star- torps- Hast- Fagel - vands - fors� Langa- mosse mosse moss en muren moss en angarna bran Bodfljot plot OP FP OP FP OP FP OP FP �OP FP OP FP OP FP OP FP OP FP

fiel d layer 13 12:9 11 14:11 10 10:9 14 13:6 15 24 :10 13 13:7 10 10:6 14 9:6 14 16:7 bottom 1 ayer 15 13:11 12 18:11 9 8:5 17 16:12 16 9:6 13 1 3 :5 4 12:1 7 1 3:2 15 9:4 total 27 25 :20 23 32:22 19 18:14 31 29 :18 31 33:16 26 26 :12 14 22 :7 21 22 :8 29 25 :11

cover and number of species. This clearly applies to considerably different inside and outside the fe rtilized Plagiothecium denticulatum, Pleurozium schreberi, plots, the latter factor is evidently predominant. The Pohlia nutans and, at Bodfljot, Polytrichum com­ vegetation then becomes more or less similar to that mune. The Stora Hastmuren site is here, as in other fo und in moist forests on mineral soil. respects, somewhat deviating. The reason is evidently On a few sites the trees have reacted only slightly that also the non-fertilized plot of this site has a fairly to the fertilization. The observed vegetation dense tree layer. In the fertilized plot it has become so differences must thus be attributed to the improved dense that all kinds of ground species are hampered. nutritional conditions, the non-existence of Sphagnum spp. in the fertilized plots being the most important result. SUMMARY Draining of peatlands has oftenproved to be an in­ Acknowledgements sufficientmeasure for obtaining good tree growth. As I am particularly indebted to the Institute for Forest Im­ was shown by, i.a., Malmstrom (1935), the provement, Uppsala and the Swedish Forest Service, nutritional status of the peat often becomes a whose experimental sites have been used in the investiga­ delimiting factor. To obtain a better understanding of tion. Financial support from the National Swedish En­ the role of different nutrients on various kinds of vironmental Protection Board is gratefully acknowledged. peatlands, fe rtilization experiments were started in I am also indebted to Dr. Roland Moberg, Uppsala, who many places around Sweden. determined some lichens and to Dr. Erik Sj ogren, Uppsala, In this study the ground vegetation on fertilized who determined a number of mosses. plots at some of these sites in central Sweden was analysed. The aim was to investigate changes that REFERENCES had occurred due to the fertilization. The vegetation in the fertilized plots was therefore compared with the Andersson, F. 1970. Ecological studies in a Scanian vegetation in neighbouring non-fertilized control woodland and meadow area, southern Sweden. I. plots. Vegetational and environmental structure. - Op. bot. Short-term changes were studied on ombrotrophic Soc. bot. Lund. 27:1-190. Bergstrom, B. 1965. Om resultaten av nagra fOrsok med pine bogs only. Two sites were investigated. The beskogning av mossmarker i Halland. - Svenska changes were strikingly small. A decrease in the total Skogsvardsforen. Tidskr. 63:367-377. cover of Sphagna as well as in their vitality seems to Braun-Blanquet, J. 1964. Pflanzensoziologie. Grundziige have occurred. A few species which are normally not der Vegetationskunde. 3. Aufl. - Wien. 865 pp. fo und on ombrotrophic sites have invaded but they Etter, H. 1949. De !'analyse statistique des tableaux de only occur more or less sporadically. The changes in vegetation. - Vegetatio 1:147-1 54. the vegetation would probably have been greater on Heikurainen, L. & Laine, J. 1976. Lannoituksen, other peatland types. The fairly closed tree layer and kuivatuksen ja lampoolojen vaikutus istutus- ja luon­ the dense layer of dwarf shrubs presumably prevent nontaimistojen kehitykseen rameilla. (Effect of fer­ the establishment of other plants. tilization, drainage, and temperature conditions on the development of planted and natural seedlings on pine Long-term changes were studied on nine sites with swamps.) - Acta fo r. fe nn. 150:1-38. plots fertilized in the 1950 s. Here the changes in the Holmen, H. 1964. Forest ecological studies on drained ground vegetation were often profound. Two peat land in the province of Uppland, Sweden. 1-lll. ­ different causes of the changes must be considered, Stud. for. suec. 16:1-236. the improved nutritional conditions and the influence 1979. Skogsgodsling i Sverige 1978. - K. Skogs- Lant­ of the growing trees. In cases where the shading is bruksakad. Tidskr. 118:349-355.

A eta Phytogeogr. Suec. 68 30 Ingvar Backeus

A A Johansson, B. & gren, A. 1959. Godsling av skogsmark. Persson, . 1961. Mire and spring vegetation in an area Intryck fdin studieresor i Norge och Tyskland samt north of Lake Tornetriisk, Torne Lappmark, Sweden. redogorelse for godslingsforsok inom SCA. - Norrl. I. Description of the vegetation. - Opera bot. Soc. bot. Skogsvardsforb. Tidskr. 1959:381-424. Lund. 6( 1):1-187. 1962. Godsling av skogsmark. II. Resultat av fdrsok Persson, H. 1975. Deciduous woodland at Andersby, utlagda inom SCA 1957-58. - Ibid. 1962:131-174. Eastern Sweden: Field layer and below-ground produc­ Lindberg, K. 1961. Om godsling av torvmarker. - Ibid. tion. - Acta phytogeogr. suec. 62:1-71. I96I:135-1 58. Poore, M.E.D. I 955. The use of phytosociological Lundqvist, G. 1930. Jordlagren. - Sver. Geol. Unders. ser. methods in ecological investigations. II. Practical Aa 168 :68-161. issues involved in an attempt to apply the Braun­ Malmer, N. I 962. Studies on mire vegetation in the Blanquet system. - J. Ecol. 43 :245-269. archaean area of southwestern Gotaland (South Reinikainen, A. 1965. Vegetationsuntersuchungen auf Sweden). - Opera bot. Soc. bot. Lund. 7( 1):1-322. dem Walddi.ingungs-Versuchsfeld des Moores Kivisuo, Malmgren, U. 1979. Forslag till metodik for oversiktlig Kirchsp. Levionmiiki, Mittelfinnland. - Commentat. vegetationsanalys. - Svensk bot. Tidskr. 73 :9-25. Inst. fo r. fenn. 50(5):1- 62. Malmstrom, C. 1935. Om naringsforhallandenas betydelse Tamm, C.O. 1956. Studier over skogens niiringsfor­ fo r torvmarkers skogsproduktiva fo rmaga. (Uber die hallanden. IV. Effekten av kalium- och fo sfortillforsel Bedeutung der Nahrstoffbedingungen fi.ir das wald­ till ett oviixtligt bestand pa dikad myr. (Studies on produktive Vermogen der Torfboden.) - Meddn St. fo rest nutrition. IV. The effects of supply of potassium .Skogsforsoksanst. 28:571 -650. and phosphorus to a poor stand on drained peat.) - 1952. Svenska godslingsforsok for belysande av de Meddn St. Skogsforskn. Inst. 46(7): 1-2 7. niiringsekologiska villkoren for skogsviixtpa torvmark. 1965. Some experiences from fo rest fertilization trials in - Commentat. Inst. for. fenn. 40( 17):1-27. Sweden. - Silva fe nn. 117(3):1-24. Melin, E. 19 I 7. Studier over de norrliindska myrmarker­ Thurmann-Moe, P. 1956. Eldre og nyere skogkultur- og nas vegetation med siirskild hansyn till deras gj0dslingsfors0k pa Asmyra. 1-II. - Norsk skogbruk skogsvegetation efter torrliiggning. - Norrliindskt 1956:269-281, 309-3 16. Handbibl. 7:1-428. Tiberg, H.V. 1908. Skogsj ordsanalysen och jordens Osvald, H. 19 I 7. Nagra anteckningar rorande skogsviixt produktionsformaga. - Wermliindska Bergsman­ pa Granarps mosse. - Svenska Mosskulturforen. naforen. Ann. 1907 :230-277. Tidskr. 31:178-1 97. Tubeuf, C. von 1908. Di.ingungsversuch zu Kiefern auf Piiiviinen, J. & Seppala, K. 1968. Hajalannoituksen Hochmoor. - Naturw. Z. Forst-Landwirtschaft 6. vaikutus lyhytkortisen nevan pintakasvillisuuteen. (Effect of broadcast fertilizer on the ground vegetation of a low sedge swamp.) - Suo 19:5I-56.

lngvar Backeus, Institute of Ecological Botany, Uppsala University, Box 559, S-75 1 22 Uppsala, Sweden

Acta Phytogeogr. Suec. 68 The History of Woodland in Fnjoskadalur Agust H. Bj atnason

General survey (each about 10 spp.). Other families have less than 10 species and in more than 20 families there are only 1 or 2 Physiognomic description species. There are no great differences in the species composi­ Iceland lies on the North Atlantic Ridge, between tion of the plant communities in the various regions. the latitudes of 63°24' and 66°32'N and longitudes of However, there are marked differences in the distribution 13°30' and 24°32'W. It has an area of 103 125 km2• The of the species by elevation, although up to 300 m there are, island is mountainous and 7 5 % of the land area is above generally speaking, few differences. The exceptions being 200 m elevation. The major part of this area consists of a maritime and weed species. 500 to 700 m high plateau, from which rise mountains, the The above enumeration does not give a correct picture highest point being 2 119 m. About 1 1.5 %of the land area of the vegetation, as grasses and low shrubs are the is covered by glaciers. characteristic species. There are no reliable figures avail­ Geologically, Iceland is young with frequent volcanic able fo r the exact extent and condition of the vegetation, activity. Volcanic eruptions have often caused severe but it is likely that the total area is less than 25 000 km 2• damage by ashfall, lava or glacial burst. Since the country Nearly half of this area is marshland. Furthermore, large was settled, 1 100 years ago, about 200 eruptions have oc­ areas of well-drained soils carry only a sparse cover of curred from 40-50 craters, an average of I every 5 years. vegetation. In many places the vegetation is intersected by During the last 10 000 years lava flowshave covered some deep gullies, fo rmed by wind and water erosion, whilst in 10 000 km2, most of which are barren or sparsely other areas small patches of vegetation stand out in a vegetated. barren wilderness like islands. Only 1 %of the land area is The climate is cool temperate, oceanic and variable. The covered by fo rest, composed of birch, Betula pubescens, most peculiar fe atures of the weather conditions are trees or shrubs. sudden, alternate invasions of polar air from the north and warm or transitional air masses from the Atlantic. Many of the vigorous North Atlantic cyclones originate or Historical aspects regenerate in the Newfoundland region, and moving north­ eastward, reach their maximum intensity in the vicinity of There has been much discussion on what the country Iceland. A depression that becomes stationary or slow­ looked like at the time of settlement in AD 874, how moving off the southwestern coast of Iceland may main­ large an area carried vegetation and what type of tain a warm or semi-warm flow of Atlantic air over the vegetation existed. country for a considerable period. This causes thaws in Historical and scientific evidence has been put winter, but rainy and rather cool weather in summer in the fo rward to show that over half of the land carried south part. In other cases the depressions may cross the vegetation and most of this was forest, dominated by country and slow down or almost stop over the sea birch trees or shrubs. Unlike most European coun­ between its eastern coast and Norway. This situation, fre­ quently combined with high pressure over Greenland, tries, Iceland has written records, from the 12th cen­ causes a persistent flow of polar air over Iceland and a tury, which describe the settlement of the country. spell of cold weather, especially in the north. The central The records give no detailed descriptions of the highlands, with mountains and glaciers, fo rm an effective natural vegetation but in many places they mention barrier against winds and weather between the various dis­ birch fo rest where now there are only eroded hills. tricts of Iceland. Therefore the northern and southern In the oldest of the Icelandic chronicles, the fslen­ coasts, or the eastern and the western ones, will seldom dingabok-Landmimabok, written by Ari the hawe quite the same kind of weather simultaneously. Learned between 1 122 to 1133, it is written: "At that The flora includes some 450 species of higher plants, time (i.e. of the settlement) Iceland was covered by which is poor compared with those of neighbouring coun­ woods from sea shores to the mountain-sides." Many tries. The species-richest families are: Cyperaceae (5 1 spp.), Gramineae (4 7 spp.), Caryophyllaceae (27 spp.), other manuscripts describe similar scenes. In Co mpositae excl. Hieracium and Ta raxacum (20 spp.), Kj alnesingasaga (written ea. AD 1300) it is written : Juncaceae and Rosaceae (18 spp.), Cruciferae and "Then (ea. AD 900) all Kj alarnes was covered by Scrophulariaceae ( 17 spp.), Saxifragaceae (16 spp.) and fo rest and the only openings were those cleared by Ponygonaceae, Potamogetonaceae and Ranunculaceae man for farmsteads and roads." To-day this area is

A eta P_hy togeogr. Suec. 68 A 32 gust H. Bjarnason

Fig. I. A typical gully caused by wind erosion in Fnjoskadalur. The white stripe is one of the layers of volcanic ash found in usual ''loess·· soil profiles. The layer is a prehistoric one originating fr om the mountain Hekla or it vicinity.

- 1974. Photo P. Jonsson.

completely deforested and the cultivated fields are proof of more extensive vegetation cover can be ob­ surrounded by eroded hillsides, infertile, sparsely­ tained by measuring the rate of soil accumulation vegetated gravels and marshes. Many other examples between volcanic ash of known age. Tephro­ can be given and whereas there is much disagreement chronological investigations have shown that prior to as to the historical accuracy of the events described settlement soil accumulation was very slow (0. 1 mm in the Sagas, it is generally agreed that both place and per year). After AD 900 the rate increased con­ geographical descriptions are accurate. Therefore it is siderably, reaching 0.5 mm per year during the last unlikely that the recording of forests will be disputed. two centuries. Before settlement a fr agile equilibrium Later chronicles, especially farm registers from the prevailed between regrowth and deposition, but later 18th century, record extensive devastation of vegeta­ on deposition increased considerably due to in­ tion. The most noteworthy of these references is the creased erosion. Farm Register (Jaroabok) compiled by Arni The causes of erosion have long been disputed. Magnusson and Pall Vidalin, after a country-wide Earlier, many blamed climatic deterioration, vol­ survey between 1 702 and 1712. Magnusson and canic eruption, avalanches and glacial bursts. Recent Vidalin visited all parishes to record revenues, research has not been able to correlate erosion with livestock, leaseholds, charges, etc. Their record is one climatic deterioration or any other natural phe­ of the most reliable sources available as all informa­ nomena, such as volcanic activity, or indeed that tion given to them by the farmers was checked by in­ these disturbed the existing equilibrium. Local excep­ dependent witnesses. The register gives an excellent tions can be fo und but it is worth mentioning that the description of the condition of the vegetation, with fe w existing woodland remnants in South Iceland are special emphasis on woodlands, stating their condi­ in fact near the most active volcanic areas. tion, utilisation and whether erosion was present. The On the other hand, most evidence shows that ero­ present article owes a great deal to this register and it sion followed destruction of the forests and therefore will be referred to later. can be attributed to human interference (Fig. 1 ). In addition to written sources on soil and vegeta­ So far only a few factors bearing on the vegetation tion destruction, many place names indicate that of earlier times have been discussed. Whilst the fo rests occurred in areas which are now devasted, sources are of varied origins they all indicate that the such as Fitjaskogar, Holaskogar, Dynskogar vegetation cover was different than it is to-day. Little ("-skogar" means woodland). Furthermore, there are research has been conducted on the history of the still remnants of vegetation on otherwise barren and Icelandic vegetation cover and much more is needed. denuded sites, which indicate that neighbouring areas The remainder of the article will be devoted to a cer­ were once vegetated. In many places the remains of tain valley in North Iceland, Fnjoskadalur, as its old charcoal pits can be found on open land, which vegetational history is well documented compared shows that they were once forested. One irrefutable with other regions. Reliable information on the

Acla Ph; logeogr. Suec. 68 The history of woodland in Fnjoskadalur 33

woodlands in the valley is available from the time of the Farm Register of 1712. Changes in vegetation which have occurred in the valley since then are by no means exceptional to that area, but typical for many districts in Iceland.

Description of the valley of Fnj6skadalur The district, known as Halshreppur, lies in the valley of Fnjoskadalur and is more often called after the valley. The valley is approx. 10 km east of Akureyri. It is enclosed by high, basaltic mountain ridges which lie in a S-N direction. Fnjoskadalur valley (hereafter called Fnjoskadalur) is about 40 km long. In the north it changes direction to E-W and opens onto the Eyjafjorour coast. This part of the valley belongs to another district and is not included in this review. A wide mountain pass, Ljosavatnsskaro, opens from the east, midway in the valley. The farms at the mouth of the pass are included in the district political­ ly but are not g�ographically a part of the valley. At its southern end the valley splits into three tributaries. While there are reports that there were a fe w farmsteads in these valleys in fo rmer days, inhabita­ tion has rarely gone beyond the main valley. A fresh­ water river, Fnjoska, flowsthrough the valley, dropp­ ing evenly in elevation to the sea from its source 121 km in the highlands. Fnjoska is fast flowing, with an average discharge of 45 m3 /sec; however, stream flow varies considerably, according to weather conditions, from 10 to 200 m3/sec. The valley floor slopes evenly northwards with a 200 m elevation drop from south to north. It is narrow, being between 1 and 2 km wide at the most. Along its western edge the mountains are 500-700 m in height with an even slope, while to the east the mountains are 700-900 m high, steep and gullied. Apart from around the farms Draflastaoir(no. 2) and Sorlastaoir (no. 18) there is little lowland (see Fig. 2). Due to the E-W direction of its mouth, moisture­ bearing north winds do not penetrate the valley directly and precipitation is low. Meteorological observations have only been recorded over the last 20 years but the Meteorological Office has estimated the climatic data fo r the period 1931-1960. Mean an­ nual precipitation is 658 mm, 70 % of which falls in the winter when snowcover is usually deep. Mean an­ nual temperature is 2.6°C, the valley bottom being at a relatively high elevation. The frost-free period is short and mean summer temperature lower than ex­ pected. Mean monthly precipitation and tempera­ tures are shown in Fig. 3.

Fig. 2.

A eta Phytogeogr. Suec. 68 A 34 gust H. Bja rnason

(b) Some were abandoned for a while but have been

Va glir reoccupied. 1931-1960 (c) The tenancy sub-holdings have fallen into disuse, 8 1------j except two which are now independent farms. (d) Several new farms have been formed out of older u 0 6 1------j ones in the middle part of the valley. dl 1lli' Despite these changes the number of farms has :s 4 t------i changed little and in 1976 there were 40 inhabited �dl a. farms. E � 2 1------i

J F M A Deterioration of the birch woodland The history of the birch forest in Fnjoskadalur from 1 712-197 3 will now be described. As mentioned earlier, the description is based on the Farm Register of 1 712. Earlier records from the Sagas and other -80 1------�� chronicles are ignored although they give similar in­ ,,,,, formation. Furthermore, journals by visitors, both . .. .. f'=.'.·'.· ::::: . , ·... ,, fo reign and Icelandic, are used, but here one must E 60 mr: ---=mr..-tl mll.,.l,,. : l ml iH!ir---""' E - rl remember that most of these travellers kept to the --- main road, which crossed the valley to the !l.l.·,·�!··,l:'·'·,il.,·',�,-,i-!!!! Lj osavatnsskaro Pass, and only describe the middle ',,,,,·',, part of the valley. Other sources of information are Tim ,,,,, the oral communications of people who lived in the 1m1 .... �:11111 1111111111! 111� valley during the 19th century. At the end of the 19th ::o� ���������==����m 111�� century three Icelandic scholars investigated the t J FMAMJ J ASOND woodlands and their descriptions are valuable Fig. 3. Monthly mean temperature and monthly mean sources of information. Finally, the distribution of precipitation at Yaglir, 1931-1960. birch woodland in 1973 is described. The author mapped the birch woodland and made, in addition, Generally speaking, soils are well-drained : silt several vegetation surveys in the valley. loams, 15- 100 cm thick with low moisture-holding capacity and low fertility over a gravelly and stoney subsoil; or sloping gravels, stoney alluviums and 1700-1799 (a) Farm Register of Magnusson and Vida ­ steep, stoney tali. There are no lakes, only a few tarns 1712, In September 1712 all the farm tenants in and small patches of marshland. Natural hot water is lzn. fo und at three places. Fnjoskadalur were called to the parsonage at Hals A detailed description of the settlement of the (no. 24) to give information on their holdings to the valley is not attempted here. The valley has, however, above authors. Furthermore, they were made to cor­ been inhabited since the time of settlement, when roborate the information given by their neighbours. Thorir Snepill settled at Lundur (no. 22). The farms The Farm Register is orderly and all descriptions are stand, more or less equidistantly, along both sides of standardised, which makes it easy to compare the the river. The average distance between farms is 2 farms. Woodlands are classified into three utilisation km, so the valley can be considered as sparsely types: settled. Most farms stand well above the river on an (i) Woodland yielding "logs" (i.e. small timber) ice-age gravel terrace. (ii) Woodland yielding charcoal This article and the accompanying map (Fig. 2) (iii) Woodland yielding firewood are based on the Farm Register of 1712. Then there Each type is further classifiedaccording to its yield were 38 farmsteads plus 6 tenancy sub-holdings. conditions: excellent - good - plentiful - sufficient - There were 49 tenants as five of the farms were joint tolerable - deficient - little (amount) - cleared. tenures. Since then there have been various changes: Furthermore, other uses of the woodlands (than (a) Several farms have been abandoned, most of timber) such as for haymaking and other agricultural them between 1920-1 960, especially in the south of purposes are mentioned. Sometimes the general con­ the valley. dition of the woodland is recorded, especially if the

A eta Phytogeogr. Suec. 68 The history of woodland in Fnjoskadalur 35 trees are suffering from wood-rot or are dilapidated. woodland was extensive and abundant. In the following (Table I), the information from the One should also bear in mind that the farmers Farm Register is classified according to the descrip­ tended to minimise the benefits and exaggerate the tions of the Farm Register (the tenancy sub-holdings disadvantages as they suspected the register would be are included with the main farms). used fo r taxation. The table shows that there is woodland on 31 of the 38 farms, 14 farms of which have woodland (b) Eggert 6iafsson, naturalist, and Bja rni 1752, yielding "logs" that varies from "deficient" to "ex­ Povelsen, physician. They stated that the woodlands cellent", and three others where the woodland has in Fnjoskadalur were superior to all others in Ice­ been cleared or almost so. Charcoal woodland occurs land, but they had greatly deteriorated since the last on 25 farms, again varying from "little" to "ex­ century. They said that 100 years earlier trees which cellent", but has been destroyed on 4 farms. were 12,5 m up to the first branches had grown Woodland yielding firewood was fo und on 31 farms, there. But now the woodlands were abused and in again varying from "little" to "excellent" but "plen­ need of protection. tiful" on most farms. On seven farms no woodlands occur. However, it (c) 6lqfu r Olavius, naturalist. Before he began 1777, is said that at Birningsstaoir (no. 25) an old hayfield his journey he had been told that the trees were so tall is so infested with birch shrubs as to be useless. Ac­ that a man on horseback, holding up his whip, would cording to the Register the largest deforested areas be hidden underneath the trees. He was greatly disap­ are: One in the north at Austari- and Vestari-Krokar pointed because at the farmstead of Hals "there are (nos. 36 & 37) and another in the south at Steinkirkja no trees, large or small, young or old, only stumps, (no. 11) and Fjosatunga (no. 12). The main thorough which show that a forest had once grown there, and a fares from the neighbouring districts, which were by small area of shrub on the opposite side of the river". then deforested, passed through both these areas. It is He fo und it incredible that such devastation should almost certain that these woodlands were cleared by happen over only 20 years, but unfortunately it was men from other parishes, who deliberately sought to true. use the woodlands in Fnjoskadalur. There was also no woodland at Vegeirsstaoir (no. 32). According to (d) Sveinn Pdlsson, physician. He wrote in his 1794, the Farm Register this farm was very small and the diary: "Fnjoskadalur is a beautiful district and had occupant had to lease both winter and summer graz­ fo rmerly been forested. The wood at Hals was ing and pay fo r it by indefinite fees and services. It is renowned over all Iceland but during the last 50 years not unlikely that he had to encroach on his has been completely destroyed ...apart from a small woodlands to pay fo r grazing. copse 1/4 mile south of the farmstead." From the Farm Register it is obvious that Fnjoskadalur was heavily wooded at this time. Apart (e) Oral communications. In addition to the above from the above-mentioned clearings, most of the references, traditions concerning the tall woodland in valley was covered by birch woodland, probably Fnjoskadalur have been passed down by the in­ some 10 000 ha. On 8 farms the occupants com­ habitants of the valley who lived there in the middle plained of the difficulties of haymaking as nearly all of the century. It would be too much to mention them their land was wooded. On the other hand, it is clear all, however tempting. A woman, who grew up at that in many places the woodlands were dete­ Hroastaoir (no. 9) and lived all her life in the valley riorating. It was not without reason that the absentee (1761-1848), said that in her youth there had been landlord of Vioivellir (no. 5), who was sheriff of a so much woodland at Hroastaoir that it was different county, fo rbade his tenant to sell charcoal or necessary to tie bells on the cattle so they could be timber without his permission. As can be seen from fo und. Now this is a barren hill-side, bearing no Table I, it is "log" woodland that was most en­ resemblance to its past glory, but the many, large, dangered, being more heavily utilised. On many charcoal pits show that woodlands once grew there farms "log" woodland was recorded as cleared or in (Fig. 4 ). Furthermore, there is a tradition that tall very small areas. At both Lundur (no. 22) and Vaglir woodland grew on the slopes above Snrebjarnarstaoir (no . 23) the "log" woodlands are classified as ex­ (no. 16) in the middle of the 18th century. One severe ceUent, but these are exceptions. At Hals (no. 24) the winter there was no hay for the sheep, so the occu­ "log" woodlands are classifiedas good and extensive, pant chopped all the branches off the trees which wh -le at Vioivellir (no. 5), Thoroarstaoir (no. 21) and stood above the snow as fodder. The woodlands were Fornastaoir (no. 28) they are classified as plentiful. completely destroyed and now the area is completely By plentiful we can be sure that it was meant that the deforested.

A eta Phytogeogr. Suec. 68 36 Agust H. Bja rnason

TABLE I. Classification of woodlands for different utilisation according to the Farm Register , 1712 .

Used for TYPE OF WOODLAND haymaking and agricult . No . Farm "Logs" Charcoal Firewood purposes Condition

l. Melar plentiful plentiful X

2. Draflasta6ir tolerable tolerable almost fallen G.�:imsger6i due to decay

3. D�li sufficient sufficient X decayed

4. Vatnsleysa deficient plentiful plentiful X

5. Vi6i'lellir plentiful ple ntiful plentiful X

6. Lj6tssta6ir sufficient plentiful plentiful X

7. Nes cleared little decayed

8. Sk6gar sufficient plentiful plentiful X

9. Hr6asta6ir sufficient plentiful plentiful X

10 . Veturli6asta6ir sufficient sufficient X

11. Steinkirkja none

12 . Fj6satunga Brunager6i none Grj6targer6i

13 . Illugasta6ir Kotungssta6ir

14 . Reykir almost plentiful plentiful X cleared

15. Tunga plentiful plentiful plentiful X

16 . Sn�bjarnarsta6 ir cleared sufficient plentiful

17 . Hjaltadalur cleared sufficient

18 . Sorlasta6ir cleared sufficient

19 . Bakki almost sufficient sufficient Bakkasel cleared

20 . Belgsa sufficient plentiful plentiful X

21. Th6r6arsta6ir plentiful plentiful plentiful X

22 . Lundur excellent excellent excellent X

23. Vaglir excellent excellent excellent X

24. Hals good and good good X decayed extensive

25. Birningssta6ir no wood land mentioned

26 . Kambssta6ir no woodland mentioned

27. Sigri6arsta6ir plentiful plentiful X very decayed

28. Fornasta6ir plentifll'l plentiful plentifu l X Fornasta6asel

29 . Hallgilssta6ir cleared plentiful plentiful X

30. Veisusel tolerable almost fallen

31. Veisa tolerable almost fallen

32 . Vegeirsta6ir no woodland mentioned

33. Bi:i6varsnes cleared sufficient

34 . Ytrih6ll sufficient sufficient X

35. Gar6ur little plentiful

36 . Austari-Kr6kar no woodland mentioned

37 . Vestari-Kr6kar no woodland mentioned

38 . Thvera sufficient plentiful plentiful X

Acta Phytogeogr. Suec. 68 The history of woodland in Fnjoskadalur 3 7

Fig. 4. The farm Hroastaoir on the west bank of river Fnjoska. The slopes above and south of the farm were covered with birchwood late in the 18th century. The small patch of young birches below the hayfield has grown from self-sown seedling within a protected area, 1974. Photo P. Jonsson.

Fig. 5. The flat bottom of Fnjoskadalur west of Hals. Note the eroded mountain slopes above the fa rm and the small partches of birch below the terrace which have survived because of thick snow cover, 1974. Photo P. Jonsson.

A eta Phytogeogr. Suec. 68 A 38 gust H. Bja rnason

Fig. 6. The terrace SW of Hals 22 years after it was protected from grazing. On the few patches left of original soil,

birches have sprouted vigorously from old roots. The eroded terrace is almost as barren as it was previous to enclosure. 1968. Photo H. Bjarnason.

Travellers in the 18th century often spoke of the 1800-1899 wood at Hills-naturally enough, fo r it was famous. (a) Ebenezer Henderson, agent of the British 1814, The farm had for centuries been a parsonage and a and Overseas Bible Society. In his travelogue he benefice and stands on the main road. It was com­ writes: "About a hundred years ago, the valley (i.e. mon talk that the parson, Jon Thorgrimsson who was Fnjoskadalur) exhibited one of the fm est fo rests in the beneficiary in 1736- 1 795, had more or less Iceland, but now there is not a single tree to be ruined the wood. He had been accused early in his seen-such has been the havock made by the in­ career fo r his greedy and destructive woodcutting. clemency of the seasons, and the improvident con­ The parson took the criticism badly and tried to clear duct of the inhabitants. The remains of this fo rest are himself of the accusations that the woods were being still visible on the east side of the river, which divides destroyed. In view of this he tried to convince people the valley, in the numerous stumps of birch trees that the wood was destroying itself and he had not which present themselves, some of which exceed two sold cutting rights to outsiders. He declared that he feet in diameter." only fe lled trees to improve the wood, sanitary clearings, otherwise it would deteriorate and be of no (b) Thorsteinn Sigurosson, then 1850-1856, use. Despite the parson's attempt to justify his treat­ labourer at Hcils. He said that at that time there was ment, the wood deteriorated annually. The story cir­ no erosion in the area to the north and west of Hals, culated that he once hired four farmers to fell 1600 the vegetation was mainly heath with low shrub but "logs" for his own disposition and 400 "logs" as their no woodland. This area is now eroded to the subsoil wages (Fig. 5). and barren.

A eta Phytogeogr. Suec. 68 The history of woodland in Fnjoskadalur 39

(c) Sremundur Eyjolfsson, agriculturalist and 1892, theologian. He says that tall woodland only occurs at three places, namely Hals (no. 24), Vaglir (no. 23) and Thoroarstaoir (no. 21), but virtually no woodland remnants occur on any other farms. He adds that the wood at Hals had almost been destroyed in the 19th century but had now regenerated itself, mainly on the slope south of the farm above the river, although it was far less exten­ sive than before (Fig. 6). Open glades and strips ex­ tended in many places into the wood and the soil was rapidly eroding on these sites. At Vaglir tall wood still occurred at places but lately it had been felled so drastically that he had never seen such treatment. He considered the wood at Thoroarstaoir to be the most beautiful in the country. For this he thanked the farmer who had lived there for many years and ap­ preciated it so much and that he had taken care to Fig. 7. Distribution of woodland in Fnjoskadalur, dark protect it. area; (a) 1712 and (b) 1973.

Thorvaldur Thoroddsen, geologist. On his (d) 1896, travels in the valley he lodged at Thoroarstaoir (no. 1900-1973 Shortly after the turn of the century a real change in 21). The farmer told him that in 1802 there had only the history of the birch woodlands occurred. In 1907 been enough timber in the forest for axe-shafts but a Forestry Law was enacted with the immediate now it was both tall and extensive. The tallest tree policy of conserving the native woodlands together which Thoroddsen measured was 8 m in height and with establishing new fo rests of exotic species. The 55 cm in circumference at ground level. He also men­ Fnjoskadalur woodlands were quickly chosen and tions the woods at Hals and Vaglir. The latter had the Icelandic Forestry Service has vigorously deteriorated considerably due to poor management. followed the policy of enclosing the woods as far as Thoroddsen remarked that there was no woodland finances allow. The fo llowing woodlands have been on the west side of the Fnjoska river although shrubs enclosed (totally 2 270 ha): grew at a few places at the beginning of the century. Otherwise he has little to say on the distribution of (1) Vaglir forest fenced in 1909. woodlands in the valley except that shrubland oc­ (2) Sigrioarstaoir fo rest enclosed on two sides in ClUrred in the valley south of Sorlastaoir (no. 18) on 1931. the east side of the river. (3) The woodlands north of Melar (no. 1) in 1942. (4) The woodland remnants at Hals, which border on to Vaglir, in 1946. Sigurour Sigurosson, agriculturalist. He (5) Part of the woods at Lundur (no. 22) and (e) 1899, visited the valley to survey the distribution of Thoroarstaoir (no. 21) and part of Belgsa (no. 20) woodland. Woodland occurred on only 5 farms, in 1946. namely, Belgsa (no. 20), Thoroarstaoir (no. 21), Lun­ As mentioned earlier, the author mapped the dis­ dlur (no. 22), Vaglir (no. 23) and Hals (no. 24), with a tribution of birch in Fnjoskadalur during the summer t tal area of less than 500 ha. In addition, shrubland of 1973. The results are shown on the map included. occurred south of Sorlastaoir (no. 18) and at Thvera Briefly it can be said that woodland covers some (mo. 38). It had recently grown at Melar (no. 1) but 2 060 ha, of which 1 260 ha have been enclosed. h.ad been cleared. At various places in the woodlands Judging by the accounts of the Farm Register of large openings broke up the canopy, especially at 1712 an area of 10 000 ha of woodlands at that time Vaglir and Hals. Sigurosson remarks that often the would not be an overestimate (Fig. 7). t1rees are leaning and crooked due to heavy snowfalls It is also interesting to compare this result with a.nd grazing. Sigurosson had acquainted himself with Sigurosson's survey of 1899. Within all the en­ f

A eta Phytogeogr. Suec. 68 40 Agust H. Bja rnason

Fig. 9. Brushwood from old roots after decreasing grazing at the southern end of the valley, 1975. Photo H. Hallgrimsson.

ing to these investigations the volume growth of young birch stands is similar to that obtained in Norway, but as the stands mature the yield of the Icelandic birch lags behind the Norwegian birch. This is because the Icelandic birch makes very little height increment after reaching middle age. Information on total volume production per hectare is lacking but for normally thinned woodland, with 1 940 trees/ha at 45 years of age the average standing volume is 30.2 1 m\ mean annual increment (m.a.i.) is 0.79 m3 and current annual increment 0.91 m3• If these figures are compared with Norwegian yield tables for birch Fig. 8. Birchwood grown up from roots and stumps after growing in similar conditions, then one can expect the 20 years of protection in the southern part of the woodland total production and m.a.i. of the Norwegian stands at Vaglir, 1930. l?hoto H. Bj arnason. to be about 25 % greater. It may also be added that birch growth in Fnjoskadalur is usually slightly less land. The woods at Vaglir and Hals have shown the than in other regions, due especially to the low greatest improvement, but at Melar there is now birch precipitation. woodland 4-6 m in height where Sigurosson Apart from mapping birch woodland, the author reported deforested land (Fig. 8). In the southern part investigated the major plant communities in the of the valley birch shrubland, 3-4 m in height, has valley, especially with the object of comparing the developed on unenclosed land. The major reasons for sward under birch woodland with that of grazing this regeneration have been the gradual abandon­ areas. The investigation has not yet been published ment of farms in this area (nos. 15, 16, 17, 18, 19 and and it is not appropriate to go into any detail. It is 20) and, since 1900, discontinuation of winter graz­ noteworthy that the same species occur on both sites, ing (Fig. 9). although there are significant differences in their Most of the woodland in Fnjoskadalur has grown respective frequencies and abundance within and out­ up from grazed trees on which signs of browsing can side the enclosures. The most common species on still be seen. It seems that the trees take a long time to both sites are Arctostaphylos uva-ursi, Calluna recover fr om the many years of grazing before they vulgaris, Empetrum nigrum, Festuca spp., Poa spp. make any real growth. On the other hand, the trees at and Vaccinium uliginosum. Vaglir, where they have been protected for nearly 70 On the other hand, there are several species which years, have developed normally and the tallest are characteristic of grazed land but are rare in specimens are 13 m in height and 22 cm in diameter enclosed woodland (see Plant list I). In the same way at chest height. Detailed yield measurements were a smaller number of species grow in the woodland, made on the birch stands at Vaglir in 1956. Accord- but scarcely thrive outside of woodland in the valley

eta A Phytogeogr. Suec. 68 The history of woodland in Fnj6skadalur 41

(see Plant list 11). The latter species are not especially Table II. Changes in the livestock in Fnj6skada1ur , 1712-1978. confmed to woodlands but often occur in sheltered Year Cattle Sheep Goats Horses Hay production sites in lava fields in other districts. 3 1712 126 3191 421 134 ± 3 100 m

1942 162 3832 111 116 ± 9 333

Plant list I: Plant list II: 1978 262 8715 101 ±30 000 Agrostis canina ssp. mont. Agrostis tenuis A lchemil/a alpina Coeloglossum viride A renaria norvegica Deschampsia flexuosa of the 17th century the ratio between cattle and sheep Armeria maritima Geranium sylvaticum was 1 :6. It then increased suddenly to about 1:25, Cardaminopsis petraea H ierochloe· odorata where it has since remained. If one examines hay Carex rupestris Ramischia secunda production fo r 1712 it is obvious that the livestock Cerastium alpinum Ranunculus acris were forced to fe nd fo r themselves out-of-doors as Draba incana Roegneria canina long as weather and fo rage permitted and this custom Dryas octopetala Rubus saxatilis was maintained for centuries. Since 1900 hay Juncus trifidus production has gradually increased. Apart from Kobresia myosuroides sheep, goats were common in Fnjoskadalur until Loiseleuria procumbens Minuartia biflora 1960 and they are even more destructive to fo rest Rhinanthus minor than sheep (see Table 11). Furthermore, in 1771 Sa/L'( herbacaea reindeer were introduced to Iceland and ranged in Silene acaufis Fnjoskadalur and neighbouring valleys to the south Tofleldia pusilla until 1857. Viscaria alpina From this review it is clear that birch will rapidly All in all, more species occur on the grazing land spread once grazing pressure is reduced, even to but the vegetation is much less vigorous and the up­ eroded land. The history of the woodlands in per soil layer suffers more from wind drying. The Fnjoskadalur shows that nature is seeking a dyna­ sward under birch differs considerably, it is far more mic balance with the succession of plant communi­ vigorous and has a highly developed moss layer with ties, resulting in birch woodland. Tomentypnum nitens, Climacium dendroides, Rhyti­ The words of Ari the Learned that the country was diadelphus triquetrus, Drepanocladus uncinatus and wooded between the mountains and the shore at the Hy locomium splendens as characteristic species. time of settlement are likely to have been true.

SUMMARY REFERENCES

The history of the woodlands in Fnjoskadalur has Bj arnason, H. 1942, Abuo og ortroo. - Skogrrektarfelag been sketched. From this one can conclude that most islands. Arsrit 1942:8-40. of the valley, about 10 000 ha, was fo rested around 194 7. Eyeing skoga i Fnjoskadal. - Skogrrektarfelag 1700. Due to imprudent management the woodlands islands. Arsrit 1947:5-19. have been destroyed on most farms, which has led to 1966. Skn1 yfir skogrrektargiroingarog grooursetningu. - Skogrrektarfelag islands. Arsrit 1966:30-38. soil erosion, as witnessed by the bare hill-sides. Short­ Brejatal a islandi 1976 (The Icelandic Farm Register of ly after 1900 work was begun on enclosing 1976). - Reykjavik. 160 pp. woodlands. This, coupled with changed agricultural Eyjolfsson, S. 1900. Fero urn pingeyjarsyslu og Fljots­ practises, has resulted in regeneration of the wood­ dalsherao. - Bunaoarrit 8:50-6 1. lands at several places over the last 50 years. Henderson, E. 1818: Iceland or the journal ofa Residence The causes of devastation have barely been dis­ in that island during the years 1814 and 1815. Vol. cussed, but they are first and foremost agricultural I-ll. - Edinburgh. 377 pp. & 412 pp. practises. During the 18th and 19th centuries many Hylander, N. 1955. List of the Plants of N.W. Europe. I. bad seasons occurred and as a result the people were Vascular plants. - Lund. 175 pp. fo rced to abuse their land and grazing led to the islendingabok - Landnamabok (The Book of the Icelanders - The Book of the Settlement) 1968. Ed. by deterioration of conditions fo r woodlands. Felling J. Benediktsson. - islenzk Fornrit 1-2. Reykjavik. was usually uncontrolled, which by itself would not 525 pp. have caused deforestation if it had not been fo llowed Johannesson, B. 1960. The Soils of Iceland. - Univ. Res. by grazing, especially year-round grazing with sheep Inst., Dept. Agric. Rep. Ser. B., 13:1-140. and goats. Kjalnesingasaga 1959. Ed. by J. Halldorsson. - XIV. From the first years of settlement until the middle fslenzk Fornrit. Reykjavik. pp. 1-44.

A eta Phytogeogr. Suec. 68 42 H. Agust Bja rnason Magnusson, A. & Vidalin, P. 1943. Jaroabok (Farm Steindorsson, S. 1964. Groour a islandi. - Reykjavik. 186

Register) 11. - Hio islenzka frreoafelag. Kaup­ pp. mannahofn . 409 pp. Teitsson, B. 1973. Ein garhald og abuo a joroum i Suour- Olafsen, E. & Povelsen, B. 1 772. Reise igenem Island. - t>ingeyjarsyslu 1703-1 930. Studia Historica Soroe. 1042 pp. 2:1-183. Olavius, 1780. Oeconomisk Reise igiennem de 0. Thorarinsson, S. 1961. Uppblastur a islandi i ljosi nordvestlige, nordlige og nordostlige Kanter af Island. oskulagarannsokna (Wind Erosion. A Tephro­ - Kobenhavn. 756 pp. chronological Study). - Skogrrektarfelag islands. Palsson, S. 1945. Feroabok. Dagbrekur og ritgeroir Arsrit 1960-1961:17-54. 1791-1797. - Snrelandsutgitfan. Reykjavik. 813 pp. Thoroddsen, Th. 1911. Lysing islands Il. - Kaup­ Sigurosson, Sigurour 1900. Skogarnir i Fnjoskadal. - mannahofn. 673 pp. Andvari 25:144-175. - 1960. Feroabok IV. Sec. ed. - Reykjavik. 392 pp. Sigurosson, Snorri 1977. Birki a fslandi. - Skogarmal. Veorattan. 1962. Arsyfirlit. - Veourstofa islands. Reyk­ Reykjavik. pp. 146- 172. javik. 127 pp. Skaptason, J. 1969. S-t>ingeyjarsysla vestan Skjalfan­ dafljots og Fljotsheioar. - Feroafelag islands. Arbok 1969:1-182.

Agust H. Bj arnason, Laugateigur 39, 105 Reykjavik, Iceland

A cl a Ph.l'logeogr. Suec. 68 Strandverschiebung und Strandvegetation auf Havero Prastang Wolfgang Cramer

Im Rahmen einer geookologischen Examensarbeit sollten dert fe stgestellt wurde, daf3 die Strandverschiebung in im Sommer 1979 Vegetationsuntersuchungen an der der nordlichen Ostsee starker war als in weiter siid­ Schiirenkuste von Uppland (Mittelschweden) durchgeflihrt lich gelegenen Bereichen und infolgedessen Meeres­ werden. Angeregt durch zahlreiche finnische Arbeiten (u.a. spiegelschwankungen nicht die Hauptursache sein Valovirta 1937; Palomiiki I963; Hiiyren I948), fur die es konnten. Der Begriff ,Landhebung" ging daraufhin auf schwedischer Seite noch kaum Entsprechungen gibt, in den allgemeinen Sprachgebrauch iiber. Er dient sollte der Zusammenhang zwischen der durch die Landhe­ bung verursachten Strandverschiebung und der Zonierung heute als erschopfende Beschreibung der im Zusam­ der Strandvegetation behandelt werden. Bei Berucksichti­ menhang mit der Strandverschiebung ablaufenden gung der zeitlich und materiell begrenzten Arbeitsmoglich­ Vorgange. keiten bedeutete dies, da/3 einer Literaturubersicht zum In pflanzenokologischen Fragestellungen ist die Thema (vgl. Cramer 1979) beispielhafte Vegetationsauf­ Verwendung des Begriffs ,Landhebung" und der in nahmen von verschiedenen Strandtypen gegenubergestellt der geologischen Literatur dafiirangegeben en Werte werden sollten. nicht sinnvoll. Obwohl er die Hauptursache fiir die Die Literaturarbeit wurde im Winter I 978/79 am Pflan­ konstante Veranderung der Standortsbedingungen zenokologischen lnstitut (V iixtbiologiska Institutionen) fur die Strandvegetation beschreibt, wird er doch von der Universitiit Uppsala durchgefuhrt. Das Projekt wurde von der Themenstellung an uberaus geduldig betreut von einer Reihe anderer Faktoren iiberlagert, die alle den Hakan Hytteborn. Unterstutzung bekam ich aber auch Verlauf der Strandlinie und die Qualitat des Substrats von vielen anderen Mitarbeitern des Instituts, insbesondere verandern und es daher verniinftig erscheinen lassen, von Hugo Sjors, der mir durch die unburokratische Auf­ zusammenfassend den BegrifT,S trandverschiebung" nahme als Gaststudent den gesamten technischen Apparat anstelle von ,Landhebung" beizubehalten (wobei des Instituts zur Verfi.igung stellte. fe stgestellt werden muf3, daf3 die in dieser Hinsicht Als geeignetes Untersuchungsgebiet fand sich die Halb­ korrigierten Zahlenangaben in der Regel die vertikale insel Hiivero Priistiing im nordlichen Roslagen, etwa 8 km Komponente der Strandverschiebung ausdriicken). von der Hafenstadt Hallstavik entfernt gelegen. Am Strand Sole he Faktoren sind unter and ere m: dieses seit I 960 wegen seiner reichhaltigen ,Laubwiesen"­ V egetation als N aturreservat ausgewiesenen Gebietes Kurz- oder langerfristige Meeresspiegelschwankung­ kommen auf engem Raum die wichtigsten Strandtypen des en, die fiir geowissenschaftliche Fragestellungen ge­ inneren uppliindischen Schiirengurtels vor. Im Archiv der wohnlich statistisch eliminiert werden - fiirdie Vege­ Stockholmer Provinzregierung fanden sich Akten und tation dagegen sind sie bedeutsam. Kartenmaterial, die gerade fur dieses Gebiet gute Anhalts­ punkte fur Aussagen uber die seit dem 18. J ahrhundert ab­ Erosion von Bodenmaterial an exponierten Strand­ gelaufene Strandverschiebung gaben. abschnitten durch Brandung oder Eisschub. Sedimentation von Material an geschiitzteren Strand­ partien. St:randverschiebung auf Havero Prastang Strandverschiebung ist also eine Wirkung samtlicher Beg riff ,Stra ndverschiebung" sich an einem Standort auswirkender morphodyna­ Als Folge der letzten Vergletscherung, moglicherwei­ mischer Faktoren und stellt daher einen gut begrenz­ se auch aufgrund tektonischer Prozesse (Morner baren Ausschnitt aus dem Wirkungsgefiige der fiir 1977), ist die Landmasse ganz Nordeuropas in einer die Strandvegetation bedeutsamen Umweltbeding­ starken Hebung begriffen. Bereits seit dem 17. Jahr­ ungen dar. hundert war die dadurch verursachte Verschiebung der Strandlinie in Richtung Meer beobachtet und Herkunft und Zusammensetzung des Bodenmate­ verschiedentlich sogar gemessen worden. Es entstand rials bald eine langwierige Diskussion, ob es sich urn eine e statische Meeresspiegelsenkung oder eine glazial­ Nach dem Riickzug des Inlandeises, der fiirdas nord­ is statische Landhebung handele, bis im 19. Jahrhun- liche Roslagen etwa auf 7500 v.Chr. datiert wird, lag

A eta Phytogeogr. Suec. 68 44 Wolfgang Cramer die heutige Kiistenlandschaft zunachst unter dem Meeresspiegel. Die nahezu plangeschliffenen Gra­ nitgneise waren praktisch iiberall von einer mehr oder weniger machtigen Schicht lockerer Mergelablage­ rungen bedeckt. Da in der spateren Phase der Verei­ sung die Eisstrome vorwiegend aus der Richtung der nordostlich (in der Gavle-Bucht) gelegenen Kalkge­ biete kamen, erreichen die Karbonatgehalte in nicht ausgewaschenem Bodenmaterial vielfach 25-35 % · (Ingmar & Moreborg 1976). Die dem Eisriickzug fo lgende Druckentlastung verursachte eine Landhebung, die vermutlich zu Beginn mit einer erheblich hoheren Geschwindigkeit ablief als heute (Morner 1977). Starke Schwankung­ en des Ostsee-Meeresspiegels bewirkten allerdings, das das Nettoresultat, die Strandverschiebung, kei­ nen kontinuierlichen Verlauf hatte, sondern zwi­ schenzeitlich stagnieren oder plotzliche Spriinge aus­ fiihren konnte (Agrell 1979). Aus diesen Griinden laf3t sich der Zeitpunkt des Auftauchens bestimmter Gebiete nur bei vergleichsweise niedrig gelegenen Flachen aus der aktuellen Hohe iiber dem Meeres­ spiegel und der gegenwartig fe ststellbaren Landhe­ bung berechnen. Als ungefahrer Zeitpunkt fiir das Auft auchen grof3erer Teile Nord-Roslagens kann Fig. I. Landhebung im Ostseeraum (mm/a), bezogen auf daher nur grob die Schluf3phase der Litorina-Zeit den jeweiligen Meeres piegeL d. h. die Werte sind nicht hin­ (3000-2000 v.Chr.) angegeben werden (Agrell ichtlich des eustati chen Meeresspiegelanstiegs von ea. 1979). Ll mm/a korrigiert (entnommen aus Ericson & Wallen­ Im Zuge des Auftauchens der von Mergel bedeck­ tinus 1979 S. 18 Fig. 6). ten subkambrischen Gesteinsoberflache werden im­ mer neue Flachen, die bisher relativ unberiihrt auf profile ermittelt, fiir kleinere oder grof3ere Gebiete dem Meeresgrund lagen, in den Bereich der Wasser­ dagegen nur Landhebungswerte angegeben werden, oberflache und damit der Brandung gehoben. Je nach die auf unterschiedlichen Mef3methoden basieren. Exposition und Neigung werden dabei die oberfla­ Untersuchungen iiber die Landhebung gibt es fla­ chennahen Ablagerungen unterschiedlich verandert chendeckend fiir den gesamten Ostseeraum. Es hat (Erosion/Sedimentation). Beim weiteren Anheben sich dabei ein Maximalwert von ea. 10 mm/a fiirein entfernen sie sich schlief3lich aus dem Bereich des Gebiet nordostlich von U mea ergeben, von de m a us standigen Wellenschlags und werden zunehmend ter­ die Landhebung abnimmt bis zu einer Nullinie, die in­ restrischen Bedingungen ausgesetzt, d.h. Nieder­ nerhalb Schwedens etwa zwischen Kalmar und Hel­ schlage, Vegetation, Bodennutzung, etc. werden zu singborg verlauft (vgl. Fig. 1). Wie aus der Karte (die die Bodenbildung beeinflussenden Faktoren, nach­ nicht hinsichtlich des eustatischen Meeresspiegelan­ dem die Brandung einen (durch lokalen Landhe­ stiegs von 1,0-1,1 mm/a korrigiert ist; Wert aus Li­ bungswert und Neigung des Strandes bestimmt) zeit­ sitzin 1964) hervorgeht, betragt die Landhebung in lich begrenzten Einfluf3gehabt hat. Die Spuren dieses Nord-Roslagen ea. 5,5 mm/a im Verhaltnis zur je­ Einflusses konnen mit der Zeit vollig verschwinden, weiligen mittleren Meeresspiegelflache. Diese Werte aber sie konnen auch iiber lange Perioden sichtbar beruhen vor allem auf langfristigen Pegelobservatio­ bleiben, wie z.B. bei Blockmeeren, bei denen die nen an zahlreichen Stationen in Finnland und Schwe­ Brandung aus den oberen dm des Bodenprofilsje gli­ den. Auf dem Festland sind sie aus wiederholten Pra­ ches Feinmaterial ausgewaschen hat. zisionsnivellements hervorgegangen. Bestatigt werden diese Messungen durch die Aus­ wertung der reichlich vorhandenen Katasterkarten Landhebungswerte fii r Hiivero Priistiing aus dem 18. und 19. Jahrhundert (Ase 1964). Durch Aufgrund des lokal unterschiedlichen Ausmaf3es der Aufsuchen und Neuvermessen der in der Regel sehr Erosions- oder Sedimentationsvorgange konnen prazise beschriebenen Fixpunkte konnten zahlreiche Strandverschiebungswerte nur fiir einzelne Strand- einzelne Landhebungswerte berechnet und mittels ei-

Acta Phytogeogr. Suec. 68 Strandverschiebung und Strandvegetation auf Hiivero Priistang 45

Gavle BJorn 5 _ ... 1' I - : Gro is' 120 H �� I"" o Osthornmor �1a..::..- 3 CJ 5 3 " �} 1 1 0 I 0J\' c).,_j Dege/1 5 r b y � Ko b�a- l ., =-t> Uppsolo ' D25/ / k t,nta�r \7 � 0 / F•no;;ko � �to NorrtolJe� _,_,"'"' 7-j;--- ��-;_��" -.G110ronskor SodertoiJe � - 105 -.., ;:-'_ JJ'--100 1/- 0-1.. 95 NyJ,op1n j�{)� '---go o ...... �o ndsort -- km �cev:rrk o ­ 85 0 ,__. � 50I --'--'100 "\em�

Fig. 2. Landhebung im ostlichen Svealand und auf Aland, Fig. 3. Strandver chiebung auf Havero Prastang berechnet auf der Grundlage von Karten verschiedenen 1 829- 1 976. Alters. Die Werte sind eustatisch korrigiert und in einem lndcxsystem mit dem Landhebung: index fur Stockholm = Im i.ibrigen muJ3erwahnt werden, daJ3bei der Kartie­ 00 dargestellt. Gestrichelte Kurven = unsichere Werte; I rung 1829, bei der Ma13stabsangleichung, bei der (e) = extrapolierte Kurven: Periode: Mitte des 18. Jahr­ Luftbildauswertung, sowie bei der nachfolgenden hundcrt bi 1960 (entnommen au : Ase 1964 S. 172 Fig. Projektion Fehlerquellen unvermeidbar waren, die 13). die Aussagekraft der Karte begrenzen. Die scheinbar ner Regressionsanalyse zur jeweiligen Breitenlage in naheliegende Moglichkeit, auf ihrer Grundlage Mes­ Beziehung gesetzt werden. Die dabei entstandene sungen im Gelande durchzufi.ihren, verbietet sich da­ Karte (Fig. 2) zeigt die Landhebung eustatisch korri­ durch von selbst. giert in einem Indexsystem, in dem der gegenwartige Hebungswert fi.ir Stockholm (5,2 mm/a) als Index Differenzierung des Substrats entlang der 100verwendet wird. Uber den Index 130 und die Zuri.ickberechnung der eustatischen Korrektur von Strandlinie 1,1 mm/a ergibt sich fi.ir das hier betrachtete Gebiet Eine Darstellung der das Bodenmaterial der Strand­ eine Landhebung von etwa 5,6 mm/a, also nur eine zone beeinflussenden Vorgange muJ3 vereinfachend innerhalb der Karten-Ablesegenauigkeit liegende Ab­ von der Annahme ausgehen, daJ3 die aus dem Meer weichung von dem durch Pegelablesungen erhaltenen auftauchenden Sedimente fi.ir das behandelte Gebiet Wert. eine etwa gleichbleibende Zusammensetzung haben. Der erforderliche meJ3technischeAu fw and machte Was die mittelschwedische Ki.istenlandschaft betrifft, ein N achvollziehen der von A se beschriebenen Ar­ handelt es sich dabei urn kalkreiche Mergelablage­ beitsschritte fi.ir das Untersuchungsgebiet bisher un­ rungen der letzten Eiszeit auf einer harten moglich. Es gelang jedoch immerhin, mithilfe einer Granitgneis-Unterlage. Dari.iberfinde t sich vielerorts Katasterkarte a us dem J ahre 1829 und den ,ekono­ eine di.inne Schicht rezenter mariner Sedimente, so­ miska kartor" von 1953 und 197 6 einen optischen wie stetig zunehmende Mengen Zivilisationsabfalle, Eindruck vom AusmaJ3 der horizontalen Strandver­ vor allem Roholbestandteile und Ausscheidungen der schiebung zu gewinnen, indem alle drei Karten (nach Zellstoffindustrie. fototechnischer MaJ3stabsangleichung) i.ibereinander­ Die Gesamtmachtigkeit der Sedimente wird u.a. projiziert wurden (Fig. 3). Bei der Betrachtung der durch die Neigung des Strandes und seine geogra­ dabei entstandenen Abbildung sollte man sich aber phische Ausrichtung bestimmt. In Steillagen oder bewuJ3tsein , daJ3die auftretenden groJ3enU nterschie­ gegen die ehemalige Hauptbewegungsrichtung des de der Abstande zwischen Strandlinien verschiedenen Eises (Nordost nach Si.idwest) ausgerichteten Berei­ Alters nicht nur auf die Neigung der betreffenden chen findet man vorwiegend polierte, kahle Felsfla­ Flache, sondern auch auf die morphologischen chen, in flachen Lagen, Mulden oder ,Leeseiten" Veranderungen des Strandes zuri.ickzufi.ihren sind. dagegen Ansammlungen von Material auch aus be-

era rl Phytogeogr. Suec. 68 46 Wolfgang Cramer nachbarten Gebieten. In keinem Fall war die Dauer der Einflul3nahme des Meeres auf die Strandoberfla­ che lang genug, urn den kristallinen Gesteinsunter­ grund nachhaltig zu verandern. Man kann stattdes­ sen sagen, dal3 im gesamten Grol3raum nur ,Freispiilungs- oder Zuschlammungskiisten" auftre­ ten. Abrasionskiisten kommen nicht vor (Hausen 1961). Bestimmend fiir das Ausmal3 des Abtrags oder der Akkumulation von Material sind vor allem die Fakto­ ren Hangneigung und Exposition. An starker geneig­ ten und/oder dem offenen Meer in starkerem Mal3e ausgesetzten Stranden wird vorwiegend Material ab­ gespiilt, in flacheren und geschiitzten Bereichen wird sedimentiert.

Strandtypen Fig. 4. Substrattypen am Strand von Havero Prastang. In vereinfachender Abwandlung der Klassifikationen von Pyokari (1978) und Hausen (1961) bin ich zu fo l­ gender Typisierung der im Gebiet vorkommenden Strande gekommen : Striinde auf kahlem Felsuntergrund (klippstriinder) Strandvegetation auf HiiverO' Priistiing sind offene,glatte Felsflachen, die keine Sedimentauf­ lage haben. Es besteht ein enger Zusammenhang Zur Erfassung der Zonierung der Strandvegetation zwischen der Exposition solcher Flachen und der wurden insgesamt 10 Bandprofile aufgenommen zuletzt aufgetretenen Eisbewegungsrichtung, d.h. der (darunter 2 Doppelprofile). Untersucht wurde jeweils friiheren Eisbewegungsrichtung entgegengesetzte ein 1 m breiter, im rechten Winkel zur Strandlinie ver­ Flachen sind weit verbreitet und in der Regel vollig laufender Profilstreifen, der an beiden Enden durch in eben, wahrend andersartige seltener und starker zer­ den Boden gerammte Eisenrohre markiert wurde. Die kliiftet sind. untere Markierung lag in der Nahe der mittleren Striinde auf stark ausgewaschenen Mergeln (renspo­ Wasserstandslinie, die obere ea. 1-2 m innerhalb der geschlossenen Strauch- oder Waldvegetation. Im lade moriinstriinder). Aus der urspriinglichen Sedi­ Anschlul3 an die Vegetationsaufnahmen wurden alle mentauflage ist hier der grol3te Teil des fein- und mit­ telkornigen Materials ausgewaschen worden. Die Profile durch ein die gesamte Halbinsel umspannen­ oberen dm des Profils bestehen vor allem aus des Nivellement vermessen, urn genauere Angaben iiber Hohenlage und vertikale Erstreckung der ein­ Gesteinsblocken von mehr als 5 cm Durchmesser. zelnen Vegetationsgiirtel zu bekommen. Striinde auf gering ausgewaschenen Mergeln Nach der Kartierung der Verbreitung der verschie­ (moriinstriinder). Die urspriinglicheSedimentaufl age denen Substrattypen entlang der gesamten Strandli­ ist hier besser erhalten, d.h. es kommen noch prak­ nie des Untersuchungsgebiets (Fig. 4) wurden die tisch alle Korngrol3en an der Oberflache vor. Weite­ Standorte fiir die einzelnen Profilaufnahmen an sub­ res Feinmaterial kann sich aber kaum ansammeln. jektiv als ,charakteristisch" fiir den jeweiligen Typ Sandstriinde (sandstriinder) kommen auf Havero erscheinenden Punkten festgelegt. Die insgesamt 12 Prastang nicht vor. Einzelprofile verteilten sich dabei fo lgendermal3en auf die Substrattypen : Striinde mit rezenter organischer Sedimentauflage (/erstriinder) gibt es in flachen, geschiitzten Berei­ kahler Felsuntergrund (2 Profile); chen, vor allem in Buchten. Sie werden mit zuneh­ stark ausgewaschener Merge) (2 Profile); mender Entfernung von der offenen See immer haufi­ schwach ausgewaschener Merge) (3 Profile); ger. Sie bestehen aus im Schilfgiirtel akkumuliertem Merge) mit organischer Sedimentauflage (3 Profile); organischem und mineralischem Material, das die Zwischentypen (2 Profile). urspriingliche Unterlage in einer bis zu mehreren m Bei der Profilaufnahme wurde jeder markierte machtigen Auflage verdecken kann. Strandabschnitt entlang eines Mal3bandes in 50 cm

Acta Phytogeogr. Suec. 68 Strandverschiebung und Strandvegetation auf HiiverO' Priistiing 4 7 langen Teilstiicken liickenlos aufgenommen. Fiirjede dieser 0,5 m2 grof3en ProbefUi.chen wurde der gesam­ te Phanerogamenbestand erfaf3t und mithilfe einer geringfiigig modifizierten Bedeckungsgradskala nach Hult-Sernander-Du Rietz quantitativ aufgenommen. Obwohl bei der Beschreibung der Vegetationspro­ file teilweise pflanzensoziologische Methodik verwen­ det wurde, war es niemals Ziel der Gelandeaufnah­ men, die Strandvegetation im Untersuchungsgebiet vollstandig pflanzensoziologisch zu bearbeiten. Es ging nur darum, die besonderen Merkmale der Vege­ tation auf den unterschiedlichen Substrattypen mog­ lichst anscha ulich wiederzugeben.

Zusammenstellung der wichtigsten vor kommen­ den Vegetationsgiirtel Auf fe lsigem Untergrund Die kahlen Felspartien am Strand von Havero Prast­ ang tragen hohere Vegetation erwartungsgemaf3 nur dort, wo Bodenmaterial, das sich in Spalten oder zwischen groben Blocken angesammelt hat, einen geeigneten Untergrund bieten. Die raumliche Vertei­ lung der Pflanzen orientiert sich daher stark am Vor­ handensein solcher Wuchsorte. Trotzdem la13t sich eine grobe Zonierung erkennen, die im Bereich der mittleren Wasserstandslinie (MWL) mit einem Fig. 5. Beinahe kahler Fels mit dunner Vegetation in den schmalen Juncus gerardii -Agrostis stolonifera - Spaltcn. I m Vordergrund Hippoph ae· rhamnoides (August Giirtel, der diinn mit kleinen Phragmites-Pflanzchen 1979). und Triglochin maritimum durchsetzt ist, beginnt. Bei anhaltendem Vorkommen von Agrostis fo lgt ein arundinacea -Aster tripolium -Giirtel, der nach oben Festuca rubra - Phalaris arundinacea -Giirtel mit von einem vergleichsweise dichten Juncus gerardii ­ (u.a.) Rhinanthus angustifolius, Festuca arundinacea Agrostis stolonifera -Giirtel abgelost wurde. Dort und Valeriana salina. Ab etwa 50 cm oberhalb der wachsen auch Carex extensa und Plantago mariti­ MWL geht die Vegetation dann in lokal variierende, ma. Ab ea. 25-30 cm steigen dann allmahlich die weniger deutlich zonierte Gesellschaften iiber, die bei Anteile von Rhinanthus angustifolius, Festuca rubra den untersuchten Profilen z.B. Geranium sangui­ und Valeriana salina bis hin zum Hippophae rham­ neum, Lysimachia vulgaris und Filipendula vulgaris noides -Giirtel, der vor allem von Geranium sangui­ enthielten. Der zum W aldrand iiberleitende Hippo­ neum, Centaurea jacea und Filipendula ulmaria phae· rhamnoides -Giirtel ist nur schmal und liicken­ durchsetzt ist. haft ausgepragt. A uf schwach ausgewaschenem M ergel Aufaus gewaschenem Merge! An diesen im Gebiet weit verbreiteten Strandab­ Dieser Typus ist nur an einer Stelle (nordostlich der schnitten ist die auswaschende Wirkung der Bran­ Badeklippe) vorhanden. Zwischen den groben Block­ dung erheblich geringer als beim vorhergehenden en der Bodenoberflache sammelt sich erst in einiger Typus. Es kann sich daher auch im Bereich des Wel­ Entfe rnung vom Brandungsbereich ausreichend lenschlags durchwurzeltes Feinmaterial iiber den Feinmaterial fiir eine dichtere Vegetationsdecke. Un­ Winter halten, wodurch die N eukeimung im Friihjahr terhalb davon werden die aus den Spalten hervor­ erheblich erleichtert wird. In der N ahe der MWL kommenden Pflanzen des Vorjahres durch Herbst­ wachst hier ein dic hter Juncus gerardii -Agrostis stiirme un d winterliche Eisbewegungen weitgehend stolonifera -Giirtel, in dem Glaux maritima und bes,eitigt, so da13 sich erst im spateren Friihjahr neue Triglochin maritimum haufig sind. Dariiber schlief3t Keimlinge halten konnen. Unmittelbar an der MWL sich ein Festuca rubra -Giirtelmi t Odontites litoralis, entwickelte sich immerhin ein lockerer Phalqris Rhinanthus angustifolius und Carex extensa an. In

A eta Phytogeogr. Suec. 68 48 Wolfgang Cramer

Fig. 6. Massig ausgewaschener Mergelstrand am nordlichen Strand der Halbinsel Lilloren. lm Yordergrund Phalaris arundi­ nacea, Juncus gerardii und Asrer tripolium. and Waldrand i.iber­ wiegend Hippophae· rhamnoides ( August 1979).

geschiitzteren Lagen treten hier auch dichte BesHinde tha palustris und Carex nigra. Ein Streifen mit Festu­ von Scirpus uniglumis und (verstreuter) Scirpus ta­ ca arundinacea, Geranium sanguineum und Lysima­ bernaemontani auf. Der Ubergang zum H ippophaii chia vulgaris bildet den Ubergang zu einem schmalen rhamnoides -Gebiisch, das in solchen Lagen seine Hippophae· rhamnoides -Giirtel, der bereits vom sich gr6J3te Breite und Dichte erreicht, wird von Sonchus anschliel3enden Alnus glutinosa -Bestand iiberschat­ arvensis, Leontodon autumnalis, Angelica archange­ tet wird. lica ssp. litoralis, Ophioglossum vulgatum und an­ deren gebildet. Die Grenzen zwischen den einzelnen Giirteln sind an solchen, meist flachen und weitUiufi­ W echselwirkungen zwischen Strandverschie­ gen Std:inden oft so unscharf, daB man eine starke bung und Vegetationsdecke Variabilitat von Jahr zu Jahr, in Abhangigkeit vor all­ Die Zonierung der Strandvegetation im Bereich der em von den Bewegungen der W asserstandslinie im mittleren und nordlichen Ostsee ist naturbedingt sich Friihjahr, vermuten kann. kontinuierlich wandelnden Umweltbedingungen un­ terworfen. Jeder von einer Pflanze im Strandsaum kolonisierte Standort wird innerhalb absehbarer Zeit Aufrezenten organischen Sedimenten aus dem unmittelbarer Einflussbereich des Meeres Dieser Typus wird von den am starksten gegen Wind entfernt und dadurch ganzlich anderen Bedingungen und Wellen geschiitzten Stranden vertreten, die vor unterworfen, z.B. durch den Fortfall der Konkurrenz­ allem in lnneren von Buchten, zum Teil auch auf der fa ktoren Salz, Uberflutung, etc. Fur die Differenzie­ Leeseite flacher lnseln zu finden sind. Kennzeichnend rung der Vegetation unterschiedlicher Strandab­ ist zunachst der viele m breite Phragmites communis schnitte sind deshalb nicht nur die aktuell vorhande­ -Giirtel, der die Wasserbewegung stark herabsetzt nen unterschiedlichen Lebensbedingungen, sondern und dadurch die Voraussetzung fiir starke Sedimen­ auch Geschwindigkeit und Ablauf der Strandver­ tation schafft. lm Schutze dieser dichten Schilfbe­ schiebung ausschlaggebend. Einige Aspekte der stande kommen im Bereich der MWL reichlich Scir­ Wechselwirkungen zwischen der Strandverschiebung pus uniglumis, Mentha aquatica und Triglochin und der Vegetationsdecke sollen zum Abschluss prin­ maritimum vor; hinzu kommen in deutlich abgesetz­ zipiell dargestellt werden. ten Giirteln S cirpus tabernaemontani ( -10 bis -4

cm) und Plantago maritima ( -5 bis + 1 cm). Der Juncus gerardii -Agrostis stolonifera -Giirtel leitet Striinde mit fe lsigem Untergrund aus dem Schilfbereich zur zweiten Hauptgruppe, der Da hier kaum Anhaufung oder Abtrag von Feinma­ Strandwiesen(,strandang")-Vegetation iiber. Dazu terial stattfindet, ist die Strandverschiebung von gehoren u.a. Galium uliginosum, Glaux maritima, Landhebung und Meeresspiegelbewegungen direkt Ophioglossum vulgatum, Peucedanum palustre, Cal- abhangig. Die Vegetation entfernt sich daher mit

A eta Phytogeogr. Suec. 68 Strandverschiebung und Strandvegetation auf Hiivero Priistiing 49

Fig. 7. Doppelprofil uber zuwach­ sende. kleine Bucht zwischen Stor­ oren und Festland. Vorn und am linken Rand Hippophae·- Gebusch, im Wasser (hinter dem Mal3band) dichte Bestande von Scirpus uni­ glumis und Se. tabernaemontani, im Hintergrund Phragmites com­ munis (August 1979, alle Auf­ nahmen vom Verf.).

konstanter Geschwindigkeit aus dem Brandungsbe­ Striinde mit organischer Sedimentaujlage reich, d.h. es konnen nach und nach immer salz- und Im Bereich dieser binnenseeartig anmutenden iiberflutungsempfmdlichere Arten einwandern, die Wiesen- und Schilfstrande findet die starkste Strand­ die eigentlichen Halophyten allmahlich verdrangen. verschiebung statt. Das dichte Wurzelsystem der Schilfpflanzen iiberdauertden Winter (von den auJ3e­ S triinde auf ausgewaschenem Mergel ren Randern abgesehen) unbeschadet und bildet so­ Diese in der Regel spiirbar geneigten und dem Wel­ fo rt nach der Eisschmelze eine gute Grundlage so­ lengang ausgesetzten Strandabschnitte weisen auf­ wohl fiir den Neuaufwuchs der Vegetation als auch grund des starken Feinmaterialabtrags die geringsten fi.ir das durch das Wasser herangetragene Sediment, Strandverschiebungswerte auf. Die Vegetation des so da/3 bald auch anspruchsvollere Arten gedeihen Vorjahres wird im Winter zum groJ3tenTeil entfernt, konnen. Die groJ3en Mengen Biomasse, die in diesem so da/3 nur kurzlebigere Arten vorkommen konnen. Okosystem jahrlich produziert werden, fiihren Erst oberhalb des unmittelbaren Brandungsbereichs schnell zur Bildung eines torfartigen organischen Se­ gelingt es mehrjahrigen Pflanzen, sich anzusiedeln diments, durch das die Strandverschiebung in giinsti­ und dadurch auch den Bodenabtrag zu verringern. gen Fallen vervielfacht werden kann. Aus der von ihnen produzierten Biomasse und an­ In alien Bereichen von Landhebungskiisten finden gewehtem Bodenmaterial entsteht allerdings einige m also Sukzessionsablaufe statt, die gegeniiber den an oberhalb bereits wieder ein Bodenprofil, auf dem sich Stranden ohne Landhebung ablaufenden noch urn selbst eine geschlossenere Strauchschicht etablieren eine Stufe komplexer sind. Die Eigendynamik der kann. Pflanzengesellschaften wird durch die zunehmende Entfernung vom Meer so stark iiberlagert, da/3einzel­ Striinde aufkaum ausgewaschenem Merge/ ne Vegetationsgiirtel kaum Zeit haben, sich voll zu Die Abtragungsvorgange, die auch in diesen Berei­ entwickeln, bevor bereits wieder neue, aufgrund der chen wirksam sind, werden durch die erheblich dich­ veranderten Bedingungen konkurrenzstarkere Arten tere Vegetation stark in ihrer Wirkung gemildert. V on einwandern konnten (Schwanck 1974). Es gibt meh­ einzelnen Sturmflutereignissen abgesehen, beschadigt rere Griinde fiir die Tatsache, daB es bisher trotz die Brandung die Pflanzendecke nur in geringem groJ3er Materialanhaufung aus dem gesamten Masse, so da/3 sich in gegenseitiger Bedingtheit eine Bereich der finnischen und schwedischen Ostseekiiste reichhaltigere Vegetation und ein feinkornigeres Sub­ nicht moglich war, ein komplexes Gesamtbild des strat ergeben. Vor allem aufgrund der gelegentlich Einflusses der Strandverschiebung auf diese Sukzes­ eintreffenden Sturmfluten oder extremer Eisschuber­ sionsablaufe vorzustellen : eignisse wird der theoretische Wert der Strandver­ Der Vergleich von Gebieten mit unterschiedlich scltiebung aber auch hier unterschritten werden. hohen Landhebungswerten wird durch die edaphi-

A eta Phytogeogr. Suec. 68 50 Wolfgang Cramer schen und klimatischen Variationen der Umweltbe­ Du Rietz, G.E. 1925. Die Hauptzuge der Vegetation des dingungen wesentlich kompliziert. ausseren Scharenhofs von Stockholm. - Svensk bot. Tidskr. 19:34 7-369. Der Verlauf der Ostsee-Wasserstandskurve ist so Ericson, L. 1977. Strandvegetation vid Hoga kusten i Ang­ unregelmal3ig, dal3 brauchbare Messungen der ermanland. - Svensk bot. Tidskr. 71:338-4 14. Strandverschiebung nur iiber langere Zeitraume Ericson, L. & Wallentinus, H.G. 1979. Sea-shore vegeta­ durchfiihrbar sind; ebenso ist es infolgedessen auch tion around the Gulf of Bothnia. - Wahlenbergia m it den A uswirkungen der Strandverschiebung. 5:1-142. Hausen, H. 1961. Kort oversikt av de alandska strandty­ Zu den storenden Abweichungen der Wasserstands­ perna. - Acta bot. fenn. 61: 23-28. kurve von ihren jeweiligen theoretischen Jahresmit­ Hayn!n, E. 1948, Skargardens langszoner. - Skargards­ teln kommt hinzu, da13 jeder Extremwert nach oben, boken. Helsingfors. pp. 242-256. wie Herbststiirme,extrem starker Eisschub o.dgl. die Ingmar, T. & Moreborg, K. 1976. The leaching and origi­ Kolonisation gro13er Strandbereiche urn Jahre nal content of calcium carbonate in till in northern zuriickwerfen kann, z.B. durch ungewohnlich Uppland, Sweden. - Geol. For. Stockh. Forh. schwere Beschadigung des Wurzelsystems der Schilf­ 98:120- 132. bestande. Knapp, R. 1971. Einfuhrung in die Pflanzensoziologie. - Stuttgart. 388 pp. Larsson, B.M.P. & Poder, V. 1962. Svenska Vaxtgeogra­ Beeinflussung des theoretischen Sukzessionsab­ fiska Sallskapets exkursion till Havero Prastang den 3.6.61. - Svenskt bot. Tidskr. 56:183-188. laufs durch anthropogene Einfliisse Lid, J. 1974. Norsk og svensk flora. 2 utg. Oslo. 808 pp. Die Beobachtung standig ,jungfraulich" aus dem Lisitzin, E. 1964. Contribution to the knowledge of land Meer auftauchenden Neulandes darf nicht dariiber uplift along the Finnish coast. - Fennia 89(4):1-22. hinwegtauschen, da13 auch in diesen Bereichen die Morner, N.-A. 1977. Past and present uplift in Sweden : glacial isostasy, tectonism and bedrock influence. - natiirlichen Sukzessionsablaufe stark durch den Geol. For. Stockh. Forh. 99:48-54. Menschen gestort werden. Bereits vor dem Auftau­ Palmgren, A. 1961. Studier over havsstrandens vegetation chen wird der Meeresboden im stark industrialisier­ och flora pa Aland. I Vegetationen. - Acta bot. fe nn. ten Ostseeraum stark verschmutzt, vor allem durch 61:1-268. Roholbestandteile und Abfalle aus der Zellstoffin­ Palomaki M. 1963. Uber den Einflussder Landhebung als dustrie. Die stoffiiche Zusammensetzung des Aus­ okologischer Faktor in der Flora flacherIns eln. - Fen­ gangsmaterials fiir die Bodenbildung im Strandbe­ nia 88(2):1 -75. reich kann also vielerorts kaum noch als ,natiirlich" Pyokari, M. 1978. Airiston alueen rantatyypeista (Shore­ bezeichnet werden. Dazu kommen die Belastungen types in the Airisto area, SW Finland). - Terra durch die intensive Nutzung der Scharenkiisten als 90:8 1-91. Schwanck, B. 1974. Vegetationens forandring pa de fr am­ N aherholungsgebiet. vaxande skaren. - Skargard i omvandling. Helsingfors. Die Vergangenheit hat gezeigt, da13 die Auswei­ pp. 79-85. sung einzelner Inseln oder kleinerer Archipele so­ Sjors, H. 1971. Ekologisk botanik. - Uppsala. 296 pp. lange keine Aussicht auf Erhalt oder Wiederher­ Skargard i omvandling. Miljo och manniska i Finlands stell ung dieses einzigartigen N aturlandschaftstyps skargard. 1974. - Helsingfors. 254 pp. bietet, wie nicht erheblich starker restriktive Vor­ Skargardsboken, 1948. - Helsingfo rs. 699 pp. schriften vor allem fiir die industrielle Nutzung des Ussisoo, I. 1977. Computation of land uplift and mean sea Ostseeraums erlassen und eingehalten werden. level in Sweden. - Geol. For. Stockh. Forh. 99:42-48. Wallentinus, H.G. 1973. Above-ground primary pro­ duction of a Juncetum gerardi on an Baltic sea-shore meadow. - Oikos 24:200-2 19. LITERATUR VERZE ICHNIS Valovirta, E.J. 1937. Untersuchungen uber die sakulare Agrell, H. 1979. Kvarti:irtiden - Uppland under 2 miljar­ Landhebung als ptlanzengeographischen Faktor. - der ar. Uppsala. pp. 4-7. Acta bot. fe nn. 20:1-161. Cramer, W. 1979. Litteratur om havsstrandvegetationen i Ase, L.-E. 1964. Nivaforandringen i ostra Svealand och Sverige och Finland - Medd. Vaxtbiol. Inst. Uppsala Aland beraknad med utgangspunkt fran aldre kartma­ 1979:4. (Mimeographed) 23 pp. teriai. - Ymer 84: 121-182.

Wolfgang Cramer, Geographisches Institut, Justus Liebig-Universitat, Senckenbergstr. 1, D-6300 Giessen, West Germany

A eta Phytogeogr. Suec. 68 The Siberian Fern Athyrium crenatum (Somf.) Rupr. fo und in Sweden at K vikkjokk, SW Lule Lappmark Sten Dahlskog

A new Swedishfern (Br0gger 1960 pp. 44-45). In newer floras the On July 12th, 1979, Ms. Birgit Dahlskog, Mr. Bj orn species is often treated under the names Diplazium Sarstad and I guided an excursion for the Swedish sibiricum (Turcz. ex G. Kunze) Jermy (Flora Tourist Association's first floristical course in Europaea 1964) or Diplazium sibiricum (Turcz. ex K vikkjokk to a wooded mountain slope about 6-8 G. Kunze) Kurata (Nakaike 1975). Kurata has km to the E of K vikkjokk village in SW Lule Lapp­ priority. However, the name used here will be Athyrium crenatum mark. When I went ahead to find a place to in accordance with Lid 1963 demonstrate the geological strata, I noticed some (whose nomenclature is used throughout this paper) ferns without having time to think more than and Hulten 1971. "phegopteris cannot look like this" and snatched some fronds. Back in the village we determined them Physiognomy to Athyrium crenatum (Somf.) Rupr. (Sw. 'ryssbdiken'; Russian fern), and sent some specimens Athyrium crenatum has a creeping rhizome and the to Prof. Bertil Nordenstam, Swedish Museum of fronds grow separately (Fig. 2). It has often been Pteridium Natural History, Stockholm, who verified the deter­ compared to a small (Dahl 1949 p. 13; mination. A later visit to the locality on August 8th, Kalliola 1958 p. 88), but in my opinion it is even 1979, confirmed that the stand was large and richly more similar to Dryopteris phegopteris, and from a sporing. distance it would be easy to confuse the two species. The find is the first in Sweden of a species with an From some meters one notes, however, that not only interesting and often discussed distribution (Fig. 1). A the lowest lobes but also a couple of the lobes im­ statement by Hartman (1879 p. 537) about a locality mediately above them are directed backwards­ A. crenatum in Jamtland in Central Sweden has by Rosendahl upwards: lacks the sharp divergence in ( 1913 p. 285) been explained to be an incorrect deter­ lobe orientation which is so characteristic of D. phegopteris. A. crenatum mination of a juvenile form of Cystopteris montana A stand of makes a very D. (Lam.) Bernh. and has been deleted from later elegant, graceful impression and makes phegopteris A. literature. look rigid and clumsy. Furthermore, crenatum grows taller and larger, the laminae are broader and more lobate, almost equilaterally Nomenclature triangular, the colour is a darker, pure green without A thyrium crenatum was first described by a yellow tinge, and the sparsely black-scaled petioles Sommerfelt in 1835a under the name Aspidium have a striking gray-green lustre. Sporing laminae crenatum from Kringen in Sel in Norwegian Gud­ have more or less straight sporangia. The laminae are brandsdalen, where the fe rn according to M.N. Blytt (very) sparsely fine-haired on the underside. When ( 1861 p. 22) was discovered by Wahlberg. The broken the petioles show two vascular bundles (but D. phegopteris). species has been included in the genera Asplenium, so do those of Polypodium, Cystopteris and Allantodia, but has generally been considered an Athyrium. The fo rm of the inducia and the chromosome number of the The K vikkjokk locality

Norwegian populations (n = 41) would, however, In the alpine fo reland in SW Lule Lappmark the make it reasonable to consider the species a member rivers run in glacially eroded valleys through a pre­ of the predominantly tropical genus Diplazium tertiary peneplain, which has been raised to

A eta Phytogeogr. Suec. 68 52 Sten Dahlskog

Fig. I. The world distribution of A thyrium crenatum (Somf.) Rupr. according to E. Hulten (unpubl.), revised in the Urals according to lgoshina 1969, and in Fennoscandia.

800- 1 000 m above m.s.l. and forms the rather flat­ about 1.7 5 m), exposed block shore, whose zonation (as peaked massifs Lastak to the SW of the Lilla Lule those of the other block shores of Lake Saggat) may be River and Kabla to the NE of the river. From about described in the terms used by Wassen (1966 pp. 94-98). The Subularia zone is poor in species, and the polygene 600 m the north-facing, conifer-wooded precipices of formation is not particularly distinct. The higher Carex Lastak drop into the uppermost big lake of the valley, juncella- and Calluna -zones (Calluna vulgaris Molinia + the still unexploited Lake Saggat (302 m above m.s.l.) coerulea) are rich in species with luxuriant stands of which occupies the overdeepened bottom of the (in demanding species as Trichophorum alpinum, Tofieldia the western !1art) double-sided U -valley. The A. pusilla, Thalictrum alpinum, Saxifraga aizoides and crenatum locality is on a northern slope in this area. Sedum rosea, particularly along creeklets flowing out of the fo rest. Carex parallel/a (the easternmost locality in the The bedrock is Cambro-Silurian, and the strata sequence valley) and Dactylorhiza maculata s.l. occur here. The is probably the same as on the opposite shore of the lake in Va ccinium vitis-idaea -zone is narrow and indistinct. The the S-facing precipice of Mt. Kabla, from where Svenonius lower limit for Vaccinium myrtillus is more or less coinci­ ( 1900 p. 29 3) described the strata along the creek Kad­ dent with those for Equisetum silvaticum, Cystopteris tj ajokk as follows: from below (a) red gneissic granite (25 montana, Dryopteris linneana, Melica nutans, Carex m), (b) conglomerate, (c) yellow sandstone, (d) sandstone capillaris, Gymnadenia conopsea, Rubus saxatilis, shale (b-d = 7.5 m), (e) clay shale (10.5 m), (f)gray quart­ Geranium silvaticum, Cornus suecica, Ramischia secunda zite (6 m), (g) clay shale, (h) blue quartz with sandstone and Phyllodoce coerulea. shale, (i) greengray shale and heavy layers of black shale Above the inundation zone the fo rest is virginal, swam­

(g-i = 40.7 m), and (j) 198 m above the lake level a steep py, partly a rich fen with Equisetum variegatum, rock-face of metamorphosized clay shale. The A. Cystopteris montana, Juncus biglumis, Eriophorum crenatum locality has a similar wall-faced and partly brachyantherum, Gymnadenia conopsea, Salix myrsinites, overhanging precipice. The layers dip towards the W. S. reticulata, Epilobium davuricum, Pinguicula alpina and The inundation zone is a narrow (the lake's amplitude is P. villosa. Among the more commonplace species in this

A eta Phytogeogr. Suec. 68 The Siberian fe rn Athyrium crenatum fo und at Kvikkjokk 53

linnaeana, D. phegopteris, Maianthemum bifolium, Poa nemoralis, Calamagrostis lapponica, Goodyera repens, Actaea erythrocarpa, Ribes rubrum (coli.), Rubus arcticus, R. saxatilis, Geranium silvaticum, Chamaenerion angustifolium, Trientalis europaea, Campanula rotun­ difolia, Saussurea alpina and Solidago virgaurea ; on blocks and in crevices can be fo und Festuca ovina, Empetrum sp., Moneses uniflora and Vaccinium vitis­ idaea. (Melica nutans, Roegneria canina, Aconitum septentrionale, Daphne mezereum, Val eriana sam­ bucifolia and Cirsium heterophyllum are more prominent outside the Athyrium crenatum stand.) The bottom layer is mostly very sparse but locally closed and dominated by Pleurozium schreberi, Hy locomium splendens and Rhytidiadelphus triquetrus. Above the scree slope there are partly overhanging precipices of clay and alun shales. In these occur, among Fig. 2. Athyrium crenatum (Somf.) Rupr. on the Kvikk­ others, Woodsia glabella, Dryopteris spinulosa, Carex jokk locality, SW Lule Lappmark, Sweden. Photo: Sten atrata, C. atrofusca, C. ornithopoda, Salix reticulata, Dahlskog. Aug. 8, 1979. Saxifraga adscendens, S. caespitosa, S. nivalis, Fragaria fen are Equisetum scirpoides, Tofieldia pusilla, Carex vesca and scattered Dryas octopetala (the easternmost capillaris, Dactylorhiza maculata s.l., Th alictrum known locality in the valley). alpinum, Bartsia alpina, Pinguicula vulgaris and Saussurea alpina. Between this wooded fen and the A. In these districts comparisons with the results of crenatum scree slope is a luxuriant, virginal spruce fo rest, earlier florists are made difficult by the fact that rich in species with, among others, Equisetum scirpoides, place-names have been in a state of flux so that a Luzula parviflora, Goodyera repens and Listera cordata; mountain may have as many synonyms as a fern, pine is almost totally absent. and on the new maps the names used by earlier in­ The largest boulders in the lower scree slope are of cot­ vestigators may be found in quite other places than tage size. The latest big slide probably occurred in the the original. It is, however, quite certain that the beginning of the 1960's. In the lowest, driest, least wooded vicinity of the A. crena turn locality was visited by E. and most bouldery part of the slope Stellaria longifolia, Nyman on several occasions between 1891 and Rubus idaeus and tall Dryopteris assimilis (det. J.A. Nannfeldt) are prominent. The upper part of the scree 1893, and that Selander did not visit the area. It is slope includes both drier areas and moister areas with possible that no botanist afterNy man visited the area smaller boulders and more species: the latter areas seem to until Birgit Dahlskog and I went there in 1967 be less exposed towards the prevailing winds and probably (without finding A. crenatum). However, the stand is have a thicker snow pack. The slope probably receives so large and luxuriant that it seems improbable that direct sunshine only during fo renoons, at the end of July A. crenaturn should have colonized the locality after hardly after 11.00 (summer-time). The slope is about 1893. The fern is probably an old inhabitant because 45-50° and sparsely wooded (spruce, birch, rowan, it has succeeded in occupying both rich and more juniper). The field layer in the drier areas is similar to those poor environments. in mixed heath woods that are not particularly rich in The inaccessibility is probably the best protection species, with Hy locomium splendens, Maianthemum bifolium, Rubus arcticus, Chamaenerion angustifolium, for the locality. The wooded strip is too narrow for Vaccinium vitis -idaea, Linnaea borealis-and, in the economical modern logging. The ground belongs to vicinity of the richer areas, Athyrium crenatum, too. The the government. No paths run in the neighbourhood, richer, moister areas have luxuriant tall-herb vegetation and none are likely to be made. The locality is (especially along the trickling creeklets) or closed, pure probably as virginal and undisturbed as any present­ one-species stands, but crevices with thin soil and boulders day part of Sweden can be. I consider it quite possible fo rm enclaves with poorer vegetation. In these rich areas that nobody set foot there between 1967 and 1979. A thyrium crenatum luxuriates among the boulders along a Whatever the means that dispersed A thyrium distance of at least 100 m, 330-370 m above m.s.l., partly crenatum to the locality, it was surely not by human in pure stands with fronds so close that a bottom layer is agency. completely missing. The petioles reach 41 cm in height (above ground), and the richly sporing laminae reach 30 cm in length the maximum length according to (= Distribution Hylander 1953 pp. 33-34). The fo llowing species were growing in the company of Athyrium crenatum (Fig. 1) belongs to a small group Atlzyrium crenatum: Lycopodium annotinum, Dryopteris of easterly (= with centra in Northern Russia and

A eta Phytogeogr. Suec. 68 54 Sten Dahlskog

Siberia) species with isolated occurrences in the most mixed woods, in rather dry mossy soil". lgoshina continental parts of S. Norway. This group includes (( 1 966) 1969 p. 193) describes the habitats in the Aster sibiricus, Carex rhynchophysa, Cinna latifolia, Urals as "dark-conifer mountain taiga and in sub­ Clematis sibirica, Cystopteris sudetica and Glyceria alpine open forests". lithuanica (Br0gger 1964 p. 184). The lichens In Finland, A. crenatum is generally considered to Ramalina dilacerata, R. obtusata and (to some ex­ be a demanding species. Separate Finnish and/or tent) Evernia divaricata may be said to have a similar Far-Carelian localities are reported or described by, distribution (Hjelmstad, in letter). Because of their among others, Auer (1944 pp. 45, 53, 56), Backman mysterious appearance in deep fo rests these plants (1914), Brandt (1933), Brenner (1899), Hjelt (1888), have been called "huldra plants" after the "lady of Kaakinen (1972), Kallio & Nikoskelainen (1954), the woods" of Scandinavian mythology (Norw. Linkola (1921 pp. 158-1 59), Norrlin (1870 p. 152, 'huldreplanter', Ove Dahl according to Nordhagen 1878 p. 14), Pankakoski (1939 p. 39), Pesola (1929 1943 p. 66), and they provide one of the most p. 209, 1934, 1955 a and b), Sonck (1964) and enigmatic problems of Nordic plant geography Valmari (1921). According to Linkola (1921) and (Nordhagen 1943 p. 66; Br0gger 1964 p. 183-186; Kalliola (1958) the Finnish habitats are shady, moist Hjelmstad 1978 p. 1 7 5). groves or grove fe ns (Fin. 'lehto') with thick mull soils, conspicuously often located at the foot of a .Closest to the Kvikkjokk locality of Athyrium crenatum mountain, in a blocky slope or at a rapid creek. are the known Norwegian localities in Gudbrandsdalen Kalliola reports that pH in the substrate according to 700 km to the SW of K vikkjokk and a Finnish locality in Utsjoki 465 km to the NE of K vikkjokk. The Utsjoki Pankakoski is mostly higher than 6.0, but the judg­ locality, discovered in 1954, is somewhat more than 300 ment seems to be based on only nine values (ranging km to the NW of the next closest localities on the Kola from 5.6 to 6.9) from Hiisjarvi in Salla (in the fo rmer­ Peninsula (Kandalaksha - Kirovsk - Hibina) and in ly Finnish part of Carelia) (Pankakoski 1939 p. 39). Kuusamo (Laine, Lindgren & Makinen 1955 p. 124). In Peso la considers A. crenatum to be fa voured by lime Gudbrandsdalen in Norway the species is known from in Carelia and to demand lime in Kuusamo (1929 p. three localities in Se! (Rosten above Laurgard, Kringen 209). and along River Ula), one locality in S0r-Fron (Storuren at Among the companion species of A. crenatum Kalliola Vig), and two localities in Ringebu (Randkleiv bru and mentions "large fe rns", Lastrea species, Agrostis clavata, across the river at Elstad) (M.N. Blytt 1861 p. 22; Dahl Actaea spicata and towards the north A. erythrocarpa, 1949 p. 13; Nordhagen 1921 pp. 137-1 38). Two more Cirsium heterophyllum, and Crepis paludosa. Brandt and localities in Gudbrandsdalen are reported to have been dis­ Peso! a give analyses of quadrats (I00- 1 600 m 2) in Carelia covered and will soon be described. Flora Europaea I and Kuusamo. Kaakinen 1972 gives summaries of ( 1964 p. 18) states that the species occurs in a small area in quadrat analyses from Sotkamo and Paltamo in the Ka­ N. Norway, but this is a misprint fo r S. Norway and refers jaani district. He assigns the fo rests in which A. crenatum to Gudbrandsdalen; the species is, as yet, not known fr om is most prominent to a Geranium - Oxalis -Filipendula N. Norway. In eastern Finland A. crenatum is considered -type, growing on soil with pH between 5.6 and 6.3 (1972 p. to be rare and local (Kallio, Laine & Makinen 1969 p. 82), 3 7) and considers A. crenatum to prefer lower than normal but scattered localities occur to the Lake District, the acidity. According to Kaakinen the most prominent westernmost in southern Tavastland at Asikkala (Norrlin species in the field layer of this forest type are (values 1870 p. 152) and Hattula (Kalliola 1958). denote percent of cover in 18 quadrats each of 25 m2 and The distribution in the European part of the Soviet percent of frequency): Filipendula ulmaria 21.4:94, Ox­ Union is discussed by Minayev (1965) 1968, who gives alis acetosella 8. 7 : 94, Geranium silvaticum 7.4 :94, Viola two maps (p. 50, 61). The Asian distribution extends, ac­ epipsila 6. 1 :89, Maianthemum bifolium 2.5 :89, cording to Flora of the USSR (( 1934) 1968 p. 46), through Equisetum silvaticum 3.0 :83, E. pratense 2. 1 :83, Trien­ Siberia to Sakhalin, Japan and China. According to talis europaea 2.0 : 78, Paris quadrifolia 6.8 : 72, Nakaike (1975 p. 166), A. crenatum occurs on Honshu. Dryopteris linnaeana 5.0 : 72, A thyrium filix-fem ina Localities in Manchuria are reported by Christensen (1924 2.9 :67, Crepis paludosa 2.1:61, Dryopteris phegopteris p. 43) and in northern Shanxi (Shensi) by Christ ( 1910 p. 3.5 :50 and Rubus saxatilis 1.7:50. Matteuccia 192). An isolated locality in Kashmir is marked on one of struthiopteris has the values 13.6 : 33, A thyrium crenatum Hulten's unpublished map manuscripts (courtesy Swedish 2.8 : 6, Chamaenerion angustifolium only 0.2 : 11 and A c­ Museum of Natural History). Var. glabrum (Tagawa) taea spicata 0. 1 :6 (p. 34). The dominating trees are Alnus Kurata occurs on Hokkaido, Honshu, Shikoku and in incana, Prunus padus, Betula pubescens, Picea abies and Korea (Nakaike 1975 p. 166). Sorbus aucuparia. Kaakinen has assigned one quadrat with Athyrium crenatum to a Geranium - Oxalis - Paris Habitats -type. Flora of the USSR ((1 934) 1968 p. 46) describes the The northernmost Finnish locality in Utsjoki seems habitats of Athyrium crenatum as "coniferous and to diverge from those just mentioned. According to

A eta Phytogeogr. Suec. 68 The Siberian fe rn thyriumA crenatum fo und at Kvikkjokk 55

VI�STRA-SOLSTAD 249m KVIKKJOK K 332 m 0.6· 584 UTSJOKI NUORGAM -1.7' 417 KAJAANI 140m 1.3· 627 KUUSAMO 266m ·0.7' 606 1931 - 1960 1931 - 1960 j 20m

Fig. 3. Climate diagrams for stations in the vicinity of some Athyrium crenatum localities in Norden. Vinstra-Solstad in Gudbrand dalen in S Norway from data supplied by Oet Norske Meteorologiske lnstitutt (Norwegian Institute of Meteorology), K vikkjokk in N Sweden from Dahlskog (1972), Kajaani and Kuusamo in E Finland and Utsjoki Nuorgam in N Finland from Walther & Lieth (1967). One scale interval = I0°C or 20 mm of precipitation. Numbers to the right of the station name indicate m above m.s.l., yearly mean temperature and yearly mean precipitation. Index of occanity. see the text.

Kallio, Laine & Makinen (1969 p. 82) it is a steep Continentality and easterly features in the flora valley at about 140 m above m.s.l., in the birch belt, of the K vikkjokk area with boulder fields, a fairly thin humus layer and with running water; the local climate may have continen­ The continentality of the climate of Nordic Athyrium tal features. The locality at Kandalaksha on the Kola crenatum localities can be expressed by the oceanity Peninsula seems also to be situated in the birch-wood index of Kotilainen, which increases with increasing region according to Hjelt (1888 p. 3 7). In Utsjoki, A. oceanity. It is 29 in the continental 0sterdalen in crenatum is growing in the company of Angelica Norway, 38 in Helsinki, and 386 in Bergen on the archangelica, Arabis alpina, Luzula parviflora, Norwegian west coast (Kotilainen 1933). Melica nutans, Myosotis decumbens, Oxyria digyna, Of the northernmost Finnish A. crenatum Roegneria canina, Thalictrum alpinum, Dryopteris localities Kajaani has an oceanity index of 25, phegopteris and Viola biflora ; Dryopteris fr agrans Kuusamo 20 (Kotilainen 1933), and Kevo in the grows in the vicinity. No calcicolous associates oc­ neighbourhood of the Utsjoki locality has an index of cur, and pH is fairly low, 5.3-5.4. 16 (from values fo r the period 1962-1971, published In Norwegian Gudbrandsdalen A. crenatum in SeppaHi 197 6). K vikkjokk village has an index of grows "at the feet of the mountains on shady places 21, and in Gudbrandsdalen the index is 17 in Vinstra­ with mull soils, among rotting fa llen logs and on Solstad, 249 m above m.s.l. and almost in the middle grassy ground in the company of subarctic plants as of the A. crenatum district (the K vikkjokk and Dryopteris phegopteris, Stellaria nemorum, Aco­ Vinstra-Solstad values are calculated for the period nitum, Geranium silvaticum, Ranunculus acris etc." 1931-1960). Climate diagrams for some Nordic A. (Axel Blytt describing the locality at Elstad in crenatum localities are shown in Fig. 3. Ringebu according to Nordhagen 1921 p. 138, in The low oceanity in K vikkjokk is caused by the Dahlskog's translation of the Norwegian original). fact that the unusually deeply glacially-eroded valley By "subarctic" Blytt, according to Nordhagen (1943 of the Lilla Lule River extends to one of the highest p. 39), meant plants distributed over most of Norway areas in the Scandes, which protects the valley from but particularly prominent in the alpine valleys and in the maritime influence of the westerly winds. The Northern Norway and with an over-all European­ conifer-forest border line is about 15 km to the N and Siberian impression. Dahl (1949 p. 13) describes the W of K vikkjokk village. Lake Saggat, at 302 m locality at Ula in Sel, where A. crenatum is abundant above m.s.l., is only 2 7 km S of peaks of 2000 m in fo r a distance of at least 2 km: it luxuriates in tall­ Sarek and only 20 km E of peaks of 1800 m. Ac­ herb vegetation of the Mulgedion alpini -type of cording to the map in Moe 1970 p. 63 (cf. Tallantire N ordhagen, but also thrives in the lower part of scree 1972), the spruce should have invaded the region slopes, associated with Rubus idaeus and Equisetum around Lake Saggat less than 2000 years ago, but the scirpoides, and occurs in conifer woods with present-day associates of the spruce have either had a Hy locomium splendens and Rhytidiadelphus tri­ rapid dispersal or reached the alpine fo reland earlier quetrus. The habitats at Ula and K vikkjokk seem to than the spruce, as the westernmost known localities resemble each other, and A. crenatum seems to dis­ in SW. Lule Lappmark for several easterly species play a rather wide ecological range on both localities. are situated within a few km of K vikkjokk. These

A eta Phytogeogr. Suec. 68 56 Sten Dahlskog species include Potamogeton natans, Eleocharis (Gjrerevoll 1963 p. 281), Kvikkjokk close to the sup­ palustris, Epipogium aphyllum, Salix myrtilloides, posed maritime coastal refugia in Nordland (map see Ranunculus lapponicus, Nuphar luteum x pumilum, Gj rerevoll 1963 p. 268), and Utsjoki close to the sup­ Nymphaea tetragona and Ledum palustre, and the posed continental coastal-tundra refugia in Finnmark easterly shrew Sorex minutissimus (Dahlskog & (Eilif Dahl 194 7 p. 240). Dahlskog 1970 and 1980; Dahlskog 1980; Dahlskog Selander (1947 p. 274) considers the Gud­ & Johansson 1979). brandsdalen localities for A. crenatum as evidence of the ability of the forest florato survive the glacial age in refugia, but he considers Cystopteris sudetica and Clematis sibirica to be either glacial survivors or Dispersal theories possibly post-glacial immigrants, and in 1950 pp. Many theories have been proposed to explain the 166-167 he considers the later theory to be the most Norwegian localities of Athyrium crenatum, probable for all three species. Minayev ( 1968: Clematis sibirica, Cystopteris sudetica and the other English translation) writes that the South Norwegian "huldra plants". They are probably a heterogeneous localities of the three species "are to be considered as group and may have been dispersed in different ways. relict, of basically late Pliocene origin, although ap­ As other markedly disjunct distributions, they parently dislocated in relation to the original ones" may be explained either by supposing that (I) the (p. 68). But he considers A. crenatum to have invaded plant formerly had a continuous distribution but Finland during Aller0d and states that A. crenatum became extinct in parts so that the presently isolated belongs to the species which partly reached the localities are relicts of a fo rmer larger area, or by sup­ mountains in the Northern Scandinavian Peninsula posing that the plant has reached the isolated during the end of the Gotiglacial (pp. 76-77). But as (11) localities by long dispersal from its original home his map on p. 50 does not show the Utsjoki locality, without ever having occurred in the intervening area. the text is probably best interpreted as a suggestion of The long-dispersal theory is all too easily used as a a post-gla cial immigration along the Scandes to the deus ex machina. Its weakest point is not the disper­ South Norwegian localities, although they were sal over long distances, it is the subsequent germina­ stated (p. 68) to be pre-glacial relicts. tion and establishment, which in closed vegetation may be supposed to be extremely rare (Cain 1944 p. The localities of A. crenatum in the Scandes are 285; Wulff (1943) 1950 p. 134). Simpson writes probably best not interpreted as indications of the (1952 p. 174): "any event that is not absolutely im­ fern having survived the ice age in the supposed adja­ possible . . . becomes probable if enough time cent refugia. It seems very improbable that any pos­ elapses"-but in Norden a very short time has elapsed sible refugia in Finnmark could harbour birch since the ice age. fo rest and sub-alpine species, and it seems far more probable that the Utsjoki locality has been colonized by dispersal from Kuusamo. Any South-Norwegian Relict theories nunataks must also be considered as very unlikely Glacial survivor? The discussion about possible ice­ habitats for sub-alpine species, and if A. crenatum did free refugia in glacial-age Norway and which species survive the ice age in Southern Norway the supposed might have been able to survive the latest glaciation coastal refugia in More seem far more likely. Refugia there (see LOve & Love 1963; Ragnar Dahl 1963) in More and in Nordland to the W of Kvikkjokk has only rarely involved the huldra plants, which might have harboured birch fo rest, but their climate presumably have been considered too pronounced is considered to have been maritime (Eilif Dahl 194 7 lowland conifer-forest plants to be conceivable glacial p. 240; Selander 1950 pp. 162-1 65) and would hard­ survivors. But Athyrium crenatum does grow in the ly have suited the continental Athyrium crenatum. It subalpine belt in Utsjoki, Kandalaksha and the Urals, seems impossible that isolated A. crenatum popu­ and its presently known distribution in Norden might lations would have survived the ice age in maritime be interpreted as a group of young localities in refugia and yet have kept such an ecological adap­ Eastern Finland and Far Carelia, the result, perhaps, tability that they would have been able to disperse of post-glacial dispersal still continuing, and three old into valleys with continental climates: if that had been relict localities with a bicentric distribution in the the case, the fern ought to be rather common and to Scandes. The alpine localities are all situated close to have remained in some maritime areas. The present­ supposed refugia: Gudbrandsdalen close to the sup­ day distribution of Athyrium crenatum is not logical­ posed inland refugia in the Jotunheimen-Dovre dis­ ly explained by considering the fern a glacial sur­ trict, which would have had a continental climate vivor, and until the existence of ice-free refugia is

A eta Phytogeogr. Suec. 68 The Siberian fe rn Athyrium crenatum fo und at Kvikkjokk 57

definitely proven, other theories ought to be I have not visited any other A. crenatum locality preferred. than the one at K vikkjokk, which I cannot consider as particularly humid and where A. crenatum occurs Post-glacial immigrant. Relict of an early con­ in both rather dry and somewhat more humid areas. (1) tinental period? The huldra plants may have im­ The same seems to be the case at River Ula in Gud­ migrated post-glacially and have had a continuous brandsdalen according to Eilif Dahl 1949, and distribution from Eastern Finland to Southern Kallio, Laine & Makinen ( 1969 p. 82) write about the Norway. This theory was first proposed by Axel Utsjoki locality that "occurrence in a boulder fieldin Blytt, who was no believer in long-dispersal but did a deep valley may indicate continental features". The believe in slow, step-by-step immigration (Nordhagen ecological preferences of the huldra plants seem to be 1943 p. 61). Blytt considered Athyrium crenatum to very incompletely known, and field studies of the be one of those Siberian species whose scattered local climate in sites with luxuriant stands as well as Norwegian localities he considered relicts of a cultivation experiments are probably needed before fo rmer, more continuous distribution. In this group any firm conclusions about most probable immigra­ he included, among others, Carex rhynchophysa, tion periods may be stated. Cinna latifolia, Cypripedium calceolus, Ga/ium trifidum, Pedicularis sceptrum - carolinum, Salix myrtilloides and Veronica longifolia. "Vielleicht sind Long-dispersal ...wenigst ens einige von ihnen ...wahrend der sub­ The validity of the relict theories for some of the arktischen Zeit eingewandert. Die tiefste huldra plants has been criticized by, among others, Wurzelschicht in unseren Mooren beweist, dass es Marklund (1948 p. 9), who fm ds it absurd that auch zwischen der arktischen und borealen Zeit eine Clematis sibirica should have become extinct in the Periode mit ausgesprochenem Kontinentalklima gab. whole area between Gudbrandsdalen and the one In dieser Zeit miissen viele aus Sibirien stammende locality in Osterbotten in W. Finland and everywhere nordische Arten nach Skandinavien eingewandert between Osterbotten and Far Carelia as the species is sein und sie miissen unter den spateren insularen adapted to closed forests. The same objection may be Perioden seltener geworden sein" (1893 pp. 29-30). raised towards the interpretation of A thyrium Blytt's theory has been supported by, among crenatum as a relict; it is difficult to explain why it others, Nordhagen (1943 p. 66, 1955 pp. 3-4), who should be able to remain in luxuriant stands only in considered it valid also for Cystopteris sudetica and Gudbrandsdalen, in K vikkjokk and in Utsjoki. The Clematis sibirica, whose Norwegian localities were relict theory offers no complete explanation, and if discovered after Blytt's time. long-dispersal is too oftenus ed as a deus ex machina, The localities of A. crenatum in Utsjoki and so is perhaps the convenient extinction in wide areas K vikkjokk are on the most probable immigration taken too much for granted by the relict theories. At path from Kuusamo to Gudbrandsdalen, and the least at Ula and in K vikkjokk, A thyrium crenatum climate of both localities shows continental features seems to exhibit an ecological range which seems too (although less pronounced than in the Gud­ wide fo r a relict species, as does Clematis sibirica in brandsdalen). The two fm ds may be said to support Gudbrandsdalen, according to Hjelmstad (1978 pp. Blytt's theory, but may be interpreted also in other 172-173). ways. Long-dispersal by wind. Wind dispersal has been Post-glacial immigrant. Relict of a more humid (2) offered as an explanation of several other isolated period? The huldra plants are often considered to be fern localities. continental. Lye (1967 pp. 104-1 06) assigns Atizyrium crenatum, Aster sibiricus and Cystopteris Thus, for the Atlantic Polystichum lobatum at Pielavesi in sudetica to "a rain-shadow subgroup" in his "strong­ Northern Savolaks in Finland (Kotilainen 1933 p. 66), ly continental (southeastern) element". But Gj rerevoll about 700-800 km distant from the easternmost localities (1974 pp. 91-92) points out that many of them on Gotland and in Dalsland in Sweden, for Asplenium prefer shady habitats with high air humidity and con­ adiantum-nigrum and Phyllitis scolopendrium in Halland 300-90 km from the nearest localities (Skottsberg 1936 siders them to be later immigrants than considered by pp. 238-239), for the only localities of the calcifuge Blytt. Hjelmstad (1978 p. 174) writes that Cysto­ Asplenium septentrionale on Gotland (Bengt Pettersson pteris sudetica and A thyrium crenatum grow in 1958 pp. 70-7 1), which are on erratic boulders, and for ha itats with a highly humid microclimate and that the only Nordic locality of Ceterach offi cinarum (Nyhlen onBy a period with a more humid climate can have 1949 p. 402) on Gotland about 500 km N of the nearest allowed a more continuous distribution. locality in the vicinity of Gdansk. Marklund suggests

Acta Phytogeogr. Suec. 68 58 Sten Dahlskog

( 1948 p. 9) that A thyrium crenatum and Cystopteris to be right in suggesting that Clematis had been sudetica have been long-dispersed to Gudbrandsdalen by epizooically dispersed to Norway by the hairy seeds winds. having caught in the feathers of invading Siberian Wind dispersal of cryptogams to Southern Finland was birds, and believes the same explanation to be valid investigated by Bror Pettersson, who found abundant for the Lappfjard locality. evidence of long-dispersal from the centre of the distribu­ tion area of A. crenatum. In a rainfall from a Siberian­ Bird dispersal is even more probable for the small Eastern European air mass he found spores of the mosses fern spores than fo r the seeds of Clematis, which as Aloina brevirostris and A. rigida (both new for Finland; Hj elmstad points out (1978 p. 173) are not a tempt­ determined after germination) in such quantities that the ing food and probably soon would be preened from fall-out was equivalent to 15000 106 spores/km\ and the the feathers. x time of the year made a long-dispersal of at least 2000 km from Siberia most probable (1940 pp. 85-87). Similar weather situations are far from rare, and Pettersson's Long-dispersal by human agency. Clematis sibirica is measurements therefore show not only that long-dispersal the huldra plant for which dispersal by human agen­ is common but, especially, that it is very ineffective. (The cy seems most probable, in spite of the fact that most two Aloina species are stated by Nyholm 1956 pp. Norwegian localities and the one West Finnish are 100-101 to occur in Sweden, Norway, Denmark and not situated close to settlements. Grondahl (1975 p. Central Asia or Siberia, but not in Finland.) 7) considers human dispersal to be a too imaginative theory, but he does mention that the Lappfjard locali­ According to Bror Pettersson (1940 p. 88) dispersal ty is criss-crossed by old paths and that Russian of fern spores over distances of 300-400 km would soldiers bivouaced on the habitat. Hjelmstad (1978 p. seem to be far from rare. Only one undiscovered (or 174) considers Clematis to have been dispersed by by now extinct) A. crenatum locality between Utsjoki man to Gudbrandsdalen during the Viking Age and and Kvikkjokk (465 km) and just two between to now be in full expansion. But why has Clematis K vikkjokk and Gudbrandsdalen (700 km) are needed not been fo und within the Svithiod of former days, to make wind dispersal a fully feasible explanation. the part of Sweden which had so very much closer Furthermore, both wind dispersal and successful es­ trade with its colonies in the Russian Gardarike than tablishment are more probable occurrences for Gudbrandsdalen ever had? A thyrium crenatum than for most forest species, as Dispersal by human agency of Athyrium crenatum A. crenatum thrives on scree slopes, where rising air and Cystopteris sudetica seems extremely improb­ currents, pioneer micro-sites and niches with weak able; their localities are too inaccessible. competition are more common than in other, more closed and stable fo rest habitats. SUMMARY Long-dispersal by birds. When Enge and Ourom in 1915 fo und wild Clematis sibirica in three localities The discovery of the K vikkjokk A thyrium crenatum in Gudbrandsdalen (Wille 191 7), the Norwegian flora locality gives no decisive support for any of the dis­ was enriched by a new species with the same distribu­ persal theories, because it may be interpreted in tion as A thyrium crenatum. Wille considered it out of several ways. Most of the theories may be valid fo r the question that the Clematis localities should be one huldra plant or the other; the least scattered relicts of a fo rmer area continuous with Eastern ( Carex rhynchophysa, Cinna latifolia, Glyceria Finland and Russia; there could be no doubt that lithuanica) may be relicts which have had a larger Clematis had been long-dispersed from Northern post-glacial distribution, Clematis sibirica may have Russia or Siberia in a rather late period, probably by been dispersed by birds to Gudbrandsdalen and by birds, because its behaviour in the three Gud­ Russian soldiers to W. Finland, Athyrium crenatum brandsdalen localities so obviously showed that it may be a post-glacial relict in Utsjoki and have been was extending its footholds (1917 pp. 252-254). In dispersed by wind from Utsjoki (or Kuusamo) to 1953 ten Gudbrandsdalen localities were known, and Kvikkjokk and Gudbrandsdalen (or to Kvikkjokk in 1977 the number had increased to 35 (Hjelmstad from Gudbrandsdalen). 1978, map p. 172). "But the interpretation ... must of course be sub­ In 194 7 Grondahl fo und a large stand of Clematis jective" ("Men opfatningen ... maa selvf0lgelig bli sibirica on the W. coast of Finland, in Lappfjard in subjektiv"), as Nordhagen (1921 p. 138) ends his dis­ Osterbotten, 700 km from the nearest locality in Far cussion of the immigration of A thyrium crenatum to Carelia (Marklund 1948 p. 5), Grondahl (1975 p. 7) Norway. suggests endozoic dispersal by birds having eaten I believe that Athyrium crenatum had better con­ Clematis seeds. Marklund (1948 p. 10) believes Wille ditions during a more continental period and that the

A eta Phytogeogr. Suec. 68 The Siberian fe rn Athyrium crenatum fo und at Kvikkjokk 59

species then managed to occupy some localities in the - 1949. En masseforekomst av russeburkne (A thyrium Scandes through long-dispersal by wind. From these crenatum). - Blyttia 7:13. firstloca lities it may have dispersed to others without Dahl, R. 1963. Shifting ice culmination, alternating ice managing to extend its continuous distribution area covering and ambulant refuge organisms? - Geogr. Ann. 45:122-138. to the Scandes. A more maritime climate and in­ Dahlskog, B. & Dahlskog, S. 1970. Vaxtfynd i Kvikk­ creasing competition in the more stable habitats may jokktrakten, tillagg till Selanders Karlvaxtfloran i have caused it to disappear from several localities in sydvastra Lule Lappmark. - Viixtbiologiska in­ the Scandes. The present alpine Scandinavian stitutionen, Uppsala. (Mimeographed). 24 pp. localities would be relicts of former long-dispersal, 1980. V axtfynd i SW Lule Lappmark. - Svensk bot. and several may remain to be discovered. The one Tidskr. 74:265-276. conclusion which is confirmed by the K vikkjokk Dahlskog, S. 1972. The Kvikkjokk Area.- UNGI Rap­ locality is that our knowledge of the florain the alpine port 13:7-13. fo reland is far less extensive than we would like to - 1980. Skogsfrun, Epipogium aphyllum, i Nordnorge believe. och Nordsverige. - Svensk bot. Tidskr. 74:145-152. Dahlskog, S. & Johansson, R. 1979. Mindre dvargnabb­ mus, Sorex minutissimus Zimm., funnen i K vikkjokk, Acknowledgements Lilla Lule iilvdal. - Fauna Flora 74:232-233. Flora Europaea. 1. 1964. - Cambridge. 464 pp. Thanks are due to the C.F. Lundstrom Foundation for Flora of the U.S.S.R. 1. 1968. -Jerusalem. 244 pp. financial support, to Bertil Nordenstam of the Swedish Gj rerevoll, 0. 1963. Survival of plants on nunataks in Museum of Natural History in Stockholm for determina­ Norway during the Pleistocene glaciation. - North tion, to Rolf Y. Berg in Oslo and Rolv Hjelmstad in Atlantic biota. Norwich. pp. 26 1-283. Trondheim for informative letters, to Det Norske 1973. Plantegeografi. - Trondheim. 186 pp. Meteorologiske Institutt for climate data from Gud­ 1974. Fra floraen i Gudbrandsdalen. - Bygd og by i brandsdalen, to Agneta Nordgren and Folke Hellstrom fo r Norge. Gudbrandsdalen. pp. 85-95. help with the figures,ki itoksin paljon Nunne Lundahl, and Grondahl, G. 1975. Sipprankan - en rymling fdin for good companionship to the participants in the K vikk­ Ostkarelen. - Fin!. Natur 1975:5-8. jokk 1979 floristical course. Hartman, C. 1879. Handbok i Skandinaviens flora, uppl. 1. - Stockholm. 616 pp. Hjelmstad, R. 1978. Utbredelsen av skogranke (Clematis REFERENCES sibirica) i Norge. - Blyttia 36:171-175. Hjelt, H. 1888. Conspectus Florae Fennicae 1. - Acta Auer, A.V. 1944. Kasvistollisia havaintoja Pohjois­ Soc. Fauna Flora fenn. 5: 1-562. Suomesta 3. - Memo. Soc. Fauna Flora fennica Hulten, E. 1971. Atlas of the distribution of vascular 19:44-57. plants in Northwestern Europe. 2nd ed. - Stockholm. Backman, A.L. 1914. Floristiska meddelanden. - Meddn 53 1 pp. Soc. Fauna Flora fennica 40:119-121. Hylander, N. 1953. Nordisk Karlvaxtflora. 1. - Uppsala. Blytt, A. 1893. Zur Geschichte der nordeuropaischen, 392 pp. besonders der norwegischen Flora. - Bot. Jb. lgoshina, K.N. 1969. Flora of the mountain and plain tun­ 17:1-30. dras and open fo rests ofthe Urals. - Vascular plants of Blytt, M.N. 1861. Norges Flora. 1. - Christiania. 386 pp. the Siberian North and the northern Far East. - Br andt, A. 1933. Hiisjarven luonnonpuiston kasvilli­ Jerusalem. pp. 182-334. suudesta. - Silva fenn. 32:1-112. Kaakinen, E. 1972. Studies on herb-rich forest vegetation Brenner, M. 1899. Observationer rorande den nordfinska in southern Kainuu, northern Finland. - Aquilo, Ser. floran. - Acta Soc. Fauna Flora fenn. 16:1-307. Bot., 11:23- 42. Br0gger, A. 1960. Morfologiske og cytologiske un­ Kallio, P., Laine, U. & Makinen, Y. 1969. Vascular flora ders0kelser av noen norske bregner. - Blyttia of lnari Lapland. 1. Introduction and Lycopodiaceae ­ 18:33-48. Polypodiaceae. - Rep. Kevo Subarctic Res. Stat. 1964. Huldreplanter - et plantegeografisk mysterium. 5:1-108. - Naturens Verden 1964:183-1 86. Kallio, P. & Nikoskelainen, R. 1954. Kasvistollisia Cain, S.A. 1944. Foundations of plant geography. - New havaintoja Pohjois-Karjalasta. - Archvm Soc. Zoo!. York. 556 pp. Bot. fe nn. Vanamo 9:135-145. Cluist, H. 1910. Die Geographie der Farne. - Jena. 358 Kalliola, R. 1958. Athyrium crenatum (Somf.) Rupr. - pp. Kolmiomainen hiirenporras. - Suuri Kasvikirja. 1. - Chlristensen, C. 1924. Plantae Sinenses a dre. H. Smith an­ Helsingfors p. 88. nis 1921-22 lectae. 3. Pteridophyta. - Acta Horti Kotilainen, M.J. 1933. Zur Frage der Verbreitung des gothob. 1:41-1 10. atlantischen Florenelementes Fennoskandias. - Annls Dahl, E. 194 7. Litt om fo rholdene under og etter den siste Bot. Soc. Zoo!. Bot. fenn. Vanamo 4( 1): 1-75. istid i Norge. - Naturen, 71:232-252. Laine, U., Lindgren, L. & Makinen, Y. 1955. Havaintoja

Acta Phytogeogr. Suec. 68 60 Sten Dahlskog

Utsjoen piUijan lansiosan kasvistosta. - Archvm Soc. Nordfinnland. - Acta Soc. Fauna Flora fe nn. Zoo!. Bot. fenn. Vanamo 9:120-135. 72( 18): 1-25. Lid, J. 1963. Norsk og svensk flora. - Oslo. 800 pp. 1955b. Das Autreten der Waldpflanzenauf verschieden Linkola, K. 1921. Studien i.iberden Einflussder K ultur auf kalkhaltiger Unterlage in Ladoga- und Grenz-Karelien. die Flora in den Gegenden nordlich vom Ladogasee. - - Archvm Soc. Zoo!. Bot. fe nn. Vanamo 9 (suppl.): Acta Soc. Fauna Flora fenn. 45(2): 1-49 1. 235-245. Lye, K. 1967. En ny inndeling av Norges plante­ Pettersson, Bengt 1958. Dynamik och konstans i Gotlands geografiske element. - Blyttia 25 :88-123. flora och vegetation. - Acta phytogeogr. suec. Love, A. & Love, D. 1963. North Atlantic biota and their 40: 1-288. history. - Norwich. 430 pp. Pettersson, Bror 1940. Experimentelle Untersuchungen Marklund, G. 1948. Clematis alpina ssp. sibirica funnen i i.iber die euanemochore Verbreitung der Sporen­ Finland. - Memo. Soc. Fauna Flora fennica 24:5-10. pflanzen. - Acta bot. fe nn. 25: 1-1 02. Minayev, N.A. 1968. Siberian taiga elements in the floraof Rosendahl, H. V. 1913. Bid rag till Sveriges ormbunksflora the Northwest of the European part of the USSR. - 2. - Svensk bot. Tidskr. 7:276-297. Distribution of the floraof the USSR. - Jerusalem, pp. Selander, S. 1947. Urtica gracilis Ait. in Fennoscandia. ­ 44-83. Svensk bot. Tidskr. 41 :264-282. Moe, D. 1970. The post-glacial immigration of Picea abies - 1950. Floristic phytogeography of south-western Lule into Fennoscandia. - Bot. Notiser 123 :6 1 -66. Lappmark. - Acta phytogeogr. suec. 2 7: 1-200. Nakaike, T. 1975. Enumeratio Pteridophytarum Seppala, M. 1976. Periglacial character of the climate of Japonicarum. - Tokyo. 375 pp. the Kevo region. - Rep. Kevo Subarctic Res. Stat. Nordhagen, R. 1921. Kalktufstudier i Gudbrandsdalen. ­ 13:1-1 1. Skr. Vidensk. Selsk. Christiania, I. Mat.-nat. Kl. Simpson, G.G. 1952. Probabilities of dispersal in geologic 9:1-155. time. - Bull. Am. Mus. Nat. Hist. 99:163-176. 1943. Axe! Blytt, en norsk og internasjonal forsker­ Skottsberg, C. 1936. Asplenium adiantum-nigrum L. i profil (1843-1 898). - Blyttia 1:21-83. Halland. - Acta Horti gothob. 11:233-240. 1955. Ett bidrag till differensialdiagnosen mellom Sommerfe lt, S.C. 1835a. Aspidium crenatum, en ny art af Cystopteris sudetica Al. Br. & Milde og C. montana ormbunke, upptackt och beskrifven. - K. [ svenska] (Lam.) Bernh. - Acta Soc. Fauna Flora fenn. 72, Vetensk. Akad. Hand!. 1834:103-104. (17): 1-8. 1835b. Aspidium crenatum, en nye Bregneart. - Norrlin, J.P. 1870. Bidrag till sydostra Tavastlands flora. Magazin Naturv. Christiania 12:80-8 1. - Notis. Sallsk. Fauna Fl. fennica, Ny ser. 8:73-1 96. Sonck, C.E. 1964. Die Gefasspflanzenflora von Pielisjarvi - 1878. Symbolae ad floram Ladogensi-Karelicam. - und Lieksa, Nordkarelien. - Acta bot. fenn. 67:1-3 10. Meddn Soc. Fauna Flora fennica 2:1-33. Svenonius, F. 1900. Ofversikt af Stora Sjofallets och Nyhlen, G. 1949. Ceterach off icinarum, en for Skandina­ angransande fj alltrakters geologi. 2. Berggrunden. - vien ny ormbunke. - Bot. Notiser 1949:395-402. Geol. For. Stockh. Forh. 22:273-323. Nyholm, E. 1956. Illustrated moss floraof Fennoscandia. Tallantire, P.A. 1972. The regional spread of spruce (Picea 2:2. - Malmo. 108 pp. abies (L) Karst.) within Fennoscandia: a reassessment. Nyman, E. 1895. Vegetationsbilder frfm Lappland. - Bot. - Norw. J. Bot. 19:1-16. Notiser 1895:1-12. Valmari, J. 1921. Beitrage zur chemischen Bodenanalyse. Pankakoski, A. 1939. Ekologis-kasvistollisia tutkimuksia - Acta fo r. fe nn. 20:1-6 7. Hiisjarven luonnonpuistossa. - Annls Bot. Soc. Zoo!. Walther, H. & Lieth, H. 1967. Klimadiagramm-Weltatlas. Bot. fenn. Vanamo 10(3):1-154. - Jena. Pesola, V .A. 1929. Kalsiumkarbonaati kasvimaan­ Wassen, G. 1966. Gardiken. Vegetation und Flora eines tieteelisena tekijana Suomessa. - Annls Soc. Zoo!. Bot. lapplandischen Seeufers. - K. svenska Vetensk. Akad. fe nn. Vanamo 9( 1):1-246. Avh. Naturskydd. 22:1-142. 1934. Die Waldvegetation feuchten Gelandes in N­ Wille, N. 1917. Atragene sibirica L. vildtvoxende i Norge. Kuusamo und SE-Kuolajarvi. - Acta for. fenn. - Bot. Notiser 1917:24 1-255. 40: 1-14. WulfT, E.V. (I 943) 1950. An introduction to historical 1955a. Uber die Vegetation der Moore und feuchten plant geography. - Waltham, Mass. 223 pp. Wiesen im Norden Kuusamos und Si.idwesten Sallas in

Sten Dahlskog, Institute of Ecological Botany, Uppsala University, Box 559, S-751 22 Uppsala, Sweden

Acta Phytogeogr. Suec. 68 The Downward Migration of Plants on a Rising Bothnian Sea-shore Lars Ericson

In spite of the overall importance of land uplift as a The amplitude for 99 % of the sea-level fluctuations vegetational factor in the Baltic area, only a few in­ amounts to 120- 130 cm (Lisitzin 1960). During 196 7-79 vestigations presenting documented changes over a given the sea-level amplitude was 128 cm in June-August and time scale have been published. Most relevant are (not con­ 226 cm in January-December. Large seasonal and annual si ering paleoecological investigations and map studies) variations occur, e.g., a cyclic variation between low-water the accurate long-term investigations published by Luther and high-water years (Fig. 2). (1961), Schwanck (1974) and Vuoristo & Rousi (1976). The "relative land uplift" (in the fo llowing "land uplift") However, our knowledge of the plant colonization on the i.e., the change of land surface relative to the theoretical en:ergent shores, and the role played by land uplift and mean sea-level, amounts to 8.5 mm/year (RAK 1971). other factors, is only fragmentary. In 1968, I analysed a number of permanent belt transects on the Bothnian coast. The prime aim was to in­ Descrip tion of the transect ve tigate the mode of plant colonization on the emergent The shore vegetation reflects both a marked salinity shores. Interest was focused upon whether this coloniza­ gradient and a development towards more sheltered tion occurs in a uniform way from year to year, i.e., that (less wave-washed) conditions. The salinity ampli­ land uplift is the main operative factor, or whether the %o colonization shows large annual differences, i.e., that other tude of the off-shore water is 2.8-3.9 (June­ fa ctors are operative. August 1971; cf. below with the salinity values for The aim of this paper, which deals with one single belt the top-soil during the same period). transect, is to present the downward migration of vascular The vegetational gradient in 1968 can be plants for the period 1968-79 and to compare it with the characterized as fo llows (cf. Figs. 3, 4): Chara aspera water-level fluctuations and land uplift during the period. - Potamogeton filiformis - Scirpus tabernaemontani However, neither the aquatic nor the terrestrial species (in (1.3-3.4%o), Eleocharis uniglumis (with Deschamp- the alder fr inge) are treated in any detail. The results and conclusions must not be considered as valid for other shore types, unless exclusively pointed out.

Study area The investigated belt transect is situated on the shore of a large, rather sheltered, bay (Sorfjarden) on the Gulf of Bothnia (the Oston peninsula, in Savar parish, Vasterbotten province, long. 63°48'N, lat. 17°4 1 'E), situated on the W side of the N. Kvarken, the shallow threshold between the Bothnian Sea, to the S, and the Bothnian Bay, to the N (Fig. 1). �0

Genera/fea tures The mean annual temperature is 3-4°C (Atlas over Sverige 32). The growth period (definition: daily mean air temperature > 6°C) is about 150 days and begins in mid­ May and ends in early October (Johannessen 1970); a fair­ ly accurate figure with regard to the vegetation here. The mean annual precipitation is 450-500 mm (Atlas over _ _ Sverige 31). Ice formation generally begins in late 0..__ ...... __ __.10 KM November and break-up occurs in early May (Angstrom /) 1974). Fig. 1. Location of the study area (arrow).

A eta Phytogeogr. Suec. 68 62 Lars Ericson

cm 72 73 74 75 76 77 34 ?_fi____J,9 � r

� • •

•

... � • I I I 31 I . t-, J- I ··-- \. . . ,' 300 ':•·t··· : ·· ···e·· ·· . I ·

I I 'V�, 29 / •11 � \....!--

I DEC I I 28 ' I JUN AUG I I • I I .6.- --.A.SEP OC T I I 11 ...... NOV DEC 270 �

Fig. 2. Variations in mean sea-level at the Ratan mareograph station during the period 1967- 1979 (for location see Fig. 1 ). N ate the annual and seasonal Fig. 3. The middle and upper part of the investigated sea­ variations. The sloping line denotes the change of the shore. Note the dense litter layer. V egetational gradient: theoretical mean sea-level. Primary data supplied by the Scirpus tabernaemontani (upper limit immediately to the Swedish Meteorological and Hydrological Institute left of the photo)-Eleocharis uniglumis (in flower) (SM H I). -Agrostis gigantea (vegetative)-Calamagrostis neglecta (with last year's floral shoots indicating the upper ice-limit sia bottnica and Agrostis stolonifera var. bottnica, during the preceding winter) - Carex nigra - Molinia coerulea . The narrow Alnus incana -border, with Myrica 0.9-3.1 %o) - Calamagrostis neglecta (with Juncus gale in its lower part, is succeeded by a Picea abies forest gerardi, %o) Carex nigra 0.2-1.6 - (in the lower part ( Vaccinium -Myrtillus type), reached by extreme high­ with Carex pulchella depressions, 0. 1-1.0 %o; in the waters. Water-level = 304 cm. 1971-07-06. upper part with Molinia coerulea, 0.0-0.5 %o) - Filipendula ulmaria - Rubus arcticus (0.0-0.4%o , together with shrubs of My rica gale and A lnus in­ However, among other characteristics the presence cana; reached by summer high-waters) - Agrostis of Scirpus tabernaemontani and Rubus arcticus, the tenuis (with a tree layer of Alnus incana; irregularly well-developed Eleocharis uniglumis belt, the low reached by high waters in late autumn-early winter). level of Agrostis gigantea, the minor importance of The large salinity amplitude evident fo r the belts Juncus gerardi and Filipendula ulmaria, and the situated lower and the lower values for the upper belts dense litter layer, are typical of the inner skerry zone. are due to freshwater-influence following both Due to the sheltered position, the litter-layer from the surface-water drainage and ground-water outflow middle part of the Eleocharis uniglumis belt (31.6 m from the hinterland. Due to this outflow, species such in 1968; 38 m in 1979) and upwards is never washed as Eleocharis acicularis, Ranunculus reptans, off during autumnal storms (Figs. 3, 4). Crassula aquatica and Subularia aquatica (Figs. The gradual change towards more sheltered con­ 5F-G, 6A-B) are able to colonize during low-water ditions is also revealed by the substrate: above 11 m a periods in early summer. wave-washed moraine, below 11 m a gravelly coarse The main vegetational gradient shows similarities sand (stony down to 36 m) and below 63 m a clayey with wave-washed, freshwater-influenced, shores sand. Due to the sheltered position, accumulation of situated on fo rested islands in the outer skerry zone. shore drift is of rather insignificant importance.

Fig. 4. Belt transect diagram showing the vegetational gradient in 1968 (filled bars) and 1979 (unfilledbars). Overlapping...,.._ areas are dotted. For species absent in 1968 the dates given in brackets denote the year for colonization, and for those ab­ sent in 1979, the year of disappearance. 1st-year seedlings of perennial species are excluded. Deschampsia bottnica in­ cludes D. bollnica x caespitosa (cf. Fig. 5H). Ground level in 1968 (solid line), in 1979 (dotted line).

A eta Phytogeogr. Suec. 68 >SO •90 TREE LAYER 1 •25 7 Picea abies •1 ,, • 2 .,_ •0.1 o0.1 GOV�fl "' Alnus incana . . ·.·.·.·..---Il...RTI .. . . I J W.:J I SHRUB LAYER Sorbus aucuparia ALONGSIDE * 1"1 Picea abies � 1979 ** • Alnus incana ��-� - Salix pentandra (1973)

Myr ica gale • Betuld pubescens (juv.) � Lnn·.·.�-­ Sal ix phylicifolia FIELD LAYER Rubus idaeus (1972) =�""' • Me 1ampyrum pratense (1978) --- Moehringia trinervia (1971) Vacc inium myrti1lus Luzula pi 1osa e•• Agrostis tenuis - !'.. !'.. - Potentilla erecta �.&11& Rubus arcticus !'.. o � Oxycoccus quadripetalus (L973) � F Trientalis europaea = � Sagina nodosa (1973) . Carex canescens (1977) Epilobium pa1ustre ..._ Potenti11a eged ii !'..!'..!'..!'.. . Filipendula u1maria ... !'..!'.. !'..<::>!'.. • ···--�ii!!l!br:.•••- Festuca rubra � Viola pa 1ustris . �---·· f'..=!'..f'../\1\1\J\M'W " Juncus filiformis (1976) Potentil1a pa 1ustris Rhinanthus angustifo1ius Mo linia coerulea Leontodon autumna 1is AAJ"'Q n t Euphrasia bo ttnica (1970) !'.. « • •• Carex pu1chel1a �t Ophioglossum vulgatum (1973) --·-- Al\1\�---1\ Parnassia pa1ustris M N Ga l ium pa 1ustre 1.1. ··- • .. ... Pedicularis palustris I I -...... !'..!'..!'..!'..!'.. !'.. Carex nigra var . recta d2 .-...A�� .....· ...... U .. ..· . Triglochin palustre !'.. /\{\/\�000 ==� Wi.� Agrostis stolonife ra M:::::ll�LI.!'..� .... AAA ...... Juncus gerardi ..... t::.IJc. AA !'.. r:::::ooo • •• Carex mackenziei • Agrostis gigantea o o • • • · ��oooo/\1' 7••• .. • • • • • • • • • • • • t Juncus alpinus (1976) 1\1\J\J. • • ...... - MAC>,..,,' cm Calamagrostis neg lecta ···· . . · ':'· I · · · (\..o...o...o. • 400 Ra nunculus reptans (1973) - Ag rostis stol . v. bo ttnica ------····· Deschampsia bo ttnica � !'.. !'..• • !'.. • • AA r, Triglochin ma ritimum • • !'..!'..!'..!'..!'..!'..!'..• •88�Aoo- •!,Slo o eo Ei Eleocharis unig1umis o o o o o o !'.. !'.. !'.. !'.. !'.. !'.. !'.. !'.. = !'.. !'.. !'..� !'.. !'.. !'.. !'..� · · · · · · · AA000e Eleocharis pa lustris (1971) .$$. '"tl AA& - c.t Scirpus tabernaemontani => • • • •A<::::::�� A !'.. !'.. !'.. !'.. !'..= A=LI•A•A•A•A«Liooooo +'"'• Subu1aria aquatica (1970) o o 320 Potamogeton filiformis !'.. o o !'.. o . . !'..!'.. !'..!'.. -• 6 ••••___ _. ______:"' Chara aspera . ·.·.· . · •-• AA Potamogeton [:::;::::::::: ...! ...! ...... � _ . . . . Leaf litter (trees and V) per foliatus ...... uL u u :;::: ...... Cl> . ... . • ...... 2{:�:::7:>: . ;; :; :;:; :7:::; .. . ;; ; : � . ; .4 ...... __...... '--· yer Uer ;; : ,_...... : _...... : :::;:;:: . . { . : ; . :: : : : :;::: . .. . : 0\ 79 70 60 ·��=��50 � �4 0 30 20 10 0 00 J:::m 64 Lars Ericson

cm ELEOCHARIS E UNIGLUMIS TRIGLOCHIN 320 CHARA POTAMOGETON C SCIRPUS D A ASPERA B FILI FORMIS TA BERNA:MONTANI MARITIMUM

. ' ' X

300 8901 23456789 jj 280

"ll 8901234 56789 890123456789 89 01 23 456789 8 9 0 1 2 J 4 5 6 7 8 9

AG ROST IS J STO LONIFERA DESCHAMPSIA I AGROSTIS H BOT TNICA (68- 751 STOLONIFERA ELEOCHARIS RANUNCULUS VAR. 80TTNICA F ACICULARIS G REPTANS D. BOTTNICA

X C�SPITOSA (77-79)

I I . HI. fiI ______I ELEOCHARIS PA LUSTRIS I I I q 9 0 , 2 3 4 5 6 1 8 9

8 9 0 2 J 4 5 6 7 8 9 8901234567 8 9 890 123456789 890 1 23456 789 1

N CAREX NIGRA MOLINIA CALAMAGROSTIS AGROSTIS JUNCUS M 0 COERULEA K NEGLECTA L GIG ANTE A GERARDI

II

illllllltl

t 'I·

8901 23456789 8 9 0 1 2 3 4 5 6 7 8 9 8 9 0 1 2 3 4 5 6 7 8 9 09 !H1HIH Fig. 5

A eta Phytogeogr. Suec. 68 The downward migration of plants on a rising Bothnian sea-shore 65

cm

CUS TRIGLOCHIN 360 A B JUN C D GALIUM E POTENTILLA ALPINUS PA LUSTRE PALUSTRE PALUSl RIS

340 CARE X GLAREOSA

- I . 320

CARE X MACKENZIE!

300 8901 23456789

8901 23456789

890123456 7 89 890 123456789 J FESTUCA 360 RUBRA PA RNASSIA F OPHIOGLOSSUM G LEONTODON VIOLA VU LGATUM PA LUSTRIS H AU TUMNALIS I PA LUSTRIS

340

• • I I

320

8901 23456 789

8901234 5678� 890123456 789

890 1234 56789 300

POT ENTI LLA EPILOBIUM CAREX CAREX L JUNCUS M N Q K PULCHELLA FILIFORM IS EGEDII PA LUSTRE CANESC ENS

riiii·

111.._ 890123456789 890123456 789

Fig. 6

Figs. 5 and 6. Changes in vertical distribution for some species during the investigation period 1968-79. 1st-year seedlings are excluded. Vertical bars and horizontal lines denote upper and lower limits of vegetative and floral shoots/in­

dividuals, respectively. Thick lines (only given for the most important species) denote cover exceeding 15 %, with the fo llowing exceptions: Eleocharis uniglumis and Scirpus tabernaemontani 3 %, Agrostis gigantea 5% in 1979, Clzara aspera 50 %. Thin lines denote cover below 0. 1 %. Medium thick lines denote cover above 0. 1 %. The sloping line gives the course of downward migration if coinciding with the present land uplift. Altitudinal scale refers to the Ratan mareograph station (Fig. 2).

A eta Phytogeogr. Suec. 68 66 Lars Ericson

Material and methods tion, so that the area occupied by any belt in 1968 has, in 1979, been wholly taken over by what had The permanent transect, 1 m broad, has been revised been the belt immediately above it in 1968. annually, in late September-early October, during In spite of this, the main vegetational gradient has the period 1968-79. Occurrences outside the remained unchanged. However, with regard to the transect but within 1 m have been noted. Collected details a number of important changes have oc­ data include, for example, percentage cover, number curred, which all point towards gradually more of individuals/shoots and vitality. For each species sheltered conditions (for details of the process see the lowermost and uppermost occurrences of Figs. 5-8). seedlings, juveniles, vegetative and floral in­ Most evident is the disappearance or decline of a dividuals/shoots have been measured as accurately number of features characteristic of the outer skerry as possible. zone (i.e. they are in different ways fa voured by a The transect has been levelled each year (every 50 more unstable habitat): cm). The altitudinal scale is adjusted to that of the nearest mareograph station at Ratan, situated 23 km (1) Deschampsia bottnica vanished in 1976 (Fig. NNE of the study area. Water-level data from the 5H) and Agrostis stolonifera var. bottnica in 1975 mareograph (Fig. 2) have been supplied by the (Fig. 51) due to the rapid expansion of dense Swedish Meteorological and Hydrological Institute Eleocharis uniglumis. (SMHI), Norrkoping. The 302 cm level equals the (2) The disappearance of the Carex pulchella (Fig. theoretical mean water-level for 1968. Due to land 6K) depressions, extensive in size in 1968-70, uplift this value is annually reduced by 0.85 cm thereafter decreasing due to the expansion of C. (RAK 197 1), i.e., a total of 9.35 cm for the actual in­ nigra, Molinia and Myrica, a process only temporari­ vestigation period. ly interrupted by frost-lifting in the winter 1971-72 The annual changes during the period are which severely injured the younger Molinia tussocks. presented as vertical diagrams (Figs. 5-8). These For a short period, 1970-73, these depressions have been obtained by using the horizontal values ob­ showed a succession towards a Carex canescens ­ tained at the annual revisions, and subsequently con­ Juncus filiformis community (Fig. 6L, with i.a. 0), verting them into corresponding vertical values, ob­ Ep ilobium palustre (Fig. 6N), Drosera anglica (pre­ tained at the annual levellings. This implies that a sent in 1971-73) and Nardus stricta (present in horizontal downward migration is not bound to lead 1974-7 5), characteristic for sandy, fr eshwater­ to a vertical downward migration (cf. Tables I, II) as, influenced depressions in the coastal zone. Parallel to fo r example, accumulation, litter decomposition, the closure of the Myrica shrubs a large number of frost-heaving, etc. contributes to the shore displace­ poor competitors were clearly misfavoured : Juncus ment (viz. 1-6 mm/year; Fig. 4). alpinus, Triglochin palustre (Fig. 6B-C), Parnassia The use of the terms outer and inner skerry zone, palustris, Leontodon autumnalis, Viola palustris and coastal zone, fo llows Luther (195 1). These terms (Fig. 6G-I) and Potentilla egedii (Fig. 6M). are here solely used when referring to conditions Juncus gerardi, prevailing in the transition area around the W side of (3) The decline of in 1968 eo­ Calamagrostis neglecta, the N. K varken. dominant with in 1979 only The nomenclature fo llows that of Ericson & languishing shoots. The clones present in 1968 and Wallentinus ( 1979). Agrostis gigantea is represented still extant in 1979 only showed a total vegetative by its var. glaucescens, Carex nigra by the tuftedvar. growth of 25 cm fo r the whole investigation period. recta, Eleocharis palustris by var. lindbergii, E. un­ ( 4) The decline of Filipendula ulmaria, which iglumis by var. fe nnica. Carex oederi var. pulchella reaches its maximum on more mull-rich sites and has

= C. pulchella and Potentilla anserina ssp. egedii = great difficulty in establishing in dense My rica P. egedii. Both Agrostis stolonifera and A. tenuis in­ thickets, unless they are not eroded by storm waves clude some hybrids with A. gigantea. Salix (Fig. 7A). phylicifolia most probably is the hybrid with S. myr­ (5) The decline of Agrostis tenuis (Fig. 7E) and the sinifolia. disappearance of Moehringia trinervia and Rubus idaeus, in 1971 and 1972, respectively (Fig. 4). Species of old decomposed shore drift, thrown up Changes during the period 1968-79 during severe storms in autumn or early winter, mis­ A comparison between the net changes during the 12- favoured by the closure of Picea abies. year period (Fig. 4), primarily reveals a marked A positive change, pointing towards more spatial change, i.e., a horizontal downward migra- sheltered conditions, is the rapid expansion of

A eta Phytogeogr. Suec. 68 The downward migration of plants on a rising Bothnian sea-shore 67

TRIENTALIS FILIPENDUL A MYRICA C EUROPA:A cm A B D RUBUS AGROSTIS ULMARIA GALE ARCTICUS E TENUIS

LLL.l..__l_ I_ 4S6 789

890 123456789 8901234� 31 Fig. 7. Changes in vertical distribution of some species in the upper geolittoral during 1968-79. For explanations see Figs. 5 and 6.

Agrostis stolonifera, now in extensive carpets in the autumnalis, Carex nigra and Calamagrostis neglec­ upper part of the Eleocharis uniglumis belt (Figs. 4, ta. For most other species of the upper part of the 51). The next step in this development is represented shore the colonization rate is very low, mostly due to by Carex mackenziei, though for the moment show­ the increasingly closed vegetation cover (Carex ing a decline (Fig. 6A). The colonization of nigra, Molinia coerulea and My rica gale). The com­ Op hioglossum vulgatum (Fig. 6F) more probably petition from My rica is primarily seen in fo r example reflects a gradual decline in vigour of the Carex nigra Filipendula ulmaria and Trientalis europaea, both tussocks which took place after 1973-74. showing a marked change of their lower limits ( + 10 and +9.5 cm, respectively), but as regards their main occurrence showing a retreat of -1.0 cm ( -0. 1 m) Changes in lower and upper limits and -2.5 cm (-0.5 m), respectively (Figs. 4, 7). Colonization (Figs. 5-7; Table I). The colonization This vertical retreat is also evident fo r the species pattern shows large differences, which mainly reflect of the Alnus incana border, viz., Agrostis tenuis, the vertical position of the species. Most species hav­ Luzula pilosa, Alnus incana and Picea abies. ing a remarkably high net colonization inhabit the lower part of the shore (vertical changes given in Dy ing-off(Figs.5-7; Table I). Most species having a brackets). S cirpus tabernaemontani ( + 14 cm; + 21 very high or high colonization rate show small ver­ cm 1968-73, thereafter retreat due to high water in tical changes of their upper limits, viz., Eleocharis 1974, see Fig. 2, and frost-lifting), Eleocharis uni­ uniglumis, Triglochin maritimum, Scirpus tabernae­ glumis ( +27 cm), Triglochin maritimum ( + 17 cm) montani, Juncus gerardi and Calamagrostis neglec­ and Agrostis stolonifera ( +22 cm). Apart from the ta. This is partly due to the rise of ground level (Fig. latter, which is favoured by litter-accumulation, the 4). However, it also reflectsthe ability of rhizomatous colonization of these species has not suffered from species to survive for a long time afterthey have been competition with any closed vegetation. overgrown by succeeding vegetation belts (in con­ With increasing height upward from the shore, the trast to the tufted Deschampsia bottnica and Agros­ downward migration will gradually decline. It is still tis stolonifera var. bottnica). This is further seen in considerable (range + 14 to +8 cm) fo r Agrostis the undulating limit fo r the uppermost occurrences of gigantea, Juncus gerardi, Potentilla palustris, Eleocharis uniglumis and Calamagrostis neglecta Galium palustre (the only two herbs of this transect (however the situation of the latter in 1975 was due to able to establish in a dense litter layer), Leontodon seeding in 1974), suggesting that suitable water-level

A eta Phytogeogr. Suec. 68 Lars Ericson 68

Table I. Vertical and hori zontal changes in lower and upper limit, respectively, for species present throughout

the investigation period, 1968-79. 1st-year seedl ings of perennial species are excl uded . The species are arranged according to the amount of thei r vertical change . +denotes downward migration and ; - denotes retreat in 1979 compared wi th 1968.

Deschampsia bottni ca includes the hybrid wi th D. caespi tosa (cf. Fi g, 5H ). Sci rpus tabernaemontani showed a net downward mi gration of its lower limit of +21 cm {+37 .95 m) during 1968-73; and 1riglochin pa lustr� +15.5 cm

{+10.20 m) duri ng 1968-74..

LOWER LIMIT UPPER LIMIT

vertical hori zonta l vert i ea1 hori zon ta 1

Eleochari s uniglumi s +27 .0 +37. 50 Pedicularis pa lustris +41 .0 +25. 45

Pediculari s pa lustris +24 .0 + 15.00 Fes tuea rubra +22 . 0 + 4. 95

Agros ti s s to1 oni fe r a +22 . 0 + 15.50 Rubus arcticus + 19.5 + 2.40

Triglochin ma ritimum + 17.0 +29 . 95 Rhi nanthus angusti fol ius + 16.0 + 10.10

Sci rpus tabernaemontani + 14.0 +24 . 85 Des champs i a bottni ea + 15.0 + 9. 60

Deschampsia bottnica + 14.0 + 10.40 Gal ium pa l ustre + 13.0 + 14.50

7 Agrostis gigantea + 14.0 + 9. 70 Leontodon autumna 1 is + 12.5 + 9.00

8 Potenti lla palustris + 14.0 + 10.65 Parnassia pa lustri s + 12.0 + 8. 25

9 Juneus gera rd i + 12.0 + 9.10 Potenti 11a pa lustri s + 12.0 + 2. 90

1 0 Ga 1 i um pa 1 us t re + 10.0 + 9. 80 10 Fi lipendula ulmaria + 10.0 + 4. 70

11 Filipendula ulmaria + 10.0 + 6.60 11 Carex pu lche 11 a + 10.0 + 3.15

12 Trientalis europaea + 9. 5 + 9. 30 12 Tri glochin pal ustre + 9. 5 + 9. 45

1 3 Le on todon a u tumna·l is + 9.0 + 4. 60 13 Agrosti s gigantea + 9. 5 + 2.00

14 Carex nigra + 8.0 + 9. 40 14 Myri ea ga 1 e + 8. 5 + 1.00

15 Calamagrostis neglecta + 8.0 + 6. 30 15 Tri enta 1 is europaea + 7.0 + 1.30

16 Parnassia pa lustris + 6.0 + 3. 60 16 Luzula pi losa + 5.5 + 1.15

17 Festuca rubra + 5.0 + 6.00 17 Potenti 11a egedi i + 5.0 + 3. 35

18 Sal ix phyl icifol ia + 5.0 + 5. 35 18 Agrostis stolonifera + 4.0 + 7. 30 19 Viola pa lustri s + J. 5 + 5. 35 19 Agrostis tenuis + 4.0 + 0. 95 20 Myrica gale + 3. 0 + 6.60 20 Scirpus tabernaemontani + 2. 5 + 5. 95

21 Potenti 11a erecta + 3. 0 + 3.80 21 Juncus gerard i + 2.0 + 4. 80

22 Rubus arcti cus + 3.0 + 3. 50 22 Calamagrostis neglecta + 2.0 + 4. 60

23 Mo 1 i ni a coeru l ea + 3.0 + 2. 30 23 Eleochari s uniglumi s - l.5 + 0. 40

24 Rh inanthus angusti fo l ius + 2. 5 + 6.10 24 Mo l inia coerulea - l.5 + 0.10

25 Triglochin pa l ustre + l.5 + 4. 40 25 Carex nigra - 2.5 - 0.15

26 Potenti 11a egedi i - 0. 5 + 4.10 26 Viola pa lustris - 2. 5 + 0.45

27 Agrostis tenuis - 1.0 + 0.40 27 Sa 1 i x phy 1 i cifo 1 i a - 3. 5 - 0. 25

28 Carex pul che1 1a - 1.0 + 0.20 28 Potenti lla erecta - 3.5 - 0. 35

29 Luzula pi losa -15.0 - 1.60 29 Tri gl ochi n mari ti mum - 6.6 - 0. 50

A 1 nus i ncana , trees + 10.0 + 2.60 A 1 nus i ncana , trees - 2.0 0.0

shrubs - 2.0 + 0. 30 -- , shrubs 1.0 0.0

juv. 3. 5 - 0. 75 -- ,juv . - 5. 5 - 1. 20

Picea abies > 5 m - 2. 5 0.0

< 2 m - 2.0 0.0 Sorbus aucuparia, shrubs + 15.5 + 3.95

conditions may provide the development of shoots at Vegetative and generative colonization higher levels than in the preceding year (cf. also the A comparison between the role of vegetative and variable upper limit of floralshoots in Fig. 5E, K, M). generative establishment fo r the downward migration A most remarkable stability is evident for Carex of rhizomatous and stoloniferous species reveals nigra and Molinia coerulea, both deep-rooted, with a (Table 11): prolonged senescence stage. The highest figures of vertical change are found (1) that for all species, with the exception of among the species of the upper part of the transect, Calamagrostis neglecta, with no successful seedling i.e., those suffering from the increased closure of survival below its expanding clone edge (cf. Fig. 5K), Picea abies in the alder fringe, viz., Festuca rubra, the generative establishment has been more effective Filipendula ulmaria, Potentilla palustris and Rubus than the vegetative, though for Rubus arcticus this arcticus, and from the expanding Myrica gale, viz., difference is small. Leontodon autumnalis, Parnassia palustris, Carex (2) that the hypothetic vegetative growth is inferior pulchella and Triglochin palustre. to generative establishment (exception Calama-

A eta Phytogeogr. Suec. 68 The downward migration of plants on a rising Bothnian sea-shore 69

Table 11. The ro le of vegetative and generative establishment in the downward mi gration of rh i zomatous and stoloniferous species . The table only includes those species for which the vegetative spread , duri ng at least one year, exceeded 15 cm . True vegetative growth is cal culated by adding the maxi mum vegetati ve spread, measured for the lower, expanding, zone duri ng each yea r. Hypotheti c vegetative growth is ca l cu­ lated by mu lti plying the maximum annual growth value measured by 11. The species are arranged according to thei r lower limits in 1978. + denotes downward mi gration and ; - denotes retreat in 1979 compared wi.th 1968.

vegetati ve ge nerati ve

true hy�otheti c

hori zonta 1 vertical hori zonta 1 vertical hori zonta 1 verti ea 1 (m) (cm) (m) (cm) (m) (cm)

Sci rpus tabernaemontani +7 0 50 +6 .0 + 120 75 + 9.0 +24 0 85 +14.0

Eleochari s uni glumi s +6 .45 +6.0 + 15.40 +15.0 +37 0 20 +27 0 0

Agrosti s stoloni fera +4 0 25 +2 .0 + 11.00 + 17.0 + 150 50 +22 .0

Ca l amagrosti s neglecta +6 . 30 +8.0 + 150 95 + 18.0

Juncus gera rdi + 1 0 45 -2 .0 + 3 0 30 + 0. 5 + 8.85 + 12.0

Ga l ium palustre +3 0 70 +40 5 + 7 0 70 + 505 + 9.80 + 10.0

Tri glochin palustre +0 . 90 -2 0 5 + 1 0 65 - l.0 + 4 0 35 + l.5

Potenti 11a egedi i -0 0 85 -2 0 5 + 4.40 - 0 0 5 + 4.10 - 0.5

Rubus arcti cus +2 .80 +2 .0 + 4.40 + 3 0 5 + 3 0 40 + 3.0

Tri enta 1 is europaea + 1.20 -0.5 + 3 0 30 + 2 0 5 + 9. 20 + 9 0 5 grostis neglecta; cf. point 1). However, for C. neglec­ about the same littoral level. This example shows the ta, Eleocharis uniglumis, Scirpus tabernaemontani advantage of generative establishment, enabling and Agrostis stolonifera it is large enough to ascer­ rapid expansion during, e.g., suitable low-water tain a downward migration which equals or exceeds years, cf. the dominance of A. gigantea below C. the rate of land uplift. neglecta in 1970-73, compared with the reverse (3) that the importance of vegetative growth will situation in 1968. But it also shows the disadvantage diminish for species higher up the shore and that it is that seedling survival is inhibited during high-water years (1974) while vegetative growth, even if im­ only for Calamagrostis neglecta, Eleocharis uni­ peded, continues. Hence the retreat of A. gigantea glumis and Scirpus tabernaemontani that vegetative growth may be of importance in maintaining the (due to unsuccessful seeding, senescence, too dense litter) which resulted in it reaching its maximum relative vertical position for the species. above C. neglecta in 1979. This changing position The result implies that vegetative growth is inferior to also reflects the general occurrence of the two seedling survival fo r the downward migration on flat species. In the outer skerry zone A. gigantea is con­ shores, while on more sloping shores, it may be of fined to the Festuca rubra belt, while C. neglecta will greater importance. establish lower down, in the Juncus gerardi belt. With Consequently an increase in shelter, leading to in­ increasing shelter A. gigantea will migrate to a lower creasing accumulation of litter, will delay the level than C. neglecta. When litter accumulates downward migration for clonal species that need an throughout the year the colonization of A. gigantea is open, unstable site fo r successful seedling-survival prevented and C. neglecta expands. Hence, once (naturally this is also evident fo r species without again A. gigantea will become confined to an in­ vegetative spread). An excellent example is provided creasingly higher littoral level on sheltered sites in the by Juncus gerardi, which during the investigation coastal zone. period has germinated (and survived) solely on sites influenced by frost-lifting. After 1971 these weakly Influence by water-level fluctuations expanding clones showed a recessive lower level (Fig.

5M; Table I), which will ultimately lead to an in­ Annuals and biennials. The annuals and the biennials creasingly higher littoral level (cf. its mid-geolittoral present in this material (Fig. 8) represent a very position in the outer skerry zone with its occurrence heterogeneous group with regard to their habitat in the upper geolittoral in the coastal zone). demands. However, thay have in common a variable The principal importance of the differencebetween vertical position, a change which mainly reflects the vegetative and generative establishment can be ex­ water-level situation during the different years (Fig. emplified by Calamagrostis neglecta (Fig. 5K) and 2). Agrostis gigantea (Fig. 5L), both with rhizomes and This is evidenced by Rhinanthus angustifolius an annual growth which may amount to 1.5 m and (Fig. 8D), abundant in the Carex pulchella and 0. 1 m, respectively. Furthermore, they often inhabit Molinia coerulea depressions in front of the My rica

Acta Phytogeogr. Suec. 68 70 Lars Ericson

cm PEDICULARIS PALUSTRIS cm EUPHRASIA E C 340 BOTTNICA MELAM PYRUM F PRATENSE

390 . , � I , 320 890 12� 4 56 789

8 � 0123 � 56 78? RHINANTHUS 370 0 SUBULARIA ANGUSTI FOLIUS /I A AOUATICA

300

I. 350 I 8901 2345678 9 B CRASSULA 280 AOUATICA

8 9 0 1 2 3 4 6 7 8 9 68 69 70 71 72 73 74 75 76 77 78 79 8 9 0 !J 1 2 3 4 5 6 7 8 9 330

Fig. 8. Change in vertical distribution of annual and biennial species during 1968-79. Explanations: Pedicularis palustris, thin lines lst-year individuals, medium thick lines 2nd-year individuals. Otherwise see Figs. 5 and 6.

shrubs in 1968 (the higher level most probably due to water in the summer (hence its absence in 1977). its occurrence and subsequent seed dispersal during While the above-mentioned species fa irly well the high-water year 1967; cf. also the situation fo r reflect the water-level situation in summer, Melam­ Pedicularis palustris, Fig. 8E), while in the low-water pyrum pratense (i.e., its lower limit, Fig. 8F), as a year of 1969 it extended down through the Agrostis species of the Alnus incana border, reflects the water gigantea and Calamagrostis neglecta belts. During level in late autumn-early winter. High waters, as in the fo llowing years its position reflectsthe water-level 1969-70, 72-73 and 78-79, connected with ice fo r­ so that it occurs at a higher level either in summers mation up into the alder fr inge, are fatal to the with a high water-level (e.g., 1974) or in years fo llow­ Melampyrum seeds, causing a retreat in the fo llowing ing high waters in autumn (e.g., 1970) and at lower year, while low-water winters (1968-69, 1975-76 levels during low-water years (1976). Its decrease and 1976-77) allow survival at lower levels. during 1977-79 reflects both the expansion of As regards the remaining annuals the fo llowing My rica (cf. also Pedicularis palustris), but also the can be mentioned. Euphrasia bottnica (Fig. 8C), oc­ ill-effects of shorter high-water periods in the summer curred in the Carex pulchella depressions in (not evident in Fig. 2). 1968-69. In 1970 its seedlings appeared at the Roughly the same trend is also shown by the bien­ 334-340 cm level, in front of the Myrica border nial Pedicularis palustris (Fig. 8E). The vertical dis­ (dispersal due to the high water in November 1969), tribution of the 1st -year individuals (i.e., the but succumbed in early July. Crassula aquatica and vegetative rosettes) reflects the water level during the Subularia aquatica (Fig. 8A, B) both appeared dur­ summer, e.g., low-water in 1969, 1971, 1973 (the ing the low-water in 1969. Crassula, as the water level in June-July was lower in 1973 com­ ecologically closely related Ranunculus reptans (Fig. pared with 1972) and 1976 and high water in 1970, 5G), vanished aftera few years, mainly due to the ex­ 1974 and 1975. The position of the 2nd-year in­ pansion of the graminids. Subularia shows a variable dividuals (i.e., in the autumn, the dead inflorescences position, partly reflecting the water level in early with ripe seeds) partly reflects the conditions during summer, but more a result of the uneven and irregular the preceding year, but above all the winter con­ seeding. ditions, viz., good survival fo llowing calm winters, e.g., 1967-68, and high mortality fo llowing winters Perennials. The established populations of perennial when the rosettes have frozen into the solid ice, e.g., species naturally do not react as clearly as annuals 1969-70 and 197 1-72. Afterthe low-water in 1976 and biennials to a changing water-level. However, a P. palustris has been confined to the upper part of the summer with an exceptionally high water-level, as in Eleocharis uniglumis belt (with Agrostis stolonifera). 1974, will cause a reduced vigour (cf. the retreat of Hence its more irregular occurrence as it is suscep­ "dominance area" or of floral shoots) fo r most tible both to dense litter and short periods of high species (see Figs. 5-7). This is most striking for (1)

A eta Phytogeogr. Suec. 68 The downward migration of plants on a rising Bothnian sea-shore 71

some species in the lower part of the shore, viz., Scir­ (Figs. SM and 6B, D, E), and less so fo r species at pus tabernaemontani, Eleocharis uniglumis, Agrostis somewhat higher levels, viz., Carex nigra, Molinia gigantea and Calamagrostis neglecta (Fig. SC, E, K coerulea and Parnassia palustris (Figs. SN, and 0 and L) and (2) some species of the alder fringe, viz., 6G). Favourable periods for generative establish­ mainly Agrostis tenuis and Trientalis europaea (Fig. ment also took place in the years 1970-73 and 1976. 7C, E) but also Filipendula ulmaria and Rubus arc­ However, the differences between the species are ob­ ticus (Fig. 7 A, D). Other examples are Galium vious, e.g., Triglochin palustre, Leontodon autum­ palustre, Potentilia palustris, Leontodon autumnalis, nalis and Viola palustris (Fig. 6C, H, I) show Pc.rnassia palustris and Viola palustris (Fig. 6D, E, successful establishment only after rainy summers G. H and I). All these species are strongly mis­ (1973, 1976), and Festuca rubra (Fig. 6J) only after favoured, the three latter being without any seed mild autumns (1970, 1973). Another reason for production in 1974. The only species for which the delayed downward migration is the absence of suit­ vegetative growth is clearly favoured by higher able habitats, viz., Carex nigra, C. pulchella, Juncus water-levels is Agrostis stolonifera (Fig. SJ), as its alpinus and Triglochin palustre primarily need stolons readily adapt to the prevailing water surface. shallow depressions with impeded drainage; Molinia However, for the seedling survival and conse­ colonizes the top of Carex nigra tussocks and quently for the downward migration of a species, the Festuca rubra also the top of Molinia tussocks. water-level situation is of crucial importance. The Furthermore, the increasing litter layer provides an very rapid colonization by, above all, Scirpus taber­ effective barrier to most species (exception Galium naemontani and Eleocharis uniglumis (Fig. SC, E), palustre and Potentilla palustris). but also by Agrostis gigantea and Deschampsia bott­ Yet another picture is obtained for those species nica (Fig. SL, H), in the beginning of the investigation which first appear around the lower limit of the alder period, was a result of successful seedings with high fringe viz., Filipendula ulmaria, My rica gale, Trien­ survival values during the low-water summers of talis europaea, Rubus arcticus and Agrostis tenuis 1968 and 1969. In 1970 the water-level was high (Fig. 7). With the exception of the establishment of enough to prevent any successful seedling survival My rica following 1968-69, and fo r the other species but low enough to ensure the survival of juveniles es­ following 1976, it is difficult to trace any fa vourable tablished in 1968-69. This, together with the suitable effects of low-water periods in summer. Instead, the water-level situation in 1972-73, favoured a rapid most prominent fe ature is the marked retreat shown establishment, which was not stopped until the high­ by all species, except My rica which only showed water period in the summer of 1974, causing heavy decreased vitality, during the summer high-water in losses both among genets and ramets. During the 1974. next low-water period, 1976-78, another expansion The susceptibility shown by Filipendula ulmaria took place but less pronounced due, fo r example, to and Trientalis fo r the combination of high-water and the ill-effects fo llowing frost-lifting in 197S-76 cold-spells in late autumn is also evident (in the same (devasting fo r Scirpus tabernaemontani, favour­ manner as Melampyrum pratense). This is also a able fo r Deschampsia bottnica x caespitosa, which decisive factor fo r the establishment of A In us incana replaced D. bottnica, earlier fo rced out by Eleocharis and Picea abies, but not so important for Salix uniglumis). This development also occurred for phylicifolia. A prerequisite for the successful es­ Triglochin maritimum (Fig. SD), but as it germinates tablishment of all these species is not only a low in autumn (the above-mentioned species germinate in water-level in summer, but also the absence of early summer), some details differ. Its generative es­ prolonged drought periods and freezing into the ice. tablishment is favoured by mild autumns with a low The effect of the latter factor is especially severe water-level 968, 1976) and misfavoured by cold­ when connected with a sudden rise of the water-level. (I spells in connection with a low water-level (e.g., 1969). Conclusions Higher up the shore the seedling survival becomes less clearly correlated to water-level, so that pro­ The present data reveal, fo r example, a striking nounced low-water periods may not automatically be difference in the net downward migration between so favourable, e.g., when connected with drought species due to their littoral position (Table I). As the periods. However high water-levels in summer (1974) values obtained only refer to one transect, it may be will prevent all seedling survival. Most species questioned whether the rapid expansion shown for showed a successful establishment during 1968-69, some species of the lower littoral (Fig. SC-E) really being most pronounced for Juncus gerardi, J. is the result of an unusually suitable water-level situa­ alpinus, Galium palustre and Potentilla palustris tion in 1968-69, or if it simply is the result of a

A eta P/1ytogeogr. Suec. 68 72 Lars Ericson sudden change in exposure. This cannot be wholly SUMMARY evaluated until data from other transects are The vegetation on a rising sea-shore on the Bothnian published. However, data on Deschampsia bottnica Bay has been studied during the period 1968-79. (Ericson 1973) fr om an exposed morainic shore, 22 The aim was to investigate the mode of plant km SW of the present locality, shows that in 1972 the colonization and whether its rate equals the rate of individuals germinated in 1968-69 had caused a land uplift in the area (8.5 mm/year) or if any other downward migration of D. bottnica amounting to 10 factors are operative. It was found: That the net cm, and if those germinated in 196 7 are included, the (I) downward migration varied according to the littoral value amounts to 16 cm. Consequently, it may be position of a species; highest values for those of the concluded that the changing water-level situation is lower littoral, and lowest values for those of the upper of the utmost importance for the short-term changes. littoral. (2) That, with the exception of Calamagrostis However, as also suggested by the present result, the neglecta, generative establishment was more impor­ downward migration taking place during the first tant than vegetative establishment. (3) That, in the rather continuous low-water period (1968-72; cf. short-term, the annual changes in water-level are of Fig. 2), might have been too rapid for Scirpus taber­ the utmost importance fo r the establishment, though naemontani and Eleocharis uniglumis, as both other factors increase in importance in the upper lit­ showed a marked decrease during the fo llowing high­ toral. However, the investigation period is too short water period, a decrease which also continued during to evaluate the importance of land uplift. the second low-water period (1976-78; cf. Figs. 4 and 5C, E). The rather regular course between low-water and REFERENCES high-water years (cf. Fig. 2), and for longer periods Angstrom, A. 1974. Sveriges klimat. 3rd ed. 188 pp. - (Lisitzin 1973; Ussisoo 1977) implies that years Stockholm. suitable for seedling survival occur rather regularly, Atlas over Sverige (National Atlas of Sweden) and the result shows that species of the lower littoral 1953-1971. Nos. 31-32. - Stockholm. are well adapted to take advantage of the low-water Ericson, L. 1973. Ekosystemens succession vid en years. Hence their downward migration in the short­ landhojdningskust. Strandvegetation. Forskn. term (about 10 years) may show a very weak cor­ Framsteg 1973(2):32-33. respondence with the rate of land uplift. Ericson, L. & Wallentinus, H.-G. 1979. Sea-shore vegeta­ As other factors (e.g., drought, litter, etc.) become tion around the Gulf of Bothnia. Guide for the Inter­ national Society for Vegetation Science Excursion, increasingly decisive higher up the shore, this implies July-August 1977. - Wahlenbergia 5:1-142. that seedling survival becomes increasingly tricky, Johannessen, T.W. 1970. The climate of Scandinavia. In: though regular enough to enable a downward migra­ Climates of Northern and Western Europe (ed. C.C. tion which, over a 1 0-year period, rather closely cor­ Wallen). - World Survey of Climatology 5:23-62. responds to the land uplift. For species in the upper­ Lisitzin, E. 1960. Seas, map 7:12-16. - Atlas of Finland. most part of the shore, and especially so for the Helsinki. lignoses, a durable establishment is a much rarer - 1973. Sea level variations in the Gulf of Bothnia. - event, as so many prerequisites have to be fulfilled. Nordic Hydrol. 4:41-53. Hence the downward migration shows a very small Luther, H. 1951. Verbreitung und Okologie der hoheren correspondence with land uplift during a short-term Wasserpflanzen in Brackwasser der Ekenas-Gegend in Si.idfinnland. - I. Allgemeiner Teil. - Acta bot. fe nn. investigation. The colonization may best be 49:1-23 1. characterized as a step-by-step process, i.e., when all 1961. Veranderungen in der Gefasspflanzenflora der prerequisites are fulfilled a rapid and lasting estab­ Meeresfelsen von Tvarminne. - Ibid. 62: 1-100. lishment may take place, as shown for Myrica gale RAK 1971. Geodetic activities in Sweden 1967-1 970. ­ during the investigation period. Another example is Rikets All m. Kartverk Meddn. A 38:1-8. provided by Picea abies for which all shrubs present Schwanck, B. 1974. Vegetationens fOrandring pa de in 1968, and still extant in 1979, in the lower part of framvaxande skaren. - Skargard i omvandling. pp. the alder fr inge, had germinated in 1960 or 1963, two 79-85. Helsingfors. pronounced low-water years (cf. Lisitzin 1973; Ussisoo, I. 1977. Computation of land uplift and mean sea Ussisoo 1977). level in Sweden. - Geol. Foren. Stockh. Forh. 99:42-48. Vuoristo, A. & Rousi, A. 1976. Changes during 35 years in the range of Hippophae rhamnoides in the archipelago of Rauma on the Gulf of Bothnia coast. - Lars Ericson, Department of Ecological Botany, Umea Aquilo, Ser. Bot. 14:1-5. University, S-901 87 Umea, Sweden

Ac!a Phytogeogr. Suec. 68 Ragvaldsmossen A pollen-analytical study of the Postglacial fo rest history at the "limes norrlandicus ", South-Central Sweden

Magnus Fries

In the neighbourhood of Stora Skedvi, the birth-place clayey gyttja (clay mud). (3) Still more favourable of Hugo Sjors, a tourist route named Husbyringen climatic conditions led to abundant micro- and was arranged through the initiative of Olof Sundell, macro-organisms and increasing deposition of gyttja, head of the Information Department of the Korsnas­ grading upward into dy (brown humic mud). (4 and Marma Forest Company. Along this route several 5) The open lake with Ny mphaea and Potamogeton objects of natural and cultural interest can be studied, species was gradually occupied by reed (Phragmites) among them a small bog, Ragvaldsmossen, about and sedges. A fe n was then built up, at least locally 180 x 70 m in size. It is situated in a highland area with birch, alder, and willows. (6) The fen was then (150-250 m above present sea level) in the south­ overgrown with Sp hagnum, forming a more or less eastern part of the province of Dalarna (Dalecarlia), thick bog peat, which extends to the surface. more exactly ea. 3 km S of Stjarnsund (Fig. 1). For the main part the area is covered by conifer fo rest. The distance to arable lowland is at least two or three kilometers. Most of Ragvaldsmossen is occupied by a northern bog (ombrogenous mire), which is slightly raised oak border above the surrounding lagg (marginal fe n). On the whole, the mire shows the fe atures typical of the eastern part of the Bergslagen area, which is in a tran­ sitional zone between the southern mostly om­ brogenous and the northern mainly geogenous mire types (Sjors 1948). Thus the bog is covered by a pine-dwarf-shrub community (Calluna, Vaccinium vitis-idaea, V. myrtil/us, V. oxycoccus and An­ dromeda) with much Eriophorum vaginatum and, in the bottom layer, mostly Sphagnumfuscum fo rming hummocks (Fig. 2). The lagg shows the normal fe n vegetation with sedges, mostly Carex rostrata, and ordinary fen mosses (also Scorpidium scorpioides). Ragvaldsmossen is situated in a small basin about 205 m above sea level. The highest shoreline in this area is at about 196 m. Thus the sea did not reach into this basin, but the highland area fo rmed an archipelago in the Yoldia Sea and Ancylus Lake dur­ ing the beginning of Postglacial time. 0 50 lOOk m The Postglacial development of the basin was in­ vestigated through borings and sampling, ac­ complished with the assistance by Forester Bjorn Fig. I. The southern half of Sweden, showing the position of the Ragvaldsmossen site at Stjarnsund in relation to the Jahnke. It is illustrated by simple sections published "limes norrlandicus" (vertically hatched belt). This tran­ in the journal "Skogen" (62/ 1975 p. 622 f.), with a sitional zone separates the northern boreal region with brief text in Swedish. Six stages are represented. ( 1) mostly conifer fo rests from the southern boreo-nemoral An ancient lake had slight deposition of fine mineral region with conifer forests interfered, among others, by material on the till substratum with very little organic thermophilous, deciduous trees (terminology by Sjors content. (2) Increasing temperature resulted in slowly 1963 and others). The oak and spruce borders are ap­ increasing organic production and deposition of proximately drawn.

A eta Phytogeogr. Suec. 68 7 4 Ma gnus Fries

pollen of shrubs such as Corylus, Salix and Hip­ pophae·, which are supposed to have formed their own communities. The percentages of other scattered pollen and spore types are calculated on the basis of the AP sum. Except for Sphagnum, Gramineae, and Cyperaceae, which are concentrated in special levels, the amount of these types is insignificant. The pollen diagram is as a principle, conventional­ ly arranged (Fig. 3). It should be noted that the scale of percentage varies in different parts of the diagram. Further, the vertical scale is doubled below the 7.25 m level because of the fact that the sedimentation rate was lower at this early stage than later on. No radiocarbon dates are available for the site. However, an attempt is made to insert a chronology in the pollen sequence on the basis of the dates from other diagrams. In this diagram the figures for each millennium (except for the two latest) are tentatively inserted. The chronological nomenclature fo llows the scheme of chronozones by Mangerud et al. (1974 p. Fig. 2. A view of the bog plane of Ragvaldsmossen 122). with pine-dwarf-shrub vegetation of east Swedish Surveys of the late-Quaternary history of the character. In early Postglacial time there was a lake in this vegetation of Sweden have been published, for in­ place. For the stratigraphy see the text. - Photo Sun­ 0. stance, by Fries (1965a and b) and Berglund (1968, dell. in Swedish). Basic facts and material No comprehensive regional treatment has been The development of the vegetational landscape carried out in order to elucidate the Postglacial On the whole the total Postglacial sequence of history of the comparatively warmth-demanding vegetational development is represented in the pollen trees, shrubs, or other plants in this transitional zone, diagram. The ice disappeared at about 7600 B.C. the "limes norrlandicus" (Sernander, unpubl.) or (Sveriges Geol. Undersokn., Ser. Ba no. 18, 1961 ), "borderland" (Fransson 1965), between the southern but the very first vegetation in the surroundings is boreo-nemoral and the northern boreal zones (Sjors probably not represented in the present sequence. 1963 ). However, several local studies are presented in The pollen and spores from this pioneer vegetation special papers and surveys, for instance in the must be included in the underlying mineral sediment. descriptions to the geological maps of the area in question (Sveriges Geol. Undersokn., Ser. Aa and Preboreal (- ea. B.C.) Ca). In the present paper such a local study is 7000 presented on the basis of pollen analysis. It deals with The first two spectra (levels 7.50 and 7.45 m) show a the principal features of the vegetational develop­ very early stage of vegetation. According to the large ment, especially the forest history, during the time amount of Betula pollen, the landscape was mainly represented by the deposits in the Ragvaldsmossen covered by a birch fo rest, the character of which we basin, i.e., from early Postglacial (Preboreal) time un­ may only guess. The insignificant values of non­ til now. arboreal pollen (NAP) indicate that fo rest-free areas The samples fo r pollen analysis were taken from of dwarf-shrubs, grasses, and herbs were very the deepest point in the centre of the bog. They were restricted. In fact, these two spectra show somewhat treated in the usual manner (acetolysis, also HF for higher values ofEricales (and other NAP as well), but the basal samples). The calculation of percentages is not as high as in the corresponding parts of diagrams based on the sum of arboreal pollen (AP), including from S. Scandinavia. In the Postglacial chronology

Fig. 3. Pollen diagram from the Ragvaldsmossen site at Stjarnsund in south-eastern Dalarna. For explanations and inter-)lo­ pretations see the text. Note the different percentage scales and the change of the vertical scale of depth between 7.20 and 7.25 m.

Acra Phyrogeogr. Suec. 68 SE. Da larna, Stjarnsund, Ragvaldsmossen, 205 m above sea level

• Pinus F Fagus S Salix 4i j E t::. Picea C Carpinus Po Populus � o � .� m Betula Ac Acer Hi Hippoph. 0 E dm D Alnus QM

Acta Phytogeogr. Suec. 68 76 Magnus Fries this Betula pollen zone stands fo r the Preboreal, quence. It is characterized by comparatively high which lasted to at least 7000 B.C. values of Alnus and partly Corylus. Especially Beside the pioneer birch forest there were some remarkable is the dominance of Ulmus among the stands of Hippophaii,probably along the shores or in Quercetum mixtum (QM). Hippophaii seems to be other places free of birch trees. In the lowermost gradually driven out by competition of Alnus, sample Populus reaches somewhat higher value than probably A. glutinosa, on the shores. Humulus pollen in the over-lying layers. However, the Populus pollen in some samples indicates the occurrence of hop in is often difficult to identify with confidence. the alder thickets. In the chronology this stage may The relatively high amount of Corylus pollen, at approximately be dated as between 6000 and 5000 least to some degree, may result from the redeposi­ B.C. tion of old pollen grains in this partly mineral sedi­ ment. Other types of pollen may also partly be Middle and Late Atlantic (ea. B.C.) redeposited (Ulmus, Quercus, Fraxinus, and Alnus). 5000-3000 My riophyllum alterniflorum was the pioneer The increase of the sedimentation rate continues. The among the higher water plants. pollen sequence between about 6.50 and 4.10 m is conspicuously uniform. The Pinus and Betula values show about the same, rather high amount (ea. 40 per Early Boreal (ea. B.C.) 7000 -6500 cent). The pine forest probably grew on more or less The fo llowing two samples (levels 7.40 and 7.35 m) dry ground in the surroundings, as previously and to­ are characterized by high Pinus values and conse­ day. The birch forest grew presumably on better soil. quently low percentages of other components. The The birch could fo rm permanent stands because the amount of redeposited pollen grains should be in­ competitor of today, the spruce, had not yet invaded significant or none. The pollen of Corylus, Ulmus, the country. Alnus, Corylus,and the QM constituents Quercus, and Alnus may result from long-distance are comparatively well represented throughout the transport from, for instance, S. Scandinavia, where whole zone, especially Tilia. No doubt these warmth­ especially hazel forests were common or even demanding trees grew in favourable places in the dominating. This stage can probably be labelled Ear­ vicinity. The wild occurrences of today are situated ly Boreal, lasting to about 6500 B.C. Hippophaiistill more or less far from this area. grew locally. The upper boundary is defmed by the Ulmus Ny mphaea alba coli. appeared (both pollen types, decline at the 4.40-4.00 m level. Evidently it is an ex­ alba s.str. and candida) as did Potamogeton sp. and, ample of the pollen-analytical index level (about 3000 sparsely, Sparganium sp. B.C.) that is described from several places in S. Scan­ dinavia and interpreted by some as an indication of a climatic deterioration at that time and by others as a Late Boreal (ea. B.C.) 6500-6000 result of cultural interference. In this case there is no The three spectra at 7.30, 7.25, and 7.20 m are doubt that the elm was fo rced aside for some climatic characterized by still comparatively high Pinus reason. Scattered Humulus pollen grains have been values. The low but significant amount of Corylus observed throughout the whole zone. (about 5 per cent) may indicate that hazel was estab­ The lake was more or less covered by Ny mphaea. lished in suitable places in the neighbourhood. Also Potamogeton sp. occurred. Seeds and fruit Whether this is applicable to Ulmus and Quercus is stones, respectively, have been found. uncertain, but probably they grew somewhere in the These stages may approximately be dated as region. Especially evident is the occurrence of Alnus, between 5000 and 3000 B.C. They represent the although it is uncertain which species (or both climatic optimum of the Postglacial. species), at first presumably A. incana (cf. Tallantire 1974 and literature cited). Hippophaii obviously Subboreal (ea. B.C.) retained its position. In the absolute chronology this 3000-500 stage is dated between about 6500 and 6000 B.C. The pollen sequence between about 4.10 and 1.90 m is fairly uniform and fo r that reason hardly possible to subdivide into Early, Middle, and Late Subboreal Early Atlantic (ea. B.C.) 6000-5000 (Mangerud et al. 1974). Possibly there is a slight From the level of about 7.15 m upwards the sedimen­ change in the forest composition at about 3.00 m. tation rate increases. Note the change of vertical Some changes may, however, have a local basis (level scale at that level. The spectra between 7.15 and 6.50 3.40 m). m form together a distinct section of the pollen se- The Pinus values dominate, indicating rather ex-

A eta Phytogeogr. Suec. 68 Ragvaldsmossen, a pollen-analytical study 77 tensive pine fo rests. The amount of Alnus, Corylus, of succession. Charcoal particles were noted in the and QM is, on the whole, slightly lower than before peat at 0.90 m. but considerably higher than during the following During this period (Subatlantic) the amount of Subatlantic time. It is quite evident that hazel and the pollen of Corylus and the warmth-demanding trees is warmth-demanding trees still occurred in the so inconsiderable that it can be explained by long­ neighbourhood although only in restricted, especially distance transport from at least some miles away. favourable sites. These trees and hazel still grow today in a few, very The fe w scattered Picea pollen grains observed restricted places in the south-eastern part of the may be interpreted as a result of long-distance province of Dalarna. However, oak especially has transport from spruce forests in a shorter distance to been planted in the cultural landscape in this area this area than before, for instance Finland (cf. Moe during the last centuries. The few Fagus and Car­ 1970; Tallantire 1972, 1980; cf. also below). pinus pollen grains, observed in the Subatlantic layers The lake was grown over by peat at about the tran­ are certainly of distant origin. sition Atlantic/Subboreal. Pollen grains of Utri­ From the 0.60 m level upwards pollen grains of cularia, Menyanthes, Drosera, Comarum, Rubus Cerealia occur (Secale, cf. Avena, cf. Hordeum). chamaemorus, and others are found. Further, the amount of weed pollen increases. They were transported from some distance. During these last centuries the forest has been more or less in­ Subatlantic (ea. B.C. -present time) 500 fluenced by selective felling and cattle grazing (cf. The invasion of spruce implies the most radical Zackrisson 1978). change in forest composition during the Postglacial. Unfortunately no material sufficientfor C dating of 14 the rise of the Picea curve (about 1.90 m) was ob­ tained. However, this event must have happened Material fo r comparison sometime between 1000 and 500 B.C., according to Although incomplete in some respects, this pollen the compilations made by Moe (1970) and Tallantire diagram fillsa gap in the network of diagrams in this (1972, 1980). Quite a new fo rest picture was estab­ transitional area characterized by fundamental dis­ lished in a rather short time, a century or two. placements of the boreal and boreo-nemoral vegeta­ With its capacity to compete against the birch, the tion through Postglacial time. Other diagrams of spruce was able to break the permanent birch woods similar kind have been recently published from an and to fo rce out the birch to restricted open places at area 50-70 km NW of Ragvaldsmossen (Hedin forest margins of different kinds and, above all, to 1975 and Hellman-Lutti 1975). They show clearly forest fire areas. The Picea curve fluctuates con­ the same main features. Further, there are some siderably, affecting the other tree pollen curves. diagrams in the descriptions to the geological maps Variations of the Picea curve were observed already which, although somewhat older, may be used for by von Post in Norrland (North-Swedish) diagrams comparison (Sveriges Geol. Undersokn., Ser. Aa ( 1930). He interpreted them as climatically depen­ 184, 185, 189, 194 and other nos., Lundqvist 195 1, dent, regional changes, and he based pollen­ int. al. diagrams pp. 92, 131, and 1963). analytical index levels (a and b) on them. The Some of these diagrams indicate that the sediment irregular course of the spruce, birch, and pine curves, plains with their ravines and the adjacent till slopes however, may be explained in a more natural way, along the River Dalalven from A vesta in the SE up to viz. as a reflexion of succession afterfo rest fires, Lake Siljan in the centre of the province were and still which in earlier days were a natural factor in the are climatically and edaphically more favourable boreal fo rest ecosystem (Zackrisson 1977, 1978, than the surrounding highland areas. The diagram 1981 and cited literature ; for boreal N. America see from the Gagnef plain (ea. 50 km NW of Wright 1974 and cited literature). A pollen-analytical Ragvaldsmossen and 150 m above sea-level) is expression of forest firesin boreal Sweden is shown in especially remarkable, for it shows a QM curve several diagrams, for instance from south-eastern reaching nearly 30 per cent during the period of the Jamtland (Fries 1956, e.g. diagram 2 B). climatic optimum. The forest fires indubitably became more frequent Thus it is quite evident that the warmth-demanding after the invasion of the flammable spruce than they trees and their companions penetrated deep into the were during earlier times. Because of the local Norrland terrain, from where they gradually dis­ irregularities of the fo rest fires the relations between appeared because of climatic deterioration, futile the course of the spruce, birch, and pine pollen curves competition with the invading spruce, soil acidifica­ may not be expected to follow constantly the scheme tion, and local interference by man. In fact, this

Acta Phytogeogr. Suec. 68 78 Magnus Fries phenomenon was understood already through among others, are produced in enormous amounts in Hedstrom's and G. Andersson's studies of the oc­ large areas of southern and south-eastern Europe. currence of subfossil Corylus nuts 893 and 1902, This source is far away, but long-distance pollen (I respectively). Through pollen analysis the picture has transport of that dimension is certainly well-known. become more substantial and detailed. The minute and rather even occurrence of these pollen types throughout the whole sequence may in­ dicate a continuous supply of pollen from a persistent Short- or long-distance transport of pollen source. The problem of the extent of long-distance transport of pollen is not solved. In the present pollen diagram Reconstructions of three vegetational stages there are some details that may be regarded from this point-of-view. On the basis of the pollen spectra and a considerable A fe w Picea pollen occur in the pre-Subatlantic amount of imagination, an attempt is made to sequence, from the 4.80 m level upwards, to an extent reconstruct the three main stages of the vegetational of 1 per cent or less. Concerning the fen peat at development in the area of this study, viz. 4.80-4.60 m the field notes indicate that the consis­ the early Postglacial time (Preboreal) with (I) tency is very loose. The risk of contamination from pioneer birch forest dominant and locally some open layers above (for instance 0-2 m) therefore exists. ground vegetation (Fig. 4); Note also the occurrence of a Fagus pollen at level 4.60 m. However, this possibility seems to be negli­ (2) the warm period, more exactly the stage of gible for the following scattered occurrences. The ex­ climatic optimum lasting approximately between istence of local spruce stands in the area during Sub­ 5000 and 3000 B.C. and characterized by warmth­ boreal time cannot be denied. Such a probability is demanding trees and hazel and by permanent fo rests pointed out in other connections (e.g. von Post 1924 of birch as well as pine (Fig. 5); pp. 125; Lundqvist 1929 p. 3 77; Florin 1944 p. (3) the late Postglacial time (Subatlantic) with spruce Ill, 567). However, an amount as low as this per cent as a most important element (beside pine) in the (I and less) may also be an example of long-distance landscape, invading most of the permanent birch transport from areas to the east, for instance Finland, forests and from time to time subjected to fo rest fires where the spruce at that time made considerable (Fig. 6). progress towards the west, in fact almost as far as to the Bothnian Sea of that time (Moe 1970, map p. 63; The reconstructions are artistically rendered by Tallantire 1972, map p. 2, 1980), i.e., only a few Forester Nils Forshed. hundred kilometers from the Central Swedish area in question. As pointed out above, pollen grains of plants of SUMMARY weed character (Artemisia, Chenopodium, Plantago, Rumex cf. acetosella) were fo und to an amount of a On the basis of pollen analysis the present study deals few per cent of the AP sum (including Cerealia) in the with the vegetational changes during Postglacial time uppermost fo ur samples, reflecting rural activity around Ragvaldsmossen, a small bog in a highland some kilometers from the sampling place. These area 50-250 m above sea level), ea. 3 km S of (I pollen types, together with Ericales, also occur Stjarnsund in SE. Dalarna (Dalecarlia). The district somewhat more frequently in the lowermost samples, in question lies in the broad transitional zone, "limes representing a stage when the fo rest vegetation norrlandicus" (Sernander ), between the southern probably had not yet totally covered all upland boreo-nemoral and the northern boreal zones (Sjors 1963 ; cf. map, Fig. 1. ground and shores. ) However, there are sc�ttered fm ds of the above­ The stratigraphy of the bog, originally a lake, is mentioned pollen grains even in the rest of the profile, shown schematically in the pollen diagram, Fig. 3. i.e. corresponding to the long Postglacial forest Note the change of vertical scale at level 7.25 m. The period. There were certainly some very restricted absolute chronology is tentatively added. In the text spots along lakes and streams in the neighbourhood, the pollen sequence is divided into the conventional where plants sensitive to competition may have periods, mainly according to the chronozones by grown. But it is uncertain whether this is true for the Mangerud et al. (1974 p. 122). types of plants in question. As in other pollen diagrams from this transitional On the other hand, we may consider that pollen zone, the curves in the diagram unveil an expansion grains of plants belonging to the genera in question, northwards through this zone of the warmth-

Acta Phytogeogr. Suec. 68 Ragvaldsmossen, a pollen-analytical study 79 demanding trees and hazel during the climatically Hedstrorn, H., 1893. Om hasselns fo rntida och nutida ut­ favourable period of Postglacial time and a cor­ bredning i Sverige. - Geol. For. Stockh. Forh. responding stepwise retreat leading to the present 15:29 1-320. state of rare occurrences, mainly in the southern part Hellman-Lutti, K., 1975. Myrar och vegetationshistoria. ­ Leksands sockenbeskrivning 6. Falun, pp. 205-224. of the zone. Hogbom, A.G., 1934. Om skogseldar fo rr och nu och The irregularity of the tree pollen curves in the up­ deras roll i skogarnas utvecklingshistoria. - Norrl. per part with Picea is emphasized, fo rming a contrast Handbibl. 13:1-101. to the more even course of curves during earlier Karta over landisens avsmaltning i Sverige i tre blad. times. This irregularity may reflectfo rest fires,which Mellersta bladet. - Sver. geol. Unders., Ser. Ba, 18 certainly have been more common afterthe invasion (1961). of the flammable spruce. Lundqvist, G., 1929: En fdrhistorisk paddel frc'm Dalarna. After a comparison with the results of other - Geol. For. Stockh. Forh. 51:367-381. pollen-analytical studies in the transitional zone, the 1951 : Beskrivning till jordartskarta over Kopparbergs problem of long-distance transport of pollen is dis­ Jan. - Sver. geol. Unders., Ser. Ca, 21: 1-213. 1963. Beskrivning till jordartskarta over Gavleborgs cussed in order to explain the occurrence in pre­ I an. - Sver. geol. Unders., Ser. Ca, 42: 1-181. Subatlantic times of scattered pollen grains of Picea Mangerud, J., Andersen, S.T., Berglund, B.E., and and of such plants which, at the top of the sequence, Donner, J., 1974. Quaternary stratigraphy of Norden, are characterized as weeds. The possibility of long­ a proposal for terminology and classification.- Boreas distance transport is emphasized more strongly than 3:109-128. usual, viz. dispersal of Pieea pollen from Finland, Moe, D., 1970. The Post-Glacial immigration of Picea (I) which was invaded by the spruce at the correspond­ abies into Fennoscandia. - Bot. Notiser 123:61-66. ing time (Subboreal period), and (2) dispersal of the Post, L. von, 1924. Ur de sydsvenska skogarnas regionala so-called weed pollen grains from the south and historia under postarktisk tid. - Geol. For. Stockh. south-eastern European vegetation, rich in species of Forh. 46:83-128. 1930. Norrlandska torvmossestudier. 2. Nagra huvud­ the genera in question. punkter i skogens och myrarnas postarktiska historia Three drawings illustrate tentatively the main inom sodra Norrland. - Geol. For. Stockh. Forh. stages of the vegetational development (Figs. 4-6). 52:63-90. Sj ors, H., 1948. Myrvegetation i Bergslagen. - Acta phytogeogr. suec. 21:1-299. REFERENCES 1963. Amphi-atlantic zonation, nemoral to arctic. - North Atlantic Biota and their History. Oxford, pp. Andersson, G., 1902. Hasseln i Sverige fordom och nu. - 109-125. Sver. geol. Unders., Ser. Ca, 3: 1-168. Tallantire, P.A., 1972. The regional spread of spruce Berglund, B., I 968. Vegetationsutvecklingen i Norden (Picea abies (L.) Karst.) within Fennoscandia: a efter istiden. - Sver. Nat. Arsb. 1968 :3 1-52. reassessment. - Norw. J. Bot. 19:1-16. Florin, S., 1944. Havsstrandens forskjutningar och 1974. The palaeohistory of the grey alder (A lnus incana bebyggelseutvecklingen i ostra Mellansverige under (L.) Moench) and black alder (A . glutinosa (K.) senkvartar tid. Allman oversikt. - Geol. For. Stockh. Gaertn.) in Fennoscandia. New Phytol. Forh. 66:551-634. 73:529-546. Fransson, S., 1965. The borderland. - Acta phytogeogr. 1980. The postglacial immigration and spread of the suec. 50: 167-1 75. spruce (Picea abies (L.) Karst.) in Fennoscandia. - Fries, M., 1956. Bensjoomradet i skogshistorisk belysning. Thesis Dept. Ecol. Bot., Umea. En pollenanalytisk undersokning fran sodra J amtland. Wright, H.E., 1974. Landscape development, forest fires, - Norrlands skogsvardsfdrb. Tidskr. 1956:527-559. and wilderness management. - Science, N.Y. 1965a. The Late-Quaternary vegetation of Sweden. - 186 :487-495. Acta phytogeogr. suec. 50:269-280. Zackrisson, 0., 1977. Influenceof fo rest fires on the North 1965b. Outlines of the Late-Glacial and Postglacial Swedish boreal fo rest. - Okios 29:22-32. vegetation a! and climatic history of Sweden, illustrated 1978. Skogsvegetationen vid stranden av Storvindeln by three generalized pollen diagrams. - Spec. Pap. under 200 ar. - Svensk bot. Tidskr. 72:205-226. geol. Soc. Am. 84:5 5-64. 1981. Forest fire frequency and vegetation pattern in Hedin, K., 1975. Daterad lagerfOijd vid Spjutmo. - the Vindelalven valley, N. Sweden during the past 600 Leksands sockenbeskrivning 6. Falun, pp. 191-204. years. - Acta Univ. Ouluensis, Ser. A. (In press.)

Magnus Fries, Swedish Museum of Natural History, Dept. of Botany, S- 104 05 Stockholm 50, Sweden or Ynglingavagen 5 B, S-182 62 Dj ursholm, Sweden

A eta Phytogeogr. Suec. 68 Fig. 4. An attempt to illustrate the first,early Postglacial stage (Preboreal) in the studied area (-7000 B.C.). Pioneer birch dominated. Locally there probably was some open ground vegetation. Some small stands of Hippophae· occurred in open spaces and perhaps along the shore.

A eta Phytogeogr. Suec. 68 Fig. 5. An attempt to illustrate the optimal stage of the warm period forest (ea. 5000-3000 B.C.), which was composed, apart fr om the dominating pine and birch woods distributed according to the soil conditions, of scattered stands of warmth-demanding trees (oak, elm, lime, ash) and hazel in the most favourable places. A eta Phytogeogr. Suec. 68 Fig. 6. An attempt to illustrate the late, sub-recent forest situation (Subatlantic, ea. 500 B.C. to present time) characterized by pine and spruce fo rests, distributed according to the soil conditions and the ravagings of fires.To the left in the picture a fire field occupied by birch thicket, after some time replaced again by a spruce generation.

Acta Phyrogeogr. Suec. 68 A Comparison between the Alpine Plant Communities of Alaska and Scandinavia Olav Gj rerevoll

Comprehensive and detailed sociological in­ dinavia, Alaska borders on the Arctic Ocean in the vestigations of the alpine plant communities have north and on a warmer ocean in the west. primarily been carried out in Central Europe and Alaska is a large country, about 1.6 mill. km2• The Scandinavia. At times the Central-European school differencesbetween oceanic and continental areas are and the Scandinavian school have had extensive dis­ even more accentuated than in Scandinavia. At Fort agreements, both about the concept of sociological Yukon a maximum of +38°C and a minimum of criteria and about the methods used. Nevertheless, in -61 °C is reported. In July the average maximum spite of these disagreements it has been possible to temperature is +25°C, in January -33°C. The cor­ draw parallels between the alpine plant communities responding data fo r Seward are + 17°C and -9°C. in the two areas (Du Rietz 1924; Nordhagen 1936). My work has been mainly carried out in the cen­ Because travel between the two regions is not dif­ tral parts of Alaska and the adjacent area of the ficult, it has been comparatively easy for botanists to Yukon (Ogilvie Mts). Between the Yukon River and obtain personal experience in both areas. its tributary the Tanana River there is a mountain The Scandinavian and the Central-European range that might be considered as the northernmost mountains have many species in common, and many part of the Rocky Mts. The highest peaks rise to an of them are sociologically important, but on the other elevation of about 1800 m. The geological variation is hand the number of species is much higher in the great. Alps. Consequently, comparisons are difficult.In ad­ The most dominating mountains of Central dition, there are also some great ecological dif­ Alaska are the Alaska Range with very high peaks, fe rences, e.g., light and temperature conditions. i.e., Mt McKinley, 6200 m. The Alaska Range has a In the northern areas the ecological conditions are great climatic influence. The north- and east-facing more equal. Most of the species occurring in the slopes, being situated in the precipitation shadow, are Scandinavian mountains are circumpolar; this dry. The south- and west-facing slopes, on the other applies to a high degree to many of the common and hand, are strongly influenced by oceanic weather predominating species. Accordingly, one might in ad­ conditions. This applies also to the Kenai Range and vance expect a considerable conformity in the struc­ the Chugach Mts. ture of the plant communities. Investigations in I have also carried out investigations in the moun­ Greenland conftrm this assumption. tains of the Seward Peninsula, but not in the Brooks In the vast northern areas the botanical field in­ Range. vestigations have been mainly focussed on the Some phytosociological investigations have been floristic phytogeographical problems. Thanks to the made by Spetzman (1959) and Hanson (1953). Apart monumental contributions of Hulten these problems from one paper of Hanson, Scandinavian methods are very well outlined. have not been used. Parallels may be drawn without When it comes to plant communities, some using the same methods but, on the other hand, the material covering the Siberian mountains has been comparison obviously will be easier and better if the published, but is not well suited for comparisons with analyses are carried out with the same methods Scandinavian investigations. and the same concept of sociological units and For quite natural reasons, especially from a Scan­ classification. In this paper only a survey will be dinavian viewpoint, Alaska is of great interest. Most given without sociological tables. The analyses have of Alaska is situated on the same latitude as Scan­ been carried out in accordance with traditional Scan­ dinavia. Point Barrow is just as far north as the dina vi an methods. North Cape. Seward in the Kenai Peninsula is on the The very first look at Alaskan alpine plant com­ same latitude as Stockholm and Oslo. Like Scan- munities gave the impression of great conformity

A eta Phytogeogr. Suec. 68 84 Olav Gjrerevoll

Fig. 1. Pieea glauca fo rming the timber line in Alaska Range.

with the Scandinavian ones. This applies especially to chionophobous to chionophilous ones. Correspond­ the continental mountains. The situation is more ingly, two parallel meadow series are encountered, complicated in the oceanic mountains where Pacific also in Alaska with more diversified and complicated species play an important part in forming the plant units than in the heath series. communities. There is an obvious difference between the con­ In the same way as in Scandinavia, the Alaskan tinental and oceanic mountains. The continental mountains are partly built up of acid and partly of mountains are characterized by communities of the more or less calcareous soils. Consequently, the same heath series; quantitatively the meadows are of less main classification in two series-one poor in importance. In the south- and west-facing slopes of calciphiles and one rich in calciphiles-may be used. the Alaska Range vast areas of meadow vegetation Furthermore-also in the same way as Scan­ are found. This situation is, of course, related to the in dinavia-the distribution of snow is a most important amount of snow and the progress of the thaw. In the ecological factor. The landscape consists of early continental areas much of the snow disappears by snow-free areas characterized by chionophobous sublimation. Consequently, no meadow vegetation species, and protected areas and depressions with develops below the snow fields, or a mixture solid snow-cover of different duration, characterized composition-community of heath and meadow by chionophilous species. species is fo und. When one combines the consequences of the un­ In the areas of oceanic influencethe melting of the even snow distribution with those of the soil snow gives much water, especially because the variations mentioned above, the result is the main melting oftentakes place during rainy weather. Large classification model of Scandinavia (Gjrerevoll & areas are accordingly irrigated for a long time and Bringer 1965): Two parallel heath series, one with even during the whole season. and one without calciphilous species, both com­ In Alaska one more ecological factor, not met with prising communities with a gradient ranging from in Scandinavia, should be added. Apart from the

A eta Phytogeogr. Suec. 68 Alpine plant communities of Alaska and Scandinavia 85

southern coastal areas permafrost is present in all and Diapensia lapponica subsp. obovata. This Alaskan mountains. This means, i.a., that in warm amphi-Beringian race of D. lapponica displays the and dry periods water will be transported to the same ecological behaviour as the Scandinavian plant. plants from the thawing layer. Hygrophilous species In places with strong wind erosion Salix may therefore have surprising occurrences in ex­ phlebophylla may grow abundantly. Hierochloe posed places. During the Pleistocene glaciation a alpina is one of the preferential species. large refugium existed in Alaska-Yukon. Only local So far so good. The analyses from these com­ glaciers occurred in the mountains between the munities, however, show a constant occurrence of Tanana R. and the Yukon R. and in the Central Dryas octopetala which, in Scandinavia, is the most Yukon. This refugium was of greatest importance exclusive species of communities rich in calciphiles. from a phytogeographical viewpoint, consequently it In Alaska it avoids the most strongly wind-eroded means not only old species, but also old vegetation. ridges, but the normal situation is to find com­ The subalpine birch fo rest so typical in Scan­ munities with Loiseleuria procumbens and Dryas oc­ dinavia has no counterpart in Alaska. The timber-line topetala as eo-dominants on non-calcareous soils is first and fo remost fo rmed by Picea glauca. The with pH 4-5. elevation of the timber-line varies from south to north From a taxonomical viewpoint, Dryas in Alaska and fr om the continental to the oceanic areas, as in offers intricate problems, and among specialists Scandinavia. In the continental districts Picea glauca opinions diffe r as to the concept of taxa. The Dryas may ascend to about 1000 m. of the heaths is octopetala, especially f. argentea In the coastal mountains, i.a., in the Kenai Range, which gives the heaths a silvery-gray colour. the timber-line is formed by Picea sitchensis and The fact is that in Alaska there are two different Ts uga mertensiana, normally at an elevation of Dryas octopetala -communities, one poor and one 500-700 m. Ts uga mertensiana grows at higher rich in calciphiles. Perhaps there is a historical ex­ altitudes than any other coniferous species. planation of this wider ecological amplitude of the In the interior of Alaska thickets of A lnus crispa Alaskan Dryas octopetala. are found above the timber-line. In the coastal moun­ In the uppermost part of the lee side with sufficient tains A lnus sinuata may play an important and un­ snow cover during the winter, but early exposure, pleasant role fo llowing the spruce forest from the Betula nana occurs as the predominating species ac­ lowland and exceeding the timber-line, very often companied by, i.a., Loiseleuria procumbens, Empe­ fo rming impenetrable thickets. trum hermaphroditum, Vaccinium uliginosum, and Let me also mention that in localities with a good Ledum decumbens. supply of moisture, well developed Salix-thickets In Scandinavia the next step down the lee side may be found in the uppermost part of the spruce would be a Vaccinium myrtillus -community, but V. fo rest and just above the timber-line-as in Scan­ myrtillus does not occur in Alaska. As in Northern dinavia. The most important species are Sa/ix Scandinavia, Cassiope tetragona occupies the cor­ glauca, S. hastata, S. commutata (in coastal moun­ responding zone. C. tetragona may produce a very tains), S. lanata richardsonii, S. barrattiana, S. acid humus. pulchra, and S. arbusculoides. In the interior of Alaska and the Yukon as well as in the mountains of the Seward Peninsula, Dryas alaskensis, endemic to Alaska-Yukon, may predominate on the lee sides. The community has a Heath communities considerable chionophilous character. As regards soil Communities poor in calciphiles (A) conditions, D. alaskensis seems to have the same wide amplitude as Cassiope tetragona. The com­ Chionophobous communities munity is found on calcareous soils in the Ogilvie The most exposed areas are characterized by species Mts. and on granite mountains on the Seward Penin­ and plant communities very similar to those fo und in sula. Scandinavia. The most striking contrast is the subor­ In the coastal mountains (Kenai Range), dinate role of lichens. Many lichens may be present, Phyllodoce glanduliflora predominates in the cor­ but I have not seen heaths completely predominated responding zone; it displays the same ecological by Cladonia-species as in the continental mountains amplitude as P. caerulea in Scandinavia. P. caerulea of Scandinavia. is one of the preferential species of the Vaccinium In the exposed communities the dominants are myrtillus -communities. Loiseleuria procumbens, Empetrum hermaphro­ Furthermore, in dryer conditions on the Seward ditum, Vaccinium vitis-idaea, Arctostaphylos alpina, Peninsula a most magnificent community is en-

A eta Phytogeogr. Suec. 68 86 Olav Gj cerevoll

Fig. 2. Phyllodoce glanduliflora ­ community in Kenai Range.

countered, with Rhododendron camtschaticum pre­ only vascular plant with its leaves resting on the dominating. ground. Where the extreme snow-beds are stony and gravelly Geum rossii may at least visually become the Chionophilous communities dominant species. Its ecological amplitude is very The most common chionophilous community of the wide as it is indifferent to soil conditions. interior mountains is characterized by Carex In the oceanically influenced mountains the situa­ bigelo wii (C. consimilis). It is mainly a community tion is considerably different. The Phyllodoce glan­ poor in species, as also in Scandinavia. duliflora zone is as a rule succeeded by a distinct When the snow remains so long that the vitality of zone of predominating Cassiope stelleriana. This Carex bigelowii is greatly reduced, Salix polaris species is the ecological counterpart of C. hypnoides subsp. pseudopolaris more and more takes over the in Scandinavia. The C. stelleriana snow-beds cover dominance. The corresponding community in Scan­ large areas in the western and southern mountains of dinavia is dominated by S. herbacea, which does not Alaska. Salixpolaris is always a constant and impor­ occur in Alaska. Salix polaris pseudopolaris seems tant companion. to be indifferent as regards soil conditions. The com­ Another common companion and at least partly munity is very poor in species. S. polaris may be the still more chionophilous is Luetkea pectinata. The

Fig. 3. Geum rossii predominat­ ing in a snow-bed on Mt Harper, Central Alaska.

A cl a Phy10geogr. Suec. 68 Alpine plant communities of Alaska and Scandinavia 87

Luetkea pectinata dominated areas are large giving Meadows the landscape a dark colour easily recognized at a distance. Below the Luetkea pectinata zone extreme In the same way as in Scandinavia, it is more difficult snow-beds are commonly encountered, predo­ to classify the meadow communities, especially those minated by Carex pyrenaica micropoda. Both com­ rich in calciphiles. Throughout Alaska there are early munities are rich in mosses, but otherwise are usually snow-free communities predominated by Festuca poor in species. As in Scandinavia, in the most ex­ altaica. This vigorous species forms dense tussocks treme snow-bed of the heath series, Polytrichum nor­ leaving almost no place for other species. F. altaica vegicum predominates. seems to have a wide ecological amplitude, but without doubt the species produces an acid humus. The pH-values of all soil samples are close to 5.0. Ac­ cordingly, the Festuca altaica -communities should Communities rich calciphiles (B) in be regarded as acidophilous. Calcareous bedrocks are common in the interior With a better supply of water and somewhat later mountains of Alaska and the Yukon. This applies to exposure Calamagrostis langsdorffiibe comes predo­ calcareous, easily disintegrated shists as well as minating. Vast areas, especially in the oceanic moun­ mountains built up by limestone. tains, are covered by this species. C. langsdorffii, On the most wind-eroded ridges where the vitality in the same way as Festuca altaica, produces an of Dryas octopetala is reduced, the predominating acid humus. The annual production is very high. species are Carex glacialis, C. nardina, C. rupestris, Every fall a thick layer of litter is left, rendering it and Kobresia myosuroides. almost impossible fo r cryptogams to grow. Nor is it First and fo remost, however, Dryas octopetala possible fo r other phanerogams to compete. Some predominates in this xerophilous plant community. may occur in places where water from snow fields As in Scandinavia, it is a community very rich in causes erosion furrows: Festuca altaica, Mertensia species, many of them exclusive ones. Sociologically paniculata, Chamaenerion angustifolium, Heracle­ important species are, fo r example, Carex rupestris, um lanatum, Sanguisorba sitchensis, and Polygonom C. scirpoidea (which is a species with very wide alaskanum. ecological amplitude), Androsace chamaejasme, In the more extreme snow-beds C. langsdorffii is Oxytropis maydelliana, Woodsia glabella, Senecio not present. The predominating species here are atropurpureus, and Rhododendron lapponicum, the Carex lachenalii, C. py renaica micropoda, Oxyria latter often occurring as a eo-dominant. digyna, Sibbaldia procumbens, and Claytonia On the limestone mountains in the interior of sarmentosa. Important followers are Saxifraga Alaska (White Mts.), Kobresia simpliciuscula may nivalis, Cerastium cerastioides, Ranunculus pyg­ occur as a predominating species (Gjrerevoll 1954). maeus, Koenigia islandica, Saxifraga fo liolosa, S. From a Scandinavian viewpoint this is a great sur­ fe rruginea (oceanic mountains), and Salix polaris. prise as K. simpliciuscula is a species of shallow rich In the oceanic mountains very wet snow-beds are fe ns. In the Ogilvie Mts. it was found in rich fe ns, but met with, partly of spring character. Predominating was also very abundant in exposed areas of the dry and important species are: Saxifraga lya llii, Lep­ limestone mountains. tarrhena py rolifolia, Petasites fr igidus, Caltha lep­ A number of rare and partly endemic species tosepala, and Poa paucispicula. belong to the exclusive ones of this Alaskan parallel When the soil conditions are better, gorgeous to Kobresieto-Dryadion (Nordhagen 1943) of Scan­ meadow communities develop, especially in the low­ dinavia, i.a., Campanula aurita, Douglasia gormanii, alpine zone. They are very rich in species with shift­ Erigeron hyperboreus, Oxytropis jordalii, Phlox ing quantitative importance. sibirica, Eritrichium splendens, Lesquerella arctica, In the interior of Alaska and the Yukon there is a Smelowskia calycina, Papaver walpolei (Seward multicoloured array of Aconitum delphinifolium, Peninsula), Anemone drummondii, and Castillej a Myosotis alpestris, Polemonium acutiflorum, Del­ hyperborea. phinium glaucum, Lupinus arcticus, A nemo ne nar­ With better snow protection the vitality of Dryas is cissiflora, and Va leriana capitata. This community re.duced and normally Cassiope tetragona becomes also covers the upper part of the subalpine zone. predominating, often followed by a zone of Salix It may be regarded as a counterpart of the Scan­ reticulata. The number of species is high with many dinavian "Aconition septentrionalis" (Nordhagen calciphilous species not found on acid soils. In the 1936) with i.a., Aconitum, Lactuca alpina, and snow-beds Salix polaris subsp. pseudopolaris My osotis decumbens. predominates. In the interior of Alaska and on the Seward Penin-

Acta Phytogeogr. Suec. 68 88 Olav Gj cerevoll sula the peculiar Boykinia richardsonii predo­ dinavia. In the present context I have decided not to minates, fo llowed by Dodecatheon fr igidum, Saxi­ give the Alaskan communities comparative alliance fr aga puncta/a, S. exilis, and Ranunculus nivalis. names. I could have introduced some preliminary In the oceanic mountains more species are added names but instead I have, in the same way as in the (whereas B. richardsonii is absent): Geranium text, used the neutral label "community" in connec­ erianthum, Aquilegiaformosa, Veratrum eschscholt­ tion with characteristic species. Only when more zii, Swertia perennis, Castillej a unalaschkensis, material is available will it be time to decide the most Erigeron peregrinus, Senecio triangularis, Fritillaria suitable sociological nomenclature. camtchatica, Geum calthifolium, Phleum com­ mutatum, several tall Arnica -species, and oftendense stands of Valeriana capitata and Sanguisorba sitchensis. REFERENCES In snow-beds exposed later on calcareous soils the Du Rietz, G.E. 1924. Studien iiber die Vegetation der sociological picture is very similar to Scandinavian Alpen, mit derjenigen Skandinaviens verglichen. conditions with Ranunculus nivalis, R. sulphureus, Veroff. geobot. Inst. Zurich. 1:31-138. and R. pygmaeus as important species together with Gjrerevoll, 0. 1954. Kobresieto-Dryadion in Alaska. Cardamine purpurea, Saxifraga bracteata, and in Nytt Mag. Bot. 3:51-54. the coastal mountains also Ranunculus eschscholtzii, Gj rerevoll,0. & Bringer, K. 1965. Plant cover ofthe alpine Draba crassifolia, and Phippsia algida. regions. - Acta phytogeogr. suec. 50:257-268. Hanson, H.C. 1953. Vegetation types in northwestern Alaska and comparison with communities in other Arctic regions. - Ecology 34: 111-140. SUMMARY Nordhagen, R. 1936. Versuch einer neuen Einteilung der subalpinen-alpinen Vegetation Norwegens. - Bergens In the table below I have tried to give a concentrated Mus. Arb. 1935, Naturv. rekke, 8:1-88. survey of the comparison between the main alpine Spetzman, L.A. 1959. Vegetation of the Arctic slope of plant communities of Alaska-Yukon and Scan- Alaska. - U.S. Geol. Surv. Prof. Pap. 302:1-58.

Table I. Survey of the ma in alpine heath and mea dow commun ities of Alaska compared wi th the corresponding ones of Scandina via. A. Communities poor in calciphiles . B. Communities rich in calciphiles. A.

heath series meadow seri es snow conditions Scandinavia Al aska Scandinavia Alaska wind-exposed, Empetrion Loiseleuria- almost snow free ( Loi se l euri eta- Dryas-com . Arctostaphyl ion) protected , but My rti ll ion Phyl l odoce ea rly snow free glandul i flora-cam . late snow free Des champs i o- Carex bigel owi i- Ranuncu l o- Ca l amagros t is Anthoxantion corn . Anthoxantion l angsdorffi i -corn. very late Herbaceon Sa lix polaris- Ste llari - Oxyria digyna­ snow free corn. Oxyrion Saxi fraga lyall i i -cam .

B.

1� i nd-exposed, Dryad ion Dryas octopetala. almost snow free corn . protected , but Cass i ope Cassiope Potent illeta- Valeri ana early snow free tetragona-com . tetragona-com. Polygonion vivipari capi tata-com . late snow free Reticulato- Salix reticulata Ranuncu l o- Boyk inia Poi on alpinae Poa paucispicula- Poion alpinae richardsonii- corn. cam . very late Po l ari on Sal i x polaris- Oppositifolio- Ranunculus snow free corn . Oxyri on ni val i s-com .

Olav Gj rerevoll, DKNVS Museum, Trondheim University, N-7000 Trondheim, Norway

Acta Phytogeogr. Suec. 68 Attached Algal Vegetation in some Streams from the N arssaq Area, South Greenland Catarina Johansson

Algal investigations were carried out in eight streams in the The mountains around the head of the N arssaq area, and also in the surroundings of the Kangerdluarssuk Fiord are in the northwest corner Kangerdluarssuk Fiord during the summer of 1977 (Fig. intersected by the boundary line between bedrock 1), in connection with the Narssaq project under the granite overlain by sandstone and dolerite sheets to leadership of Prof. Chr. Overgaard Nielsen, Dr. I. Rose­ the west and the rocks of the Ilimaussaq intrusion to Hansen and Prof. Henning S0rensen. The area has been the east. The boundary zone is developed as a gorge thoroughly investigated by the Geological Survey of Greenland (GGU) both from geological and ecological cut into syenite of a rusty brown colour. The vegeta­ points of view (see Rose-Hansen et al. 1977). tion here is very sparse. Earlier freshwater algal studies in Greenland have been The Lil/e elv valley system from the source of the performed by, among others, Foged 1953, Nygaard 1954, river in Bukses0 (alt. 280 m above m.s.l.) to the Bocher 1950 and Bachmann 1921, whose sampling mouth of the river at the head of the fj ord is entirely localities were in lakes, ponds and streams. Most of the surrounded by the rapidly weathering naujaite moun­ material in these works was collected in connection with tains and practically without any cover of vege­ Danish and Swedish expeditions to Greenland. tation. Only in the valley around Bukses0, and in pat­ ches along the river alley, are there areas of uncon­ Physiography of the three areas studied solidated rocks covered by vegetation. The Narssaq river is fe d by meltwater from the N arssaq glacier at the head of the valley and by numerous tributaries of which the most important ones are the Taseq stream draining the big Lake Taseq to the south of the valley and the Kvane stream draining the K vanefjeld plateau to the north of the valley. The glacier is located between the Ilimaussaq and Nakalaq mountains. The former is made up of lavas with lujavrite cropping out below the talus covering the foot of this mountain. The Ilimaussaq intrusion is one of the intrusions of the Gardar igneous province which forms a belt across South Greenland. The intrusions are dominated by syenites while gabbro and granite are of subordinate importance. The kakartokite and the syenite mountains are practically devoid of vegeta­ tion. Four of the studied streams are situated in the Ilimaussaq intrusion (Grense elv, Lille elv, Lakse elv, "Kringlerne elv"). The lower part of the N arssaq river and K vane elv flows across an alluvial plain composed of fluvio-glacial deposits, in parts as terraces. The slopes of the valley support vegetation, partly in the fo rm of overgrown tal us, partly vegeta­ tion growing on the numerous alluvial terraces and in Fig. I. Map of the Narssaq area in SW Greenland where the small gorges cut into these by tributaries. This numbers indicate the eight streams investigated in the area is called the Dyrnaes region and here also one of summer of 1977. l. Narssaq river, 2. Kvane stream, 3. the GGU's camps is situated; the other one used is Talut stream, 4. Reference Ill stream, 5. Grense stream, 6. located in the Kangerdluarssuk area. Lille stream, 7. Lakse stream, 8. Kringlerne stream.

Acta Phytogeogr. Suec. 68 90 Catarina Johansson

Table I. A surrrnary of the qeology of the three areas investigated 2a) which is common in cold oligotrophic rivers but ------·------·------1977. can also be present in more or less eutrophic waters. s a �22.q_ _ _r� Na rs saq river '"'];'""'" The stream is about seven metres wide and, at the (1} Narssaq Kvane stream �lkaligrani te - sampling occasion, had a mean depth of about 20 (2} 1 intrus ion Ta lut stream ( 3} alkaligrani te _ cm. There is no shading vegetation here. The stones

Ref. I I I area_ were predominantly covered with the diatoms Ref . I I I stream ( 4} granite Ceratoneis arcus v. linearis, C. arcus (Fig. 2e) and Diatoma hiemale v. mesodon (Fig. 2h), the latter Kangerdluarssuq urea preferring running waters in the mountains. Another Grense stream (5) augite syenite and foya ite Lillt> stream najai te stream in this delta had a vegetation of partly ( 6 } !l imaussaq Lakse stream superfi cial depos its ( 7 } intrusion Hy drurus fo etidus and partly Zygnema b (the letters Kringlerne stream (8} kakortokite after the Zygnemal species refer to the classification by Israelson 1949) (Fig. 2b) and a dense vegetation of the moss Amblystegium riparium. Here the predomi­ The bed of Lakse elv runs along the southeast side nant diatom was Diatoma hiemale v. mesodon and to of the lujavrite ridge. The north side of the river bed a lesser extent Ceratoneis arcus and also Meridion consists of lujavrite in the lower parts and in the up­ circulare and Cymbella ventricosa (Fig. 2d). The per reaches kakortokite. The river bed is strewn with character of the vegetation of the rapids near the 'big boulders. The kakortokite plateau termed camp was different. The macroalgae were composed Kringlerne is practically devoid of vegetation and of Zygnema a, To lypothrix distorta v. penicillata, through this area runs the "Kringlerne elv" from Mougeotia c, and Oedogonium sp. The dominant which algal samples were taken. diatoms were Anomoeoneis brachysira v. thermalis Reference area stream The Ill has its source ap­ and Eunotia pectinalis v. minor (Fig. 2f) and to a proximately 350 m above m.s.l. and has a tributary lesser extent Achnanthes kryophila, which prefers from Lake 9 5 (at a height of 95 m). The firstpart runs colder areas. This area could be characterized as an on lujavrite material and then on a granite slope with alpine heath with low shrubs of Salix glauca ssp. a fall of about 3 metres into a deep gully which it callicarpaea, flowering herbs such as Pinguicula fo llows to the sea. Here both sides of the river bed are vulgaris, Polygonum viviparum and Silene acaulis, covered by a thin layer of talus with a dense vegeta­ and a grass vegetation of Anthoxanthum odoratum v. tion. alpina. At the edges of the streams were scattered The climate of the Ilimaussaq area is decided by cushions of Thalictrum alpinum. the fact that it is situated in the low pressure system The water temperature varied between 6 and extending from Hudson Bay to the Atlantic Ocean 8.5°C on the 16th of June. east of Iceland. Continental polar air masses and In a canyon with fast-flowingwater at an elevation oceanic polar winds are responsible fo r the weather. of 170 m samples were taken of a very sparse algal The yearly average temperature is 1.5°C at Narssaq. vegetation of diatoms mainly consisting of Diatoma The average temperatures in January and July are hiemale v. mesodon, Tabellaria flocculosa, Cera­ respectively -3.3°C and + 7.5°C. The yearly toneis arcus, Meridian circulare, Fragilaria capu­ precipitation varied in the period 1966-73 from cina v. lanceolata, Cy mbella ventricosa and a green about 1000 mm to 2500 mm. Strong gales are fre­ alga Microspora palustris v. minor. A bit higher up, quent, partly in the form of fohn gales from the east at 330 m, the stream bottom is made up of boulders or southeast. The warm, dry fOhnwinds with relative­ and gravel. Here there was also a sparse algal vegeta­ ly low humidity may increase the temperature by tion of diatoms similar to that mentioned at 170 m. In 15-20°C in the course of less than one hour (Rose­ pools beside the river a thick green vegetation of Hansen et al. 1977). Microspora palustris v. minor existed and, further­ more, diatoms such as Pinnularia borealis and Achnanthes linearis were found, the former being es­ The algal vegetation pecially frequent in mountain streams. The Kvane elv (Fig. 4) is about 2 km long. It begins The Narssaq area at 510 m above m.s.l. and runs into the N arssaq river The lower part of the Narssaq river (Fig. 3) is divided at 80 m. Most of the stones studied in the locality just into 4 or 5 smaller channels because of the flat floor before the flow into the Narssaq river were covered of the valley just before the mouth. The main stream by the water mosses Hygrophypnum ochraceum, and some of the smaller channels had a dense vegeta­ Schistidium alpicola v. rivulare and Scapania ul­ tion of the chrysophycean Hydrurus fo etidus (Fig. iginosa. The algal vegetation was sparse, probably

A eta Phytogeogr. Suec. 68 Attached algal vegetation in some streams from the Narssaq area, Greenland 91

10 �m

10 �m

e f k

Fig. 2. (a) Hydrurus fo etidus, (b) Zygnema b, (c) Cymbella gracilis, (d) Cy mbella venlricosa, (e) Ceratoneis arcus, (f)

Eunolia peclinalis v. minor, (g) Meridian circulare v. constricta, (h) Diatoma hiemale v. mesodon, (i) Eunolia lridentula v. perminuta, (k) Eunotia bigibba v. pumila. because of the competition by the mosses. There were diatom Ta bellariaflocculosa which gave the stones a but few colonies of algae on the mosses and the slippery surface. Other diatoms found were stones. Threads of Microspora palustris v. minor, Ceratoneis arcus, Achnanthes minutissima v. cryp­ Phormidium autumnale, Zygnema c and Oscillatoria tocephala, usually common in all types of habitats, spp. could be seen in places. The diatoms were, Pinnularia divergens, most frequent in alpine regions, among others, Diatoma hiemale v. mesodon, Meri­ Cymbella delicatula, most common on over-flowed dian circulare, and the v. constricta (Fig. 2g), cliffs and this also applies to Frustulia rhomboides v. Eunotia pectinalis, and the v. minor, Pinnularia in­ saxonica. All these characteristics apply very well to terrupta and Cy mbella ventricosa. The surrounding the conditions in Greenland. vegetation could be characterized as a dwarf-shrub heath (Rose-Hansen et al. 1977), with, e.g., Salix The Reference area glauca ssp. callicarpaea, Anthoxanthum odoratum Ill ssp. alpinum, Veronica alpina, Luzula multiflora, The Reference Ill stream (Fig. 5) in the outlet of the Phleum commutatum and Rhinanthus minor. Lake 95 m (in the area of Reference in the Ill Angelica archangelica ssp. norvegica, after which the Narssaq project 1977) flows rather slowly the first brook has been named, grew along the watercourse. 100 metres and then there are a few waterfalls before The Ta lut stream, another of the Narssaq river it discharges into a bay of the Kangerdluarssuq tributaries, is a rather small stream a little less than 2 Fiord. The surrounding vegetation along the south­ km long and comes from the slope of the Tal ut moun­ facing slope is exceptionally abundant in meadow tain at 790 m above m.s.l. The ground consists of plants such as Ranunculus acris, Chamaenerion ba altic rock with manganese oxide along which the angustifolium and low Betula pubescens and rock cracks to pieces. The surrounding vegetation is Juniperus communis. Just before the stream empties an alpine heath mostly with grass vegetation. The into the bay there are large numbers of Elymus algal growth is dominated by Tolypothrixdistorta v. arenarius and Lathyrus japonicus along its banks. penicillata, small patches of Phormidium autumnale, The algal vegetation was also abundant. The upper Ph . valderiae and Zygnema b. More frequent was the reaches had a flora of Bulbochaete sp. (sterile),

Acta Phytogeogr. Suec. 68 92 Catarina Johansson

Fig. 3. Narssaq river. Fig. 4. K vane stream.

Stigonema mamillosurrz, To /ypothrix distorta v. The Kangerdluarssuq area penicillata growing in a dense vegetation of diatoms such as Ta bel/ariafloccu/osa, and also such algae as The Grense stream runs through an area with very Gomphonema angustatum, Cy mbella gracilis (Fig. few vascular plants. Among the boulders could be 2c) and Eunotia pectinalis. Further down, just before found a fe w individuals of, e.g., Chamaenerion a fall, the green algae Microspora palustris v. minor altifolium, Saxifraga aizoides and low Sa/ix g/auca and Chamaesiphon confervicola together with ssp. cal/icarpaea. The algal vegetation were mostly U/othrix aequalis were most predominant. A number composed of diatoms such as Achnanthes of diatoms such as Ta bellaria flocculosa, Fragilaria minutissima v. cryptocepha/a, Ceratoneis arcus v. intermedia and F. pinnata were also frequent. linearis and Ta bel/aria flocculosa. Zygnema a and In one of the small rapids with high velocity there To /ypothrix distorta v. penicillata occurred in places. was a dense growth of Hormidium rivulare. This alga The very few algae found in the Kringlerne stream is characteristic of habitats with rapid currents. Here were To lypothrix distorta v. penicillata and a few there were also many diatoms: Eunotia bigibba v. diatoms, Anomoeoneis brachysira v. cryptocephala pumila (Fig. 2k), Tabellaria flocculosa, Pinnularia and Ceratoneis arcus v. linearis. The surrounding appendiculata, Eunotia arcus, Achnanthes minutis­ vegetation was nearly devoid of vascular plants. The sima v. cryptocephala and Gomphonema longiceps v. samples were taken just below a waterfall and subclavata. patches of the moss B lindia acuta were found along One of the small rills which discharge into the the stream from the waterfall and down to the fiord. Lake 9 5 m had a very exceptional algal flora of The Lille stream (Fig. 6) had mostly a very diverse almost exclusively Microspora palustris v. minor and algal vegetation, whereas the surroundings had only a the diatom Eunotia tridentula v. perminuta (Fig. 2i). few vascular plants such as Betula glandulosa, Salix

A cla Phy10geogr. Suec. 68 Attached algal vegetation in some streams fr om the Narssaq area, Greenland 93

Fig. 5. Reference Ill stream. Fig. 6. Lille stream. glauca ssp. callicarpaea and Oxyria digyna. In a toneis arcus. A bit higher up along the stream, ea. i rapid part of the brook a dense vegetation of diatoms km, the algal vegetation was about the same. was found, consisting of Frustulia rhomboides v. sax­ Small rills ending in the Lakse stream had another onica, Ta bellaria flocculosa, Ceratoneis arcus v. algal vegetation: A blue-green algae, Schizothrix linearis and Eunotia pectinalis, and somewhat fewer fu scescens dominated together with many diatoms individuals were fo und of Eunotia monodon v. maior such as Ceratoneis arcus v. linearis, Ta bellariafloc­ f. bidens and Diatoma hiemale v. mesodon. Green culosa, Achnanthes minutissima v. cryptocephala, tufts on stones near the stream edge consisted of Sy nedra ulna, Anomoeoneis brachysira v. thermalis, species such as Tolypothrix distorta v. penicillata Nitzschia dissipata, Gomphonema intricatum v. sub­ and Zygnema a. clavata, Frustulia rhomboides v. saxonica f. capitata Overflowed rocks also had a very dense algal and Eunotia diodon. vegetation of about the same qualitative composition For a complete list of algae found in the 8 streams, as above but with a quantitative change of the algae. see Table Ill. Tolypothrix distorta v. penicillata together with Zygnema b were dominants and the diatoms a little less conspicuous. Comparison between the streams studied The Lakse stream runs across ground composed In order to investigate the similarity of the streams of superficial deposits with a valley floor of augite with regard to the algal composition, a Spearman syenite. The vegetation forms a Salix heath with Jun­ rank correlation coefficient between the ranked cus trifidus and a thick carpet of mosses and lichens. species in each stream has been calculated (Table 11). About 200 metres before the outlet into the On account of the very frequent ties in the present Kangerdluarssuq Fiord a dense algal vegetation of data of relative abundance, the Spearman rank cor­ Hydrurus fo etidus, Zygnema b and To lypothrix dis­ relation coefficient(rho) was calculated and also the torta v. penicillata, together with an abundance of Pearson ditto on the ranked observations (Conover Ceratoneis arcus v. linearis existed. Other diatoms 197 1). These correlation coefficients were tested in were: Anomeoneis brachysira, Achnanthes kryo­ the corresponding two-tailed nonparametric test. phila, A. minutissima v. cryptocephala and Cera- Positive correlations were found between the Lakse

A eta Phytogeogr. Suec. 68 94 Catarina Johansson

Table I I. The Spearman rank correlation coefficients (rho ) for eight 2nd axis (1 :4l strear;JS , where the algal spec ies have been ranked.

99.9 xxx = 1. significance level XX = 99 .0 % X = 95.0 t

X 2 -.4289 XX 3 -.1152 -. 5700 XXX XXX 4 -.4909 -. 5435 -.0522 XXX 5 1191 +. -. 7259 +. 0972 -. 1190 XX I 6 -. 1000 -. 5964 +. 1060 -. 0955 +. 3457 X XX / 7 +. 1829 -.4430 +.2857 -. 2568 +. 64 17 +.3593 XXX XX 8 -.0376 .. . 9052 +. 1554 -.0319 ... '. 5333 +.319 2 +. 5316 ... -- - ' �, ,,''' stream and the Grense stream and between the Lakse stream and the Kringlerne stream both at a Fig. 7. Ordination of the sampling occasions from eight .significance level of 99 %. streams in the Narssaq area in SW Greenland on two axes The species composition of the K vane stream was (2:8/ 1 :4) in a polar ordination. The circles surrounding the fo und to be negatively correlated with each of the numbers indicate the number of species where each circle remaining streams investigated, highly so at the = 5 species. For explanation of numbers see Fig. 1. 99.9 % level of significance,with Ref. Ill stream, the Grense stream and the Kringlerne stream. Further, condition appeared to be the relative amount of wet Ref. was found to have a high negative correlation flushing on the run-off area above the sampling Ill with the Narssaq river. localities. Both environmental gradients showed cor­ The regional variations in species composition relations with the exhibited successions in the ordina­ were studied applying the technique ofpolar ordina­ tion (Fig. 7). Other investigated factors proved less tion (Poole 1974). The technique was modified with successful in explaining the ordination pattern. No respect to the matrix of dissimilarities, using an index hydrodynamic variates could explain any systematic d= 1-{J running from 0 to 2 calculated directly from variation in the ordinated species composition. the Spearman rank correlation coefficients (rho). No comprehensive analyses of the stream water in In the ordination the localities are placed along the the area have been conducted. But all water samples first axis according to their relative similarity with from the area, predominantly lake samples, show each of the two most dissimilar localities, the K vane very low nutritive values. The four localities stream (2) and the Kringlerne stream (8), see Table positioned at the head of the Kangerdluarssuq Fiord Il. The end points of the second axis are made up of (Fig. 1), delineated by the dashed line in the ordina­ the most dissimilar central positioned locality pair tion diagram (Fig. 7), constitute a distal cluster with along the firstaxis. The Narssaq river and the Ref. short distance from the shore and little wet flushing Ill stream appeared to be the best choice of pairs. over the vegetation in their run-off area, but also In order to interpret the polar ordination diagram located within the Ilimaussaq intrusion in close (Fig. 7) in terms of environmental conditions, the mutual connection. One alga, A nomoeoneis localities have been ranked with respect to gradients brachysira v. lanceolata, is specific for this group of in each of several more or less complex factors. The localities and several others to some extent are con­ directions giving the highest rank correlation coef­ fined to this area in the ordination, for instance, ficient between the factor ranks of the localities and Ceratoneis arcus v. linearis, Achnanthes minutissima the ranked positions of the localities along the direc­ v. cryptocephala and Frustulia rhomboides v. sax­ tion in question have been calculated. Two complex onica. factors appeared to be correlated with the positions Two localities, the K vane stream and the Ref. Ill of localities in the ordination diagram. stream, deviate from all other localities. The diverg­ A meaningful and apparently ecologically signifi­ ing position of Ref. stream is predominantly due Ill cant factor seemed to be the distance fr om coast: to a number of algae unique to this locality in the pre­ measured as the declining gradient of sea exposure sent investigation, such as Chamaesiphon confer­ and quantified as the re-ranking of the means of the vicola, Hormidium rivulare and Eunotia tridentula v. ranks of vertical and horizontal distance from the perminuta, whereas the position of the K vane stream shore. Another seemingly adequate environmental in the ordination is to some degree due to the absence

A eta Phytogeogr. Suec. 68 Attached algal vegetation in some streams fr om the Narssaq area, Greenland 95

Table Ill. List of algae from 8 streams in the Na rssaq area investigated during the surrrner of 1977. Anomoeonei s brachysira (Breb.) Grun. CYANOPHYTA A. brachysira v. lanceolata (May.) Cleve Gloeocapsa sp. A. brachys ira v. thenna lis (Grun.) Cleve Chamaesiphon confervicola A. Graun Ceratoneis arcus (Ki.itz .) C. rustafinski Hansg . C. arcus v. linearis Holmboe S t i gonema mami 11os urn ( Lyngb . ) Ag . Cymbel la affinis Ki.i tz . Tolypothrix distorta v. penicil lata (Ag.) Lerrrn . C. cesati (Rabenh.) Grun. Nostoc spp . C. del icatula Ki.itz . Osci llatoria sp. C. gracilis (Rabenh .) Cleve Phormidium autumnale (Ag.) Gom . C. parva (W. Smi th ) Cleve P. valderiae (Delp. ) Gei tl . C. v�ntri cos a Ki.itz . Lyngbya 1 eptonema Skuja Diatoma hiemale v. mesodon {Eh r.) Grun. Sch i zothri x fuscescens Ki.itz . D. hiema le v. mesodon f. typica May Eucocconeis flexel la (Ki.itz.) Rabenh . CHLOROPHYTA Eunotia arcus Ehr.

Ulothricales E. bigibba v. pumula Grun. E. diodon Ehr. Ul othri x aequa 1 is Ki.itz . Honnidium rivulare Ki.itz . E. monodon v. maior fo . bidens (W. Smith) Hust. E. pecti nalis (Ki.itz.) Rab�nh . Mi crospora pa1 ustri s v. minor Wi chmann E. pectinalis v. minor (Ki.itz.) Rabenh . Oedoaon i a 1 es E. tridentula v. penninuta Grun. Oedogonium sp. (cel lwidth 10 pm ) Fragilaria capucina v. lanceolata Grun. Bulbochaete sp. (sterile) F. i ntennedi a Grun . F. pinnata Ehr. Zygnemata es 1 Frustulia rhomboides {Ehr.) de Tony Zygnema ( Israel son � 1949 ) F. rhomboides v. saxonica (Rabenh.) de Tony b -- z. F. rhomboides v. saxonica f. capi tata A. Mayer - 11 - z. £ Gomphonema angus ta turn Rabenh. Mougeotia - 11 - £ G. longiceps v. subcl avata Grun. G. parvulum Ki.itz . CHRYSOPHYTA Meridian circulare Ag . Chrome 1 i na 1 es M. circulare v. constricta (Ralfs) van Heurck Hydrurus foetidus (Vill.) Trev. Nitzschia dissipata (Ki.i,�z .) Grun. Pinnularia appendiculata (Ag.) Cl eve Centrophyc i dees P. borealis Ehr. Cyc ote a comta ( Ehr.) Ki.itz . 1 11 P. d i vergens W. Smi th

Penna tophyc idees P. gibba Ehr. Achnanthes fragi larioi des Petersen P. i nterrupta W. Smith A. kryophi la Petersen Synedra rumpens Ki.itz. A lineari s W. Smi th S. ulna (Nitzsch .) Ehr. A. minuti ssima v. cryptocephala Grun. Tabellaria flocculosa (Roth.) Ki.itz .

of algae present in all the other streams : flora outside the intrusion, where granite rock 3 To lypothrix distorta v. penicillata, Achnanthes mi­ predominates, than inside. Concerning taxonomic nutissima v. cryptocephala and Ta bellaria floc­ characterization of the streams, diatoms dominate in culosa. But also the presence of several unique algae streams except Reference Ill stream, where the all are important, such as: Oscillatoria sp., Zygnema c, Cyanophyceae together with green ftlamentous algae Melosira spp. and Meridion circulare v. constricta. are prevailing. The diatoms are rare in the lakes and In this investigation the number of species varies ponds apart fr om Ta bellaria flocculosa, which was between 5 and (Fig. The Grense stream has present in most of the localities examined. This ac­ 33 7). 7 and the Kringleme stream 5 species. Both of these are cords well to this investigation where the above­ situated inside the Ilimaussaq intrusion where the mentioned alga was very fr equent often in mass mean species number is while those outside have growth in streams except the N arssaq river and 13, all Lakes and ponds investigated in the area the K vane stream. 21. (Nygaard, unpublished) have shown a richer algal Compared to studies in West Greenland north and

Acta Phytogeogr. Suec. 68 96 Catarina Johansson south of the island Disco (Foged 1958), the Narssaq REFERENCES area is poor in species. The diatomaceous floranear Bachmann, H. 1921. Beitrage zur Algenflora des Disco usually is rich in species, the mean number Siisswassers von West-Gronland. - Mitt. Naturforsch. from 14 of the streams investigated by Foged was 47 Ges., Luzern 8:1-181. and only two small mountain sources had 7 and 8 Bocher, Tyge, W. 1950. Structure and biology of four respectively. In the Disco area Meridion circu/are species of the Stigonemataceae from a shallow pool at was one of the predomiant algae in the streams Ivigtut. - Meddr Gr0nland 147(5):1-21. together with Ceratoneis arcus, Nitzschiafrustulum, Conover, W.J. 1971. Practical nonparametric statistics. ­ N. pa/ea, Achnanthes linearis, A. kryophi/a, Cym­ New York. bella ventricosa and Diatoma hiemale. This differs Foged, N. 1953. Diatoms from West Greenland collected land 147(10):1-87. from the Narssaq area as a whole where the following by Tyge W. Bocher. -Meddr Gr0n 1958. The diatoms in the basalt area and adjoining diatoms were dominating: Achnanthes minutissima v. areas of archean rock in West Greenland. - Meddr cryptocepha/a, Anomoeoneis brachysira v. therma/is, Gr0nland 156(4):1-147. Ceratoneis arcus v. linearis, Frustulia rhombiodes v. Israelson, G. 1949. On some attached Zygnemales and saxonica and Ta bellaria flocculosa. their significance in classifying streams. - Bot. N otiser 1949:3 1 3-358. Nygaard, G. 1954. A new diatom species and two new -SUMMARY varieties from plankton in Lake Taserssuatsiaq. - Meddr Gr0nland 148(1):3 11-3 13. A study of the attached algal vegetation in running - (unpubl.) Phytoplankton from lakes and ponds of the waters was carried out in three areas in SW N arssaq area, South Greenland. - Freshwater Greenland, the N arssaq area, the Reference Ill area Biological Laboratory, University of Copenhagen. and the Kangerdluarssuq area. The first one is Poole, R.W. 1974. An introduction to quantitative dominated by alkaligranite and is called the N arssaq ecology. - Tokyo. 532 pp. intrusion, the second has a bedrock mostly consisting Rose-Hansen, J., Nielsen, C.O. & S0rensen, H. (eds) 1977. of granite and the last area is located inside the The Narssaq Project. - Progress Report No. 1, Inst. Ilimaussaq intrusion where syenites dominate. Petrology, Univ. Copenhagen. 82 pp. From the eight streams studied, diatoms Literature not referred to dominated in most of the streams, Hydrurusfoetidus Brun, J. 1900. Diatomees de Jan Mayen et du Groenland. in one (lower parts of the Narssaq river), Bih. K. svenska Vetensk. Akad. Handl. cyanophyceans in another (the Talut stream) and 26(18): 1-22. watermosses such as Hygrohypnum ochraceum in a Bocher, Tyge W. 1949. Studies on the sapropelic flora of third (the Kvane elv). the Lake Flynders0 with special reference to the In order to investigate the similarity of the streams Oscillatoriaceae. - Bioi. Meddr 21(1): 1-46. with regard to the algal composition, a Spearman B0rgesen, F. 1894. Ferskvandsalger fra 0stgr0nland. - rank correlation coefficient between the ranked Meddr Gr0nland 18: 1-41. species in each stream has been calculated. The 1910. Freshwater algae from the "Danmarkexpedition" regional variations in species composition was to north-east Greenland. Meddr Gr0nland 43:71 -90. studied applying the technique of polar ordination. It Cleve, P.T. 1882. Farskvattensdiatomaceer fr fm Gronland was fo und that a richer algal floraexisted outside the och Argentinska republiken. - Ofvers. K. Vetensk. Illimaussaq intrusion, where granitic rock pre­ Akad. Forh. Stockh. 38(10):3-13. dominates, than inside where loose layers of syenites Foged, N. 1955. Diatoms from Peary Land, North exist. Greenland collected by Kjeld Holmen. - Meddr Gr0nland 128(7): 1-91. Larsen, E. 1907. The freshwater algae of East Greenland. Acknowledgements - Meddr Gr0nland 30:75-110. I am grateful to Prof. H. Sjors for valuable comments on Petersen, J. Boye. 1924. Freshwater-Algae from the North the manuscript and I also would like to thank Fil. kand. L. Coast of Greenland. Coil. by the late Dr. Th. Wulff. - Kronborg for kind help in the identification and confirma­ Meddr Gr0nland 64:305-3 19. tion of some intricate algal taxa and also for the checking 0strup, E. 1898. Ferskvands-Diatomeer fra 0st­ of the manuscript. Doe. E. Sjogren has been helpful in the Gr0nland. - Meddr Gr0nland 15:25 1-290. identification of some mosses. For the statistical work I 1912. Diatoms from North-East Greenland. - Meddr wish to thank my husband Cand. scient. H. Mosegaard. Gr0nland 43:193-254.

Catarina Johansson, Institute of Ecological Botany, Uppsala University, Box 559, S-75 1 22 Uppsala, Sweden

Acta Phytogeogr. Suec. 68 Three new subspecies of the apomictic Ranunculus auricomus L. (s.str.) from the Flora Upsaliensis area in Uppland, Sweden Erik Julin John Axel Nannfeldt &

Uppland is a part of the area of most intense medio rotundate cuneato, crenato, lobis lateralibus postglacial evolution of new taxa of Goldilocks in latis, lobulatis vel crenatis. Folia intermedia multipar­ North-West Europe. The area also comprises Soder­ tita, segmentis obtuse lanceolatis vel - posterioribus manland, the Aland Islands and the south-west part - fere pari latitudine. Folia interiora foliis ex­ of the Finnish mainland. The occurrence of a very terioribus similia sed tripartita, segmentis lateralibus great number of indigenous subspecies with small profunde lobatis vel incisis. Folia aestivalia utrim­ ± ranges is characteristic of this area. The three que passim pilosa, ambitu orbicularia, sinu basali described here are very likely examples of such ones. angusto vel obtecto, tripartita, segmento medio In a previous paper (Julin & Nannfeldt 1967) we cuneato, segmentis lateralibus latis, segmentis have described 32 subspecies of R. auricomus (s.str.) irregulariter vel acute vel obtuse dentatis. Laciniae and recorded the occurrence of a fu rther 13 in fo liorum caulinorum obtusae et - praesertim Uppland. Other papers (Julin 1963, 1967, 1977, validiores folii infimi - late breviterque oblanceolatae 1978, 1980; Marklund 1961) contain information et inter medium et apicem dentibus brevibus porrectis about Goldilocks in Uppland, of which there are 8 nonnullis praeditae. Flares plerumque imperfecti, new subspecies. Finally, 15 subspecies have been petalis nonnullis abortivis, interdum tamen sat perfec­ found in U ppland, although their presence in the ti, ad 25 mm diametro. Sepala prasina. Stamina province has yet to be published. In all c. 70 sub­ numerosa apicem capituli pistillorum fere aequantia species of R. auricomus (s.str.) are known from vel paulo superantia. Torus ovatus, dense pilosus, 5 Uppland. This may seem a small number in com­ mm longitudine. Carpellophora sat brevia. An­ parison with Sodermanland (249 subspecies; Julin droclinium teres intervallo fere duplo longius. 1980) and the Aland Islands, where 76 subspecies oc­ cur in the parish Vardo alone (Olofsson 1961), but may possibly be a result of the insufficientin ventory, Localities inter alia in the zone closest to Lake Malaren and the Sweden : Uppland. Skuttunge, section II, Askarback, Alnus incana large islands in the east part of the lake, regions wayside at the locality of 1943 H. Sj ors probably rich in taxa. Moreover, there are still collec­ (UPS); Askarback, southward, grove W of the main tions in UPS that so far have not been worked up. road, round an old stump 1968 J.A.N. 20344 (UPS); Askarback, southward, shady rivulet W of the main road, in coniferious wood 1968 J.A.N. 20345 (halo­ R. auricomus L. ssp. sj oersii Julin Nannf.ssp. type, UPS). & nov. (Figs. 1, 2) Holotypus Ssp. sjoersii is named in honour of Hugo Sj ors, its first collector, who found it as early as in 1943 while working Suecia, Uplandia, par. Skuttunge, ab Askarback with an inventory for the Flora Upsaliensis project. meridiem versus, juxta rivulum umbrosum in silva conifera 29.5.1968 J.A.N. 20345 (UPS). L. ssp. Julin Nannf. R. auricomus vaginifer & Planta sat alta. Caules plerumque erecti, inferne ssp. nov. (Figs. 4) violacei, ramum infimum sub angulo <30° 3, emittentes, intra reliquias foliorum rosulariorum anni Holotypus proxum vaginis aphyllis 1-2 instructi. Folia Suecia, Uplandia, par. Skuttunge, Myrby, prope in­ rosularia magna, exteriora, intermedia et interiora troitum in vicinia viae communis 18.5.1943 J.A.N. ambitu reniformia. Folia exteriora trilobata, lobo 6375 & H. Smith (UPS).

A eta Phytogeogr. Suec. 68 98 Erik Julin John Axe! Nannfeldt &

Fig. 2. R. auricomus ssp. sj oersii. Basal leaves of the type collection. - Below right: Receptacle.

vel plurimis abortivis. Sepala prasina. Capitulum If( 1 :h ...., ...... pistillorum magnum mox stamina numerosa superans. Rostra carpellorum sat longa et tenuia, Fig. I. R. auricomus ssp. sj oersii. Holotype. dimidium inferius eorum pilis sat longis dense instruc­ tum. To rus ovatus, 6 mm longitudine, dense pilosus.

Planta alta. Ca;ules · erecti, sat graciles, inferne Carpellophora sat brevia. Intervallum nullum. An­ violacei, ramos sub angulo fe re 45° emittentes, intra droclinium sursum crassius. reliquias fo liorum rosulariorum anni proximi saepe vagina aphylla una instructi. Folia rosularia magna, ambitu reniformia, sinu basali aperto vel apertissimo. Localities Folia exterior a basi truncata, trilobata - tripartita, Sweden: Uppland. Billinge, damp glade at the main lobo (segmento) medio fere pari latitudine, in­ road c. 400 m E of Akerlfuma 1968 J .A.N. 2034 1 tegerrimo vel apice acute dentato, lobis (segmentis) (UPS). - Skuttunge, Myrby, at the approach near the lateralibus crenatis vel lobulatis, sinibus inter lobos main road 1943 J.A.N. 6375 & H. Smith (holotype, vel segmenta latis. Folia intermedia multipartita, UPS), J.A.N. cult. 1943:37; Myrby, damp and segmentis divergentibus, anguste lanceolatis, in­ shady coniferous wood at the way from Bjorkfallet, tegerrimis vel dentibus parvis acutis singulis praeditis, just S of a ditch (formerly a rivulet) 1968 J.A.N. segmentis anterioribus saepe bi-vel trilacinulatis. 20339 (UPS); Myrby, Bjorkfallet, shady ditch along Folia interiora acute triangulari-dentata, tripartita, a spruce hedge 1968 J. A. N. 20338 (UPS). segmento medio cuneato, apice bi-vel trilobulato, segmentis lateralibus multilobulatis - fe re indivisis. Ssp. sjoersii and ssp. vaginifer belong to those sub­ Folia aestivalia utrimque parce pilosa, inaequaliter species of R. auricomus, which approach R. fa !lax - acute mammiformi- vel triangulari-dentata, indivisa through the occurrence of leafless sheaths of basal vel levissime incisa. Laciniae fo liorum caulinorum leaves, relatively considerable magnitude, and, con­ longae, anguste lanceolati-lineares, integerrimae vel cerning the former subspecies, also the configuration dentibus vel lacinulis angustis patulis singulis of the segments of cauline leaves. Yet, our opinion, in praeditae. Flares valde imperfecti, petalis omnibus they relate better to R. auricomus.

Acta Phytogeogr. Suec. 68 Three new subspecies of Ranunculus auricomus in Upp land 99

Fig. 4. R. auricomus ssp. vaginifer. Basal leave of the type collection. - Below right : Receptacle.

tulis lanceolatis, acutis, plerumque integerrimis, inter­ dum prorsum curvatis. Folia interiora tripartita, segmento medio lanceolato-cuneato, segmentis lateralibus latis, lobatis, segmentis lobisque acute Fig. 3. R. auricomus ssp. vaginifer. Holotype. mammiformi-dentatis. Folia aestivalia foliis in­ terioribus similia sed trilobata, lobis triangulariter serratis. Laciniae fo liorum caulinorum mediocriter L. ssp. Julin Nannf. R. auricomus nontunensis & longae, lanceolati-lineares, obtusulae, integerrimae ssp. nov. (Figs. 5, 6) vel dente vel lacinula una alterave praeditae, Holotyp us validiores fo lii infimi vulgo petiolulatae. Folium caulinum infimum saepe petiolatum. Flares perfecti, Suecia, Uplandia, par. Danmark, in coryleto Non­ petalis omnibus plerumque evolutis, ad 19 mm tunae a N parte intercessionis naturae juxta 2 7.5. diametro. Sepala apice rubro-violacea. Stamina 1979 E. Evers (UPS). numerosa apicem capituli pistillorum superantia. Torus anguste ovatus, c. 5 mm longitudine, viridis, Caules Planta mediocriter alta. erecti vel curvati, in­ dense pilosus. Carpellophora sat brevia. Intervallum fe me violacei, ramos sub angulo plerumque 30°-45° obscurum. Androclinium teres vel sursum aliquanto Folia emittentes, caules laterales subtus saepe pilosi. crassius. rosularia vulgo reniformia, sinu basali aperto. Folia exteriora lobata - partita, lobis vel segmentis crenatis. Folia intermedia tripartita, segmento medio Locality angusto, vel lanceolato vel cuneato vel fe re pari Sweden: U pp land. Danmark, N fmtuna hasseUund latitudine, integerrimo vel apice dentato, segmentis (Nfmtuna hazel grove), just N of the nature reserve lateralibus lobatis vel iterim partitis, lobis vel segmen- 1979 E. Evers (UPS).

Acta Phytogeogr. Suec. 68 100 Erik Julin John Axe! Nannfe ldt &

t .. ,.- ,_,: '>. ' � � ...... �, , . '· _' ·� .! o, ' ' o •; • o w ;• ' ·: • o •o - ... 0 � l,. I W -· � 0 �:� ®'

� " -

Fig. 6. R. auricomus ssp. nomunensis. Basal leaves of the type collection. - Below right: Receptacle.

REFERENCES

Julin, E. 1963. Der Formenkreis des Ranunculus auricomus L. in Schweden 1. - Ark. Bot. Ser. 2, 6:1-28. 1965. Der Formenkreis des Ranunculus auricomus L. in Schweden 2. - Ibid. Ser. 2, 6:29- 108. Fig. 5. R. auricomus ssp. nontunensis. Holotype. 1967. Der Formenkreis des Ranunculus auricomus L. in Schweden 4. - Ibid. Ser. 2, 6:243-308. 1977. Some Bothnian subspecies in the Ranunculus auricomus complex: origin and dispersal. - Bot. Notiser 130 :287-302. 1978. New subspecies of Ranunculus auricomus from Vastmanland, Sweden. - Ibid. 131 :3 1 7-348. 1980. Ranunculus auricomus L. (s.l.) in Sodermanland, East-Central Sweden.- Op. bot. Soc. bot. Lund. 57. In press. SUMMARY Julin, E. & Nannfeldt, J. A. 1967. Der Formenkreis des Ranunculus auricomus L. in Schweden 3. - Ark. Bot. Ranunculus auricomus L. (s.str.) ssp. sj oersii Julin & Ser. 2, 6:163-241. vaginifer non­ Nannf., ssp. Julin & Nannf., and ssp. Marklund, G. 1961. Der Ranunculus auricomus-Komplex tunensis Julin & Nannf. from the Flora Upsaliensis in Finnland 1. - Flora fenn. 3:1-128. area in Uppland are described and illustrated. In all c. Olofsson, P. 1961. Ranunculus auricomus-gruppen i 70 subspecies of R. auricomus L. (s. str.) are at pre­ Vardo (Aiand). - Acta Soc. Fauna Flora fenn. sent known from the province of Uppland. 76:(3):1-47.

Erik Julin, Borgaregatan 11, S-611 00 Nykoping, Sweden John Axel Nannfeldt, Institute of Systematical Botany, Uppsala University, Box 54 1, S-7 5 1 21 U ppsala, Sweden

Acta Phytogeogr. Suec. 68 Fine-root Dynamics in a Scots Pine Stand with and without Near-optimum Nutrient and Water Regimes Hans Persson

Plant ecologists studying root system development centages of the aboveground parts of the plant crop and maintenance in relation to their function in (cf. Persson 1978, 1980a), are, furthermore, of no terrestrial ecosystems, often over-emphasize the part relevance as regards the small-diameter roots and played by the long-lived and therefore more accessi­ may lead to serious underestimates. ble plant parts. Thus, studies of the growth of sup­ It is important to realize that we are only beginning porting structures of a fo rest tree (branches, stems, to understand some of the complex processes that large-diameter roots, etc.) are prerequisite for an un­ determine death and replacement in the belowground derstanding of how the tree allocates its resources part of the ecosystem. In this respect, studies by the with age, but studies of the growth of leaves (or present author have increased our knowledge of the needles) and small-diameter roots are of an im­ role played by the belowground plant components in mediate vital functional importance. Most large­ some coniferous fo rests in relation to total growth diameter roots die with the tree itself, while the fine and turnover (cf. Flower-Ellis & Persson 1980). Since root-ramifications are in a constant flux, with death few such determinations are available at the moment, and replacement taking place simultaneously they have also opened new perspectives on what we (Persson 1978, 1979b, 1980a, 1980b). hope to reach in the future: a fundamental knowledge Studies by, e.g., Orlov (1955), Sonn (1960), of the periodicity of root growth in relation to en­ Reynolds (1970), Santantonio (1979) and Persson vironmental factors. The present paper is restricted to (op. cit.) show that the amount of fine-root investigations in situ of the effect of fe rtilization and necromass in many fo rest ecosystems may be larger irrigation on root growth. Preliminary results have than the amount of fm e-root biomass. The fine-root earlier been reported in Persson (1979b) and Ericsson population has been shown to replace its weight & Persson (1980). several times during the growing season (Persson op. cit.). Since a large carbohydrate reserve is always available in the root system during the growing Material and methods season (cf. Ericsson & Persson 1980), there is no reason to suppose that carbohydrate supply should The investigation was conducted in a 15- to 20-year-old Scots pine (Pinus sylvestris) stand (cf. site description in normally limit root growth. On the contrary, since Axelsson & Bdikenhielm 1980), situated on sandy sedi­ there is a growing body of evidence from a variety of ments of glacifluvial origin at Ivantjarnsheden in Central ecosystems (cf. Caldwell 1979; Agren et al. 1980; Sweden, at the field station of the Swedish Coniferous Harris et al. in press) that the energy investment in Forest Project (60°49'N, 16°30'E, 185 m above m.s.l.). the belowground part of the ecosystem is consider­ The fo rest type is of lichen to lichen-dwarf-shrub type (cf. able, there is reason to conclude that the fine roots in Ebeling 1978). The nomenclature used for vascular plants particular constitute an important sink of car­ fo llows Lid (1974). The stand is self-sown from seed trees, bohydrates. after clear-felling in 1957, and had an average height of It must be remembered that all comparisons of net about 2.2 m in the spring of 1974. Parts of the stand were primary production at an ecosystem level, which are cleaned in August 1972, i.e., before the investigation began. A less extensive cleaning was undertaken late in the based solely upon aboveground estimates, assume winter of 1972-1973, in an attempt to create the same constant ratios of aboveground to belowground density throughout the area. The stand has previously been production. The latter ratio may, however, diffe r con­ described by Flower-Ellis et al. (1976). siderably (even in the same species) in different com­ The soil is a pod sol, with a F /H layer thickness of ap­ munities (Persson 1975, 1979a). Conventional ways proximately 2.5 cm and an Ahh/ Ahu layer (Babe! 1970) of of assessing root production, from growth per- about 2.5 cm in 1974 (cf. Persson 1980a), i.e., before the

Acta Phytogeogr. Suec. 68 102 H ans Persson

' Fig. I. The Ih 11 research area on 7 October, 1975, shortly after sampling had started. To the left is one of the control plots and to (O) the right one of the irrigated and fertilized plots (IF). The sheet of aluminium in the middle of the figure separating the two plots was forced down to 33 cm in the soil before treatment started, to pre­ vent roots from one plot growing into the neighbouring plot. The treatment started in July 1974. (Photo : S. Oscarsson.)

Fig. 2. The Ih 11 research area, at the same place as in Fig. 1, on 24 August, 1976, almost one year later. Notice the considerable growth of the Scots pine in the foreground to the right on the IF­ plot. (Photo: S. Oscarsson.) investigation had begun. The mineral soil consists of sandy was about 3 I m-2·day-1 and the total amounts of sediment (coarse sand with fractions of medium sand). The nitrogen supplied to the IF-plots were 7, 10, 15 and soil surface was treated with a tractor scarifier after the 15 g m-2 during the years 1974-77, respectively. All 19 57 clear-felling, resulting in fairly regular patches (ap­ other essential elements were given in proportion to proximately 1 m in diameter) of exposed mineral soil. nitrogen according to Ingestad (1967). The fertilizer was distributed to the IF-plots dissolved in the irriga­ The experimental area is divided into plots subjected tion water. Details of the irrigation and fe rtilization to experimental treatment with the intention of programme are given by Aronsson et (1977). al. eliminating water and mineral nutrients as growth­ The temporal variation of fine-roots was estimated limiting factors (Figs. 1 and 2). The treatments were: in ingrowth cores (containers) removed on successive irrigation (I) every day during the growing season, occasions regularly during the growing season. The fertilization (F) with solid fertilizers once a year, and containers were placed in the holes left by the daily irrigation plus fertilization (IF) fivedays a week withdrawal of the soil cores (diameter 6. 7 cm) in a during the growing season. The irrigation and fer­ long steel corer. They consisted of cylindrical net tilization started in 1974 (cf. Ericsson & Persson "stockings" ftlled with sifted sand of local origin and 1980 fo r the periods of irrigation and fertilization peat instead of raw humus at the F/H layer. The during 1974-77). The water supplied by irrigation mesh size of the "stockings" was 7.5 mm, thus allow-

A era Phytogeogr. Suec. 68 Fine-root dy namics in a Scots pine stand 103

ing the roots to penetrate freely. In all, 600 containers I968 p. 204). Thus, the probability of obtaining at the were installed in the 0-, 1-, F- and IF-plots from Au­ most k negative outcomes is given by the formula: gust to the middle of September I974, equally distri­

= buted between the different experimental plots. Ran­ P = I� o<:) dom samples of 9 cores were taken from each treat­ 2m ment, beginning in the autumn of I975, and continu­ P ing throughout the growing seasons of I976 and where is the significance level and: I977 (cf. Persson I979b; Ericsson & Persson I980). m 1) = m! m (m- 1) (m-2) ...(m -n+ The surrounding roots and soil were removed, and ( ) --­ 3·2· 1 sand sifted away and the samples were then stored in n ' n! (m-n)! n (n-1) (n-2) . .. a deep-freeze at -20°C until the finalsorting could be The null hypothesis (i.e., that root variables did not carried out (for details of the complicated sample differ between the 0- and the IF-plots) was rejected treatment, cf. Persson I978, I980a, I980b ). at P = 0. 1. However, the specific significance levels However, due to the extensive work involved, cores are given separately in the text (see Results). from only seven samplings in the 1- and F-plots, respectively (54 cores in all) could be sorted during the period 1975-77. The total number of samplings in the 0 and IF-plots, respectively, was I4 (126 cores Results in all). The living and dead root fragments (biomass and It may be concluded from the accumulation of new necromass, respectively) were sorted into different roots in the ingrowth cores (Figs. 3, 4, 6) that a con­ diameter fr actions, fo llowing diameter measurements siderable fine-rootaccretion had taken place in the 0-, made in the middle of each fragment with vernier 1-, F- and IF-plots. As regards the increments in callipers (Persson 1978, I980a, I980b). The total fine-rootlength of the < I mm diameter fraction, the lengths in each diameter-class (cf. Persson 1978 for ingrowth of new roots evidently proceeded more the diameter classification) were estimated with a rapidly in the IF-plots than in the 0-plots, since the "line intersect method" similar to that used by New­ mean length generally fluctuated at a higher level. man (I 966) and further elaborated by Tennant This is also confirmed by the test given above (see

(I 97 5). The root tips were separated into long and Material and methods) at P = 0.001. When short roots and counted; for definition cf. Hatch & samplings from the 1- and F-plots were compared Doak (1933), Hatch (I937). The long-root tips are with the samplings from the 0-plots, the amount of readily distinguished from short-root tips by their new roots in the latter was significantly lower (P = white, often swollen terminal parts, which are nor­ 0.008 and at P = 0.063, respectively). mally more than I 0 mm in length, depending on the As regards the fine-root weight of the < 1 mm size of the root. The short roots are often converted diameter fraction, the latter fluctuated generally at a into complex mycorrhizal structures, either irregular­ lower level in the IF-plots (Fig. 3); the test shows that ly or racemosely branched, and form coralloid types the fine-root weight was significantly lower at P = of mycorrhiza. They may also dichotomize repeated­ 0.029. It is therefore reasonable to conclude that the ly and thus form branches in successions of higher < 1 mm fine roots were considerably longer and order towards the apex of the long roots. A special thinner in the IF-plots than in the 0-plots. No type consisted of aggregations of forked short-root significantly different estimate of the

"ball mycorrhiza" sensu "Knollen-typus"; Melin the F-plots the estimates were generally lower at P = I922). The number of root tips was impossible to es­ 0.063. timate in these aggregations and they were accord­ The length of the fineroot s in the diameter fraction ingly classified into balls < or > 2 mm in diameter. 1-2 mm, also fluctuated at a significantly higher Tests were made of the magnitude of the root level in the cores from the IF -plots than from the 0- variables (i.e., root length and weight in varying plots (P = Fig. 4). The related corresponding O.OOI, diameter fractions and the number of short- and long­ weight was larger in the IF-plots (P = 0.007). No root tips) in the 0-plots compared with 1-, F- and IF­ significant difference in length or weight of the fine plots, respectively. In the test the number of negative roots I-2 mm in diameter was found fo r the 1- and outcomes (k) in the total number of samplings (m F-plots. All these observations agree with data on the ) (viz., a higher estimate of the root variable in one of mean diameter of the excavated root fragments the samplings from the 0-plots) was compared with (Persson, in prep.). the binominal distribution (cf. Snedecor & Cochran The seasonal pattern of the fluctuations in fine-

Acta Phytogeogr. Suec. 68 104 Hans Persson

Fig. 3. Changes in biomass and length of the root fr action <... I mm 150 diameter during the period of study. The sampling began in autumn 1975 and continued throughout the growing seasons of

1976 and 1977. • = control plots

(0). � = irrigated plots (1). .&. = fertilized plots (F). • = irrigated and fe rtilized plots (I F). Standard

errors (11 = 9) are indicated.

: .. .

/ . +

0 -- 1975 OCT 1976 MAY JUN JUL AUC SEP OCT NOV 1977 MAY

50

Fig. 4. Changes in biomass and length of the root fraction 1-2 0 �50 mm in diameter during the period of study. For explanations of sym­ bols see Fig. Standard errors (n 0 3. _: 1975_: OCT 1976 MAY JUN ..U. AUC SEP OCT NOV 1977 MAY JUN JUL AUC SEP OCT = 9) are indicated.

Acta Phytogeogr. Suec. 68 Fine-root dy namics in a Scots pine stand 105

Fig. 5. Changes in the number of hort-root and long-root tips per square metre during the period of study. For explanations of sym­ bols see Fig. 3. Standard errors (n

= 9) are indicated.

,;200E g 150

;:�

�100 _ t - _ _ · . ··-. . __ _ .. _ g 50 +-- _ � t _

0  1975 OCT 1976 JUL AUG SEP OCT NOV 1977 MAY JUN JUL AUG SEP OCT

PINUS SYLVESTRIS l

200

CALLUNA VULGARIS

Fig. 6. The amount of root biomass and necromass of Pinus syh·estris, Cal/una 1 ulgaris and Vaccinium vitis-idaea in the in­ growth cores during the period of study. For explanations of sym­ bols see Fig. 3 (filled symbols = biomass; unfilled symbols necromass). Standard errors (n = 9) are indicated.

Acta Phytogeogr. Suec. 68 106 Hans Persson

-2 -2 root length and weight was followed by large fluc­ Table I. The ingrowth of new roots (g DW m yr ) into root free cores (containers ) removed on succes sive tuations in the number of root tips (Fig. 5). Short-root sampl ing occas ions . The es timates are the difference betwe en the highest and lowe st root amount in the cores tips were the most numerous, the average number of duri ng the peri od autumn 1975-75. Estimates are given as long-root tips during the whole period of study was me an ± standard errors (n = 9 ). Included in the calcul a­ tion were both liv ing and dead root fragments . Most root only 2.3, 3.3, 4.6, and 3.0 % of the total number of members , except for a few cases of Pi nus syl ves tri s �1e re less than 2 mm in diameter. root tips in the 0-, 1-, F- and IF-plots, respectively. Note the different scales in the diagrams fo r the number of long-root and short-root tips (Fig. 5). The experi mental trea tment Spec e IF number of both long-root and short-root tips fluc­ .i s'------­ tuated at a higher level per m 2 in the IF-p lots than in Pinus sylvestri s 139 ± 80 163 :': 72 the 0-plots. The test confirms that the numbers of Ca lluna vulgari s 26 :': 10 49 = 1 9 long-root and short-root tips were larger in the IF­ Vaccinium vi t is-idaea 26 ± 16 67 = 24 than in the 0-plots (P = 0.00 1 and P = 0.007, respec­ tively). There were no significant differences between the I- and F -plots. the ingrowth cores for Pinus sylvestris, Calluna It should, however, be noted that these estimates vulgaris and Vaccinium vitis-idaea fluctuated at are given per unit area. If the number of root tips is significantly higher levels in the IF -plots than in the calculated per unit length of the fineroots < 1 mm in 0-plots at P = 0.090, P = 0.00 1 and P = 0.00 1, diameter, averages of 0.07 and 2.85 cm-1 fo r the long respectively (Fig. 6). The test also confirms a and short roots, respectively, are found in the 0-plots significantly higher level of the fluctuations of C. and 0.08 and 2.60 cm-1 in the IF-plots. A large vulgaris in the !-plots at P = 0.063; otherwise there difference is obtained if the average number of were no significant differences between 1- and F-plots aggregates of "ball mycorrhiza" is included in the and 0-plots. Since the death of fineroots was propor­ calculations (see Material and methods for tionately low during the 1975-76 growing season definition). These aggregates were considerably fewer and the ingrowth of new roots took place fairly rapid­ in the cores from the IF-plots. Thus, the average ly, the accumulated standing crop must be considered number of aggregates, < and > 2 mm in diameter, as an indication of the annual below-ground produc­ respectively, was 6.6·1 0-2 and 7.1·10-2 cm-1 in the 0- tion (cf. Table I). The estimates given in Table I are plots and 1.1·10-2 and 0.6·10-2 cm-1 in the IF-plots. derived fo r the 0- and IF -plots, no estimates are given

Since the number of short root tips included in the fo r the 1- and F-plots, due to the smaller number of aggregates may comprise up to ten or even more root samplings. These production estimates are not re­ tips per aggregate unit, it is obvious that mycorrhiza markably different from earlier results from a se­ formation takes place to a lesser extent in the IF­ quential sampling programme of soil cores in the plots. A comparison between the I- and F -plots and same stand (Persson 1978, 1979b, 1980a). The the 0-plots as regards averages of the number of root numerical differences between the 0- and IF-plots tips per unit length is not valid, since rather few (Table I), in spite of large standard errors of the es­ samplings were sorted from these treatments. It is, timates, are likely to be real bearing in mind the however, quite clear from the present data, that the overall tendency as indicated by the test. number of "ball mycorrhiza" aggregates was con­ siderably lower in F-plots than 0-plots. The accretion of fine-rootlength and weight in the Discussion above diameter fractions ( < 1 mm, 1-2 mm), was at In spite of the growing interest during recent decades its greatest in the IF -plots, and was fo llowed by a in studies of the productivity and fu nctioning of forest successive increase in fine-root length and weight in ecosystems the effectof fertilization and irrigation on the diameter fraction >2 mm in diameter, the in­ the root growth of forest trees (cf. review in Sutton crease being substantial at the end of the growing 1969, Hermann 1977), has remained incompletely in­ season of 1976 and during 1977. A considerable ac­ vestigated. According to in situ experiments, con­ cretion of Pinus sylvestris fine roots <2 mm in fined to fairly young tree plants (Heinsdorf 1966; diameter evidently took place both in the 0-plots and Paavilainen 1967, 1968, 1974; White et al. 1971), the IF-plots during the growing season of 1976, as fertilization causes an increase in the total weight of judged fr om the amount of root biomass in the cores the fine roots. Application of solid fertilizers (Fig. 6). As regards Calluna vulgaris and Vaccinium (especially of phosphorus) seems to stimulate super­ vitis-idaea, the ingrowth of new roots took place at a ficial rooting since vertical transport in the soil takes somewhat slower rate. place slowly. Thus, Buchholz & Neumann (1964) The total weight of both living and dead roots in found a considerable increase in the number of super-

Acta Phytogeogr. Suec. 68 Fine-root dy namics in a Scots pine stand 107 ficial fine roots ( < 1 mm in diameter) in a 57-year-old fertilization (e.g., the most favourable experimental Scots pine stand in East Germany, treated with conditions) induces changes in the growth strategy of

NPK + Mg. The number of fine roots in the deeper the small-diameter roots, other than changes in the soil-layers was lower as compared with that in the 0- total root weight alone. The changes may be advan­ plots. Saffo rd (1974) found increased average con­ tageous as a means of regulating growth rates in centrations of fm e roots ( <3 mm in diameter) in a different environments. Thus, increased length northern hardwood stand (beech-birch-maple) in growth or increased development of mycorrhiza may

New Hampshire, USA, treated with NPK + lime, be regarded as alternative ways fo r the plant to both in the soil surface and in the mineral-soil regulate the growth of the fine roots, both leading to horizons. Tamm (1979), referring to unpublished intense exploitation of available moisture and data by Flower-Ellis concerning a Scots pine stand in nutrients (cf. Persson 1980a). A similar change in the Central Sweden, established after wind-felling 1954 growth of the fine roots was also found in the F-plots, and fertilized annually from 1969 onwards with NPK since the weight of the < 1 mm diameter fraction was

+ Mg, showed that while there was a significant in­ generally lower there, while there was no differencein crease in the biomass of almost all fractions of the root length. In both cases these changes were aboveground tree structures on the fe rtilized plots, probably related to the reduced infection by the there was remarkably little differencein the fine-root mycorrhizal symbionts as a result of the fertilization biomass ( <2 mm in diameter) between treatments. (cf. Melin 1953 ). As regards the larger diameter frac­ Both excess and shortage of soil water seem to tions (1-2 mm and > 2 mm) both length and weight affectroot growth negatively (cf. Hermann 1977). On were significantly larger in the IF than in the 0-plots, the whole, fm e-root ramifications seem to be fairly while no difference was found in the or F-plots. 1- sensitive to desiccation. Substantial losses of fine-root The same accretion of fm e-root length and weight biomass occur whenever precipitation fails to main­ of P. sylvestris as above in the < 1 mm and 1-2 mm tain the soil moisture tension near zero (Ford & diameter fractions was also calculated in the collec­ Deans 1977; Deans 1979). Decline in the amount of tive < 2 mm diameter fraction (cf. Fig. 6). As regards fm e roots during the summer is largely related to Calluna vulgaris and Vaccinium vitis-idaea, a con­ periods of drought. Zak ( 1964) reviewed the works of siderable accretion in both length and weight in the several investigators and concluded that the elonga­ < 2 mm diameter was observed. The average level of tion of roots and mycorrhizal fo rmation may cease the variations fo r all species was significantly larger during periods of insufficient water supply and be in the IF than in the 0-plots. It is therefore justifiable resumed when environmental conditions are once to conclude that the increased growth of the fm e more favourable. roots in the IF -plots was a result of the improved soil There is a general lack of information concerning moisture and nutrient conditions. the rate of death and renewal of root systems in forest One striking feature of the results from the in­ ecosystems from which water, mineral nutrients or growth cores in all treatments is that few dead roots both have been eliminated as growth-limiting factors were found and that the bulk of the excavated root (as in the present study). Most data on root weight mass consisted of biomass. Similar experiences have and distribution in relation to environmental con­ been described by Ovington & Murray ( 1968), from ditions have been obtained by sampling on specific observations made the year after the installation of occasions rather than by observing changes over a vertical observation windows in a birch stand. Over a period of time. One of the more striking observations longer period of study the rate of production and from the present study was that the fine roots of mortality would undoubtedly be equal in the cores Pinus sylvestris ( < 1 mm in diameter) in the IF-plots and the necromass content would consequently in­ were significantly longer and thinner than in the 0- crease. Dead roots, when found in the cores, were plots, because root length generally fluctuated at a composed of whole branches that had died ofT.A fu n­ higher level and weight at a lower level (cf. Fig. 3). As damental feature of the growth dynamics of tree regards the and F-plots as compared with the 0- roots is the shedding of entire fine-root ramifications 1- plots, there were either no significant differences, or (cf. Kolesnikov 1968). This shedding mechanism is significantly lower estimates, of both root length and undoubtedly necessary, otherwise the soil horizons weight. In a teleological sense, in agreement with the would be completely interwoven with an endless conclusion from Tamm (1979) it is reasonable to ex­ tangle of small-diameter roots and further growth pect that the trees do not need as many roots if they would be hindered. are well supplied with nutrients. However, since fm e It is well established that mycorrhizal infection root length was significantly greater in the IF-plots causes increased longevity of rootlets (Orlov 1968; than in the 0-plots this indicates that irrigation plus Harley 1969). Mycorrhizal short roots may persist

A eta Phytogeogr. Suec. 68 1 08 Hans Persson fo r a period of a fe w weeks to several years. Non­ & Marx 1979), is costly in terms of energy for the mycorrhizal short roots, in contrast, may in many plant host. However, since a considerable car­ cases have a lifespan of a few days only (Kinman bohydrate reserve is present in the tree roots 1932; Kolesnikov 1968), probably because their throughout the whole growing season (Ericsson & apical meristems depend on an external source of Persson 1980), it is reasonable to conclude that the growth regulators to enable them to continue their ac­ tree invests in its fu ngal symbionts fr om a relative tivity (Wilcox 1968). The factors determining carbohydrate surplus. According to a carbon budget longevity of short roots are imperfectly known. made fo r a Scots pine in the 0-plots (Agren et al. Although, as inferred from some experimental 1980), about 60 % of the total net photosynthetic studies, it is not clear that the mycorrhizal structure production is invested in growth and maintenance of offers any direct advantage in preventing desiccation the root system. In the long run, knowledge of the under dry conditions (cf. Reid 1979). However, some complex processes determining root growth in a mycorrhizal fungi can grow better at a far lower fo rest ecosystem may be expected to indicate ways of water potential than higher plants alone (Ruehle & managing fo rests in such a way that a larger share of Marx 1979). The availability of soil moisture may the existing carbohydrate production may be directed also have a considerable effect on the composition of to the production of utilisable structures instead of the mycorrhizal association (Worley & Hacskaylo being used fo r the maintenance of the system, as is 1959). the case at present. The results of the present study indicate a larger average number of root tips per unit area in the IF­ plots than in the 0-plots. However, if the calculations SUMMARY are made per unit length and the number of "ball mycorrhiza" aggregates are included, considerably The temporal vanatlon in fine root growth was fewer root tips are found in the IF -plots, which im­ studied in a young Scots pine (Pinus sylvestris) stand, plies a decline in the number of short-root rami­ subjected to experimental treatment with the inten­ fications (e.g., the mycorrhiza). The period of the tion of eliminating water and mineral nutrients as experimental treatments was probably too short in growth-limiting factors. The sub-soil in this area con­ the I- and F -plots to show unequivocal changes in sists of sandy sediments of glacifluvial origin, which the ramification pattern of the short roots. One con­ is easily penetrated by sampling devices. The annual spicuous feature, however, in the temporal variation increments of fine-roots were estimated from the ac­ of the number of short-root tips was the low estimates cumulated amounts in ingrowth cores (containers)

in the 1-, F- and IF-plots on the last sampling occa­ removed on successive occasions during the growing sion during 1977 (cf. Fig. 5). During the same period, season. The containers were installed during autumn there was a gradual in�rease in the number of root 1974, resampling began in autumn 1975 and con­ tips in the 0-plots, which led to a fairly large estimate tinued throughout the growing season of 1976 and at the last sampling. Thus the fine roots in the 0- 1977. The vegetation structure is fairly simple with a plots, which are subjected to periodical drought and few dominating vascular species with distinctive deficiency in essential nutrients, may benefit from morphological rooting features, e.g., P. sylvestris, their symbiotic fungal partners. These findings arein Calluna vulgaris and Vaccinium vitis-idaea. The agreement with Bjorkman's (1941, 1942, 1944) treatments were: irrigation every day during the classical works, in which he showed that the number growing season (I), fertilization with solid fertilizers of short-root ramifications (sensu mycorrhizal fre­ once a year (F), and daily irrigation plus fertilization quency) drastically decreased with application of fe r­ fivedays a week (IF) during the growing season. The tilizers. experimental treatments started in 1974. A considerable part of the total fine-root weight The annual belowground production of fine-roots consists of mycorrhizal mycelia in immediate con­ <2 mm in diameter for P. sylvestris, C. vulgaris and tact with the root surface (Harley 1969). Because no V. vitis-idaea was estimated at 139, 26 and 26 g m-2 attempt was made before weighing to remove the in the 0-plots and to 163, 49 and 67 g m-2 in the IF­ fungal sheath of the mycorrhiza, it is to a varying plots. The fine-roots < 1 mm in diameter were degree included in the estimates of new root amounts significantlylong er and thinner in the IF -plots than in in the cores (Table I). There are strong reasons for the 0-plots and there was a larger average number of suspecting that this symbiotic relationship, which is root tips per unit area. However, if the calculations of extremely beneficial for the uptake of relatively im­ the number of root tips are made per unit root length mobile ions such as phosphates, zinc, copper, and the "ball mycorrhiza" aggregates are also in­ molybdenum and sometimes ammonium (cf. Ruehle cluded, fewer root tips are found in the IF-plots,

A eta Phytogeogr. Suec. 68 Fine-root dy namics in a Scots pine stand 109 which probably implies a decline the number of Mykorrhiza in mit Asche gediingten und ungediingten in short-root ramifications (e.g., the mycorrhiza) due to Teilen von entvassertem Moor. - Meddn. St. the fertilization. SkogsfOrsoksanst. 32:255-296. U The average level of variation of the fine-root 1942. ber die Bedingungen der Mykorrhizabildung bei Kiefer und Fichte. - Symb. Bot. Ups. 6(2): 1-190. weight ( <2 mm in diameter) for all species was 1944. Forest planting and soil biology. - Svenska significantly larger in the than in the 0-plots, as a IF- Skogsvardsforen. Tidskr. 5 :334-335. result of the improved moisture and nutrient con­ Buchholz, F. & Neumann, E. 1964. Der Einfluss der ditions. Since fewer samplings were sorted from the l­ Bodenbearbeitung und Diingung auf den and -plots, comparison with the conditions in the 0- F Wasserhaushalt und die Durchwurzelung sandiger plots is not valid and no attempt has been made to Waldboden Brandenburgs. - Albrecht Thaer Arch. calculate the annual belowground production. The 8:525-536. period of experimental treatments was also probably Caldwell, M.M. 1979. Root structure; The considerable too short in the I- and -plots for unmistakable cost of belowground fu nction. - Solbrig, Q.T., Jain, S., F changes in the ramification patterns of the short­ Johnsson, G.B. & Raven, P.H. (eds.). Topics in Plant Population Biology. London, pp. 408-427. roots to be observed. However, very low estimates of Deans, J .D. 1979. Fluctuations of the soil environment the number of root tips were obtained in the and I-, F- and fine root growth in a young Sitka spruce planta­ -plots on the last sampling occasion during 197 7, IF tion. - PI. Soil 52:1 95-208. at the same time that a fairly large estimate was Ebeling, F.E. 1978. Nordsvenska skogstyper. - Svenska arrived at in the 0-plots. Skogsvardsforb. Tidskr. 4:340-3 81. Ericsson, A. & Persson, H. 1980. Seasonal changes in starch reserves and growth of fineroots of 20-year-old Acknowledgement Scots pines. - Ecol. Bull. 32 (in press). I wish to thank J.G.K. Flower-Eilis for taking an active Flower-Ellis, J., Albrektsson, A. & Olsson, L. 1976. Struc­ part in the planning of this work; H. Sjors for providing ture and growth of some young Scots pine stands: (1) laboratory facilities at the Institute of Ecological Botany Dimensional and numerical relationships. - Swed. and, together with C.O. Tamm, for general interest and Con if. For. Proj. Tech. Rep. 3:1-98. support; M. Dahlgren, B. Andersson and I. Asplund for Flower-Ellis, J.G.K. & Persson, H. 1980. Investigations of valuable technical assistance; G. Agren, A. Lindgren and structural properties and dynamics of Scots pine K.B. Sundstrom for statistical advice and computer stands. - Ecol. Bull. 32 (in press). programming; and F. Andersson, B. Axelsson and E. Ford, E. D. & Deans, J.D. 1977. Growth of a Sitka spruce Sjogren fo r making useful comments on the manuscript. plantation : spatial distribution and seasonal fluc­ The work was carried out within the Swedish tuations of lengths, weights and carbohydrate concen­ Coniferous Forest Project supported by the Swedish trations of fm e roots. - PI. Soil 47:463-485. Natural Science Research Council, the Swedish National Harley, J.L. 1969. The biology of mycorrhiza. - London. Environmental Protection Board, the Swedish Council of 334 pp. Forestry and Agricultural Research and the Wallenberg Harris, W.F., Santantonio, D. & McGinty, D. (in press). Foundation. The dynamic belowground ecosystem. - Environmen­ tal Sciences Div., Oak Ridge Nat. Laboratory, Pub!. 1459:1-28. Hatch, A.B. 193 7. The physical basis of mycotrophy in Pinus. - Black Rock Forest Bull. 6:1-168. REFERENCES Hatch, A.B. & Doak, K.D. 1933. Mycorrhizal and other Agren, G.l., Axelsson, B., Flower-Ellis, J.G.K., Linder, S., fe atures of the root system of Pinus. - J. Arnold Arbr. Persson, H., Staaf, H. & Troeng, E. 1980. Annual car­ 14:85-99. bon budget fo r a young Scots pine. - Ecol. Bull. Heinsdorf, D. 1966. Uber die Feinwurzelverteilung in 32 (in press). verschiedenen Sandboden unter Kiefernkulturen. Aronsson, A., Elowsson, S. & lngestad, T. 1977. Elimina­ Arch. Forstwes. 15:591-603. tion of water and mineral nutrition as limiting factors in Hermann, R.K. 1977. Growth and production of the a young Scots pine stand. 1. Experimental design and roots: a review. - Fort Collins: Range Science Dept. some preliminary results. - Swed. Conif. For. Proj. Sci. Ser. 25 :7-28. Tech. Rep. 10:1-38. lngestad, T. 1967. Methods fo r uniform optimum fe rtiliza­ Axelsson, B. & Brakenhielm, S. 1980. Investigation sites of tion of forest tree plants. - 14th IUFRO Congress, the Swedish Coniferous Forest Project-biological and Miinchen Sect. 22:265-269. physiographical features. - Ecol. Bull. 32 (in press). Kinman, C.F. 1932. A preliminary report on root growth Babe!, U. 1970. Gliederung und Beschreibung des studies with some orchard trees. - Proc. Am. Soc. hor­ Humusprofils in mitteleuropaischen Waldern. - Mitt. tic. Sci. 29:220-224. dt. bodenkundl. Ges. 10:289-293. Kolesnikov, V.A. 1968. Cyclic renewal of roots in fruit

Bjorkman, E. 1941. Die Ausbildung und Frequenz der plants. - Ghilarov, M.S., Kovda, V.A., Novichkova-

A eta Phytogeogr. Suec. 68 110 H ans Persson

Ivanova, L.N., Rodin, L.E. & Sveshnikova, V.M. - 1980b. Death and replacement of fine-roots of a mature (eds.). Methods of Productivity Studies in Root Scots pine stand. - Ecol. Bull. 32 (in press). Systems and Rhizosphere Organisms. - Int. Symp. Reid, C.P.P. 1979. Mycorrhizae and water stress. - USSR, USSR Academy of Sciences. Leningrad, pp. Riedacker, A. & Gagnaire-Michard, J. (eds.). 102- 106. Physiologie des Racines et Symbioses. C.R. des Lid J. 1974. Norsk og svensk flora. - Oslo. 808 pp. Reunions du Groupe d'Etude des Racines. Nancy pp. Melin, E. 1922. Boletus-Arten als Mykorrhizenpilze der 392-408. Waldbaume. - Ber. dt. bot. Ges. 40:94-97. Reynolds, E.R.C. 1970. Root distribution and the cause of - 1953. Physiology of mycorrhizal relations in plants. ­ its spatial variability in Pseudotsuga taxifolia (Poir.) A. Rev. PI. Physiol. 4:325-346. Britt. - PI. Soil 32:501-5 17. Newman, E.I. 1966. A method of estimating the total Ruehle, J.L. & Marx, D.H. 1979. Fiber, fo od, fuel, and length of root in a sample. -J. appl. Ecol. 3:139-145. fu ngal symbionts. - Sciences 206:419-206. Orlov, A.J. 1955. The role of feeding roots of fo rest vegeta­ Safford, L.O. 1974. Effect fe rtilization on biomass and tion in enriching soils with organic matter. - nutrient content of fine roots in a beech-birch-maple Pochvovedenie 1955 (6):1-14. stand. - PI. Soil 40:349-363. 1968. Development and life duration of the pine feeding Santantonio, D. 1979. Seasonal dymanics of fine roots in roots. - Ghilarov, M.S., Kovda, V.A., Novichkova­ mature stands of Douglas-fir of different water Ivanova, L.N., Rodin, L.E. & Sveshnikova, V.M. regimes-a preliminary report. - Riedacker, A. & (eds.). Methods of Productivity Studies in Root Sys­ Gagnaire-Michard, J. (eds.). Physiologie des Racines et tems and Rhizosphere Organisms, Int. Symp. USSR, Symbioses, C.R. des Reunions du Groupe d'Etude des USSR Academy of Sciences. Leningrad, pp. 139-145. Racines. Nancy, pp. 190-203. Ovington, J.D. & Murray, G. 1968. Seasonal periodicity Snedecor, G.W. & Cochran, W.G. 1968. Statistical of root growth of birch trees. - Ibid. pp. 146-1 54. Methods, 6th ed. - Ames, Iowa. 593 pp. Paavilainen, E. 1967. The effect of fe rtilization on the root Sonn, S.W. 1960. Der Einfluss des Waldes auf die Boden. systems of swamp pine stands. - Folia Forestalia - Jena, 166 pp. 31:1-9. Sutton, R.F. 1969. Form and development of conifer root 1968. Root studies at the Kivisuo fo rest fertilization systems. - Commonw. Agr. Bur., Techn. Communs area. - Communs. Inst. forestal. fe nn. 66(1): 1-31. 7:1-131. 1974. Die Einwirkung der Dungung auf die Tamm, C.O. 1979. Nutrient cycling and productivity of Wurzelverhaltnisse der Kiefer auf Moorboden. - HafT­ fo rest ecosystems. - Leaf, A.L. (ed.). Proceedings. Im­ man, G. (ed.). Ecology and Physiology of Root pact of Intensive Harvesting of Forest Nutrient Cycl­ Growth, II. International Symposium. Berlin, pp. ing. State Univ. of New York, College Environmental 255-26 1. Science and Forestry, Syracuse, N.Y. pp. 2-21. Persson, H. 1975. Deciduous woodland at Anderby, Tennant, D. 1975. A test of modified line intersect method Eastern Sweden. Field-layer and below-ground produc­ of estimating root length. - J. Ecol. 63:995-1001. tion. - Acta phytogeogr. suec. 62:1 -72. White, E.H., Prichett, W.L. & Robertson, K. 1971. Slash 1978. Root dynamics in a young Scots pine stand in pine biomass and nutrient conditions. - Young, H.E. Central Sweden. - Oikos 30:508-5 19. (ed.). Forest Biomass Studies. Maine at Orono pp. 1979a. The possible outcomes and limitations of 165-176. measuring quantitative changes in plant cover on per­ Worley, J.F. & Hacskaylo, E. 1959. The effectof available manent plots. - N aturvardsverket Rapp. SNV PM soil moisture on the mycorrhizal association of Virginia 1151:81-87. pine. - For. Sci. 5:267-268. 1979b. Fine-root production, mortality and decomposi­ Wilcox, H.E. 1968. Morphological studies of the roots of tion in fo rest ecosystems. - Vegetatio 41:101-109. red pine, Pin us resinosa 2. Fungal colonization of roots 1980a. Spatial distribution of fine-root growth, mortali­ and the development of mycorrhizae. - Am. J. Bot. ty and decomposition in a young Scots pine stand in 55:686-700. Central Sweden. - Oikos 34:77-87. Zak, B. 1964. Role of mycorrhiza in root disease. - A. Rev. Phytopathology 2:377-392.

Hans Persson, Institute of Ecological Botany, Uppsala University, Box 559, S-75 1 22 U ppsala, Sweden

A eta Phytogeogr. Suec. 68 V egetationsentwicklung in einem W eidegebiet auf Siid-Oland, Schweden Lars Rodenborg

In den J ahren 1965-1 967 veroffentlichte ich eine Abhand­ N und W an andere Teile des Haines. Der Berggrund lung uber Flora und Vegetation in und bei dem Haine von aus Kalkgestein tritt an einigen kleinen Flecken im 0- Albrunna (,Albrunna lund") im Suden der Inset Gland Teil zutage. Diinne Erdlager, bis 20 cm Tiefe, decken (Rodenborg 1965, 1966, 1967). Im sudostlichen Teil des den Hauptteil des UG im Gegen W nimmt die 0. Haines von Albrunna wurde ein Gebiet von ungefahr 8000 Machtigkeit der Erdlager erheblich zu. Eine Beschrei­ m2 (das Spezialgebiet) fur nahere Untersuchung abgeschie­ bung der physikalischen und chemischen Bodenver­ den. Im ersten Teil der genannten Abhandlung (Roden­ borg 1965) wurden die Vegetationsverhaltnisse des Spe­ haltnisse des Haines von Albrunna ist bei Rodenborg zialgebietes wahrend der Jahre 1957-1 965 unter Ruck­ (1976 S. 42-44) zu finden. sicht auf die Bodennutzung beschrieben. In Tabell I wur­ Das Temperaturklima Olands ist maritim, die Nie­ den die Gefasspflanzen des Gebietes nebst einigen Moos­ derschlage sind gering (400-500 mm im Jahr im all­ und Flechtenarten auf die Vegetationstypen des Gebietes gemeinen). Mit Riicksicht auf die Humiditat wird verteilt, mit Angabe der Abundanz-Dominanz und in vie­ Oland als ein subarides Gebiet bezeichnet. Fiir nahe­ len Fallen auch mit Angabe des Verhaltens der betreffe n­ re Angaben iiber das Klima Olands sei auf Roden­ den Art zum Weidefaktor. Wahrend der Jahre 1966- 1979 borg (1976 S. 45 fT.)mit da angefiihrterLiteratur ver­ wurde die Entwicklung des Pflanzenlebensdes Spezialge­ wiesen. bietes weiter beobachtet. In der vorliegenden Arbeit werden die Ergebnisse dieser weiteren Beobachtungen beschrieben unter Vergleich mit Bodennutzung den Verhaltnissen in den J ahren 1957 -1965. Fur das Spe­ zialgebiet wird unten das Wort Untersuchungsgebiet (UG) Bodennutzung in vorigen Jahrhunderten benutzt. Mit Bezug auf Artabgrenzung und Nomenklatur Im Archiv des Vermessungsamtes (Lantmateristyrel­ sei auf Rodenborg ( 1965 S. 411 f.) hingewiesen. sens Arkiv) fm den sich fo lgende Karten beziiglich des Dorfes Albrunna: Geometrische Vermessung 1641, 1682, 1 7 34, 1771, Lage, Boden und Klima Teilung (,Enskifte och abodelning") 1816. Der sogenannte Hain von Albrunna, der dem nord­ Nach der Karte 1641 scheint der ganze Hain von westlich davon liegenden Dorf Albrunna gehort, ist in Albrunna dann Wiesenland (,Hardhwalds Engh") der Gemeinde Sodra Mockleby im Siidwesten von gewesen zu sein. Auf der Karte von 1682 wird der N­ Oland gelegen. Er grenzt im Osten unmittelbar an das Teil des Haines als junger Eichenwald (,ung Ek­ Gross-Alvar, das grosse siidoUindische Kalkgestein­ skog") bezeichnet und mit Bezeichnungen fiir Baume plateau mit meistens mehr oder weniger diinnem Erd­ versehen; der iibrige Teil wird als eine etwas gereutete lager und mit einer eigenartigen Flora und Vegeta­ Wiese beschrieben (,Engen ar nagot rojdh men tion, worin weit getrennte Florenelemente und Pflan­ mackta skarp, och bar foga ofwer N agot mulbete") zengesellschaften sich treffen, unter anderem solche und ist mit Baumbezeichnungen nicht versehen. mit Ankniipfung an die siidosteuropaisch-siidsibi­ Nach der Karte von 1734 scheint der ganze Hain rischen Steppengebiete. Der Hain von Albrunna ist Wiesenland (,Ang och Hobo!") gewesen zu sein, zum nur zum Teil ein Hain im gewohnlichen Sinn des Teil jedoch mit Eiche und Esche bewachsen (,Ang W ortes; da gibt es auch steppenhafte und wiesen­ uti gardet som en del med Ek Ask och Torn ar artige Vegetation, mit Gebiischen untermischt. Der bevaxt"' ,skog ails intet, undantagandes nagra ekar i Hain wird von Steinmauern begrenzt, die meistens gardet, som ej far huggas eller utodas"). Auch auf der aus Kalkblocken bestehen. Karte von 1771 werden die dem jetztigen Hain ent­ Das UG, das dem Grundstiick Albrunna 1:23 sprechenden Gebiete als Wiese bezeichnet, im N je­ entspricht, aus einem Teil des GrundstiicksAlb runna doch mit Baumzeichen, im S ohne Baumzeichen. Als 1:3 gebildet, grenzt im Osten an das Gross-Alvar, im Wiesenland wird das Haingebiet auch auf der Karte

A eta Phytogeogr. Suec. 68 112 Lars Rodenborg von 1816 bezeiehnet, mit Ausnahme eines Teiles im Andere okologisehe Faktoren (Klima usw.) haben N des Haines der als Aeker bezeiehnet wurde. Das sieh wahrend dieser Zeit im Mittel nieht erheblieh UG diirftedem Gebiet ,Batsmansangen" dieser Kar­ verandert. Die Bodenverhaltnisse sind natiirlieh von te wenigstens zum Teil entspreehen (auf der Karte den Veranderungen des Weidefaktors beeinflusst, von 1771 ,Batsmansangarne"). z.B. Abnahme bzw. Aufhoren des Tritts und der W ahrend seiner olandisehen Rei se im J ahre 1741 Diingung der Weidetiere. besuehte Linne den Hain von Albrunna am 17. Juni. Im Reiseberieht wird der Hain als ein sehoner Ei­ Arten und Artengruppen ehenhain erwahnt, in welehem Laserp itium latifo­ lium, A nemone hepatica, Pulmonaria of ficina/is, A Zu den auffallendsten Veranderungen in der Vegeta­ 1- liaria petiolata, Astragalus glycyphyllus, Viola hirta tion wahrend dieser J ahre gehort das sehr starke Ab­ und V. mirabilis wuehsen (Linne 1745 S. 86 f.). Die nehmen und in mehreren Fallen aueh das ganzliehe genannten Arten waehsen da immer noeh, im N-Teil Versehwinden annueller Arten. Das kommt am deut­ des Haines. liehsten zum Ausdruek mit Bezug auf die fr iiher sehr zahlreiehen, den Friihlingsaspekt der Pflanzengesell­ sehaften des flaehgriindigen Kalkfelsenbodens oft ± Bodennutzung in diesem Jahrhundert dominierenden Friihlingstherophyten, wie Erophila N_aeh Angaben der Besitzer dieses Teiles des Haines verna, Hornungia petraea, Cerastium spp., My osotis ist das UG jedenfalls nieht seit dem Anfang dieses spp., Veronica arvensis. Die meistens zufalligen und J ahrhunderts als Wiesenland benutzt worden son­ individuenarmen Sommertherophyten, iiberwiegend dern, zusammen mit einem anderen, angrenzenden Unkraut- und Ruderalpflanzen, sind jetzt fast ganz Gebiet, nur als Weide unter dem gemeinsamen Na­ versehwunden, z.B. Polygonum aviculare and Gera­ men Oehsenweide (,Oxhagen"). Bis zur Zeit des er­ nium pusillum. Einzige Ausnahme der allgemeinen sten Weltkrieges weideten namlieh Oehsen da. Dann Tendenz zur Abnahme oder Versehwinden der an­ wurde das UG als Weide fiir Pferde bis urn 1945 be­ nuellen Arten bei Abnahme oder Aufhoren der Be­ nutzt. Danaeh war es Weide fiir Rinder (Kiihe, Far­ weidung seheint das hemiparasitisehe, in massig sen, Kalber). Dagegen fand keine Beweidung mit xerophytiseher Saumvegetation waehsende Melam­ Sehafen da statt; sie war auf das Gross-Alvar be­ pyrum cristatum zu sein. Die nordisehen Melam­ sehrankt. Oehsen und Pferde, 2 beziehungsweise 7-8 pyrum - Arten seheinen iiberhaupt weideempfindlieh an der Zahl, weideten auf dem UG den Sommer hin­ zu sein. dureh. Die Rinder dagegen wahrend zwei ziemlieh Aueh mehrere Rosettenpflanzen haben als Folge kurzer Perioden im Jahr, teils 2-3 Woehen mit An­ der Abnahme bzw. des Aufhorens der Beweidung fang am Mittsommer, teils einiger Woehen im Sep­ kraftig abgenommen oder sind ganz versehwunden. tember. Die Anzahl der Weidetiere war seit etwa Sie sind natiirlieh im Kampf urns Lieht in hoher,

1950 8-9 (Kiihe), im · September ausserdem 5-6 nieht abgeweideter Vegetation unterlegen. Beispiele Jungtiere. In den Jahren 1957-1962 war ein Gebiet sind Primula fa rinosa, Plantago maritima, Cirsium von ea 3000 m2 auf dem UG umzaunt und vor Be­ acaule, Hypochoeris maculata, Leontodon autumna­ weidung ganz gesehiitzt. Der iibrige Teil des UG lis. wurde dagegen wie gewohnlieh beweidet. Danaeh Im allgemeinen sind niedrigwaehsende Arten wah­ wurde das ganze UG ungefahr wie vorher beweidet rend der besproehenen Zeit zuriiekgegangen oder ± bis zum Jahre 1973. In den darauf fo lgenden Jahren versehwunden. Viele dieser Pflanzen sind ja zugleieh (1974- 1979) ist das UG nieht beweidet worden. Zu Annuellen und/oder Rosettenpflanzen. Beispiele, bemerken ist also dass die oben erwahnte Beweidung nieht oben erwahnt, sind Carex panicea, C. caryo­ seit dem Anfang des Jahrhunderts nieht auf das UG phyllea, C. ericetorum, Luzula campestris, Polygala besehrankt war sondern ganz ,Oxhagen" umfasste, amarella, Viola rupestris, Antennaria dioeca (vgl. ein Gebiet von insgesamt ea 4 ha, wovon das UG Rodenborg 1967, S. 355). fo lglieh einen kleineren Teil betrug. V on den 6 Orehideenarten des UG sind Orchis us­ tulata, moria, Platanthera bifolia und Gymnade­ 0. nia conopsea deutlieh weidebegiinstigt und sind darum in den letzten Jahren erheblieh zuriiekgegang­ Vegetationsentwicklung en (op. eit. S. 358 ff.). Die Vegetationsentwieklung des UG wahrend der Calluna vulgaris ist seit 1965 auf dem UG nieht Jahre 1957-1979 steht wesentlieh unter dem Ein­ beobaehtet worden (vgl. dazu Rodenborg 1967 S. fluss der in diesen Jahren abnehmenden und zum 358). Sehluss ganz aufhorenden Beweidung des Gebietes. Auf dem flaehgriindigenKa lkfelsboden haben sieh

Ac1a Phylogeogr. Suec. 68 Vegetationsentwicklung in einem Weidegebiet auf Siid-Oland 113 mit der Abnahme der Beweidung Strauchflechten, achtungszeit kraftig verbreitet und bildet an einigen vor allem Cladonia rangiformis und C.furcata, aus­ Stellen grosse, zusammenhangende Bestande. Auch gebreitet, auf Kosten von Moosen und annuellen Ge­ Rubus Sekt. Corylifolii hat sich da verbreitet. fasspflanzen. Quercus robur hat mit der Abnahme der Bewei­ Arten mit vegetativer Vermehrung sind von der dung stetig an Zahl zugenommen und zwar schon vor Abnahme und dem Aufhoren der Beweidung deutlich dem Anfang meiner Beobachtungen im Jahre 1957. begiinstigt, z.B. Brachypodium pinnatum, Elytrigia Das gilt hauptsachlich auf dem mittleren und westli­ repens, Prunus spinosa, Rubus caesius, Rosa spp. chen Teil des UG mit tiefgriindigem Boden. Jetzt ist und Artemisia oelandica. da an mehreren Stellen, vor allem auf dem SW-Teil, Einige hochwiichsige Graminiden haben sich mit ein ziemlich dichter Gebiischwald m it junger Quercus der Abnahme der Beweidung stark verbreitet, ausser in der Baumschicht entwickelt. Im Jahre 1979 sind Brachypodium und Elytrigia der nachst vorherge­ auf dem flachgriindigen 0-Teil m it steppenartiger ± henden Gruppe, auch Arrhenatherum elatius und Vegetation viele Schosslinge von Quercus beobachtet Dacty/is glomerata (vgl. Rodenborg 1967, S. 356 worden. Fraxinus excelsior hat freilich auch zuge­ f.). Brachypodium, Elytrigia und Arrhenatherum elatius nommen aber bei weitem nicht so sehr wie Quercus. bilden oft ausgedehnte, dichte Bestande. Auf flachgriindigem Boden ist kein Schossling von Mehrere perenne Xerothermen mit siidostlicher Fraxinus angetroffen. Hauptverbreitung und/oder mit nahe verwandten Ar­ Betula verrucosa breitet sich in mesophiler bis ten von siidostlicherVe rbreitung halten seine Stellung schwach hygrophiler Vegetation auf dem W-Teil des auf dem UG nach dem Aufhorender Beweidung oder VG aus. Da kommen jetzt sowohl Schosslinge als verbreiten sich sogar, wie Adonis verna/is, Aster lino­ mittelgrosse Baume dieser Art vor. Betula pubescens sy ris, Artemisia oelandica (vgl. oben) und A. campes­ dagegen zeigt auf dem UG keine Verbreitungsten­ tris. Dagegen ist die konkurrenzempfindlichePio nier­ denz; wie 1965 kommt noch 1979 nur ein Ex. der Art art Artemisia rupestris zuriickgegangen (op. cit. S. vor, an der W-Grenze des UG. Rhamnus fr angula 360 ff. ). hat in schwach hygrophiler Gebiischvegetation mit Agrimonia eupatoria und Geranium sanguineum dominanter Potentilla fr uticosa auf dem W-Teil des haben sich mit der abnehmenden Beweidung in UG zugenommen aber bei weitem nicht in dem Mas­ steppen- und wiesenartiger Vegetation kraftig ver­ se wie es Betula verrucosa getan hat. Betula verruca­ breitet und bauen da eine xerophile Saumvegeta­ sa scheint iibrigens mehr offene, wiesenartige Vegeta­ ± tion auf(Origanetalia; Geranion sanguinei und Trifo­ tion vorzusiehen als Rhamnusfrangula die nur in Ge­ /ion medii). Geranium sanguineum bildet an mehre­ biischvegetation auftritt. ren Stellen im Haine von Albrunna, nachdem die Wie aus Tabelle I hervorgeht, hat die Anzahl von Weide aufgehort hat, ausgedehnte, dichte Bestande. Gefasspflanzenarten auf dem UG von 238 im Jahre Kennzeichnend fiir die Vegetationsentwicklung 1965 zu 199 im Jahre 1979 abgenommen. Fast ohne auf dem UG in den letzten Jahren ist die Verbreitung Ausnahme diirftediese Veranderung in der Artenzahl oder vielleicht auch Einwanderung einiger Hainarten. der Abnahme bzw. des Aufhorens der Beweidung im Paris quadrifolia ist jetzt erheblich haufiger als vor­ Jahre 1973 zuzuschreiben sein. Unter den 1965 an­ her. Anemone nemorosa ist wahrscheinlich neu auf getroffenen annuellen Arten sind 16 im Jahre 1979 dem UG; jedenfalls ist sie nicht vor dem Jahre 1977 nicht wiedergefunden. gesehen. Wahrend der Jahre 1966-1979 sind insgesamt 19 Auf dem 0-Teil des UG, auf flachgriindigem Gefasspflanzenarten erstmals beobachtet worden. ± Boden, bilden nunmehr, mit dem Aufhoren der Wei­ Mehrere davon sind zweifelsohne auch fr iiher auf de, Prunus spinosa und Rosa spp., vor allem R. du­ dem UG gewachsen aber sind da iibersehen worden, malis aber auch R. villosa und R. obtusifolia, niedrige wie Viola rupestris, V. odorata, V. canina, Hiera­ und lichte aber ausgedehnte, vegetativ sich vermeh­ cium peleteranum, Ta raxacum Sekt. Palustria, wohl rende Gebiische (vgl. oben). auch Globularia vulgaris. Andere Arten dagegen Auf dem mittleren und westlichen Teil des UG, mit sind nur zufallig, in einem oder einigen wenigen Jah­ tiefgriindigem Boden, werden nunmehr Juniperus ren, auf dem UG angetroffen; alle diese sind nur in und Potentilla fru ticosa von hochwiichsigen Laub­ den Jahren der Beweidung, 1966-1973, gesehen. baumen, vor allem Quercus aber auch Fraxinus, all­ Eine dritte Gruppe bilden Arten, die weideempfind­ mahlich unterdriickt.Das Gebiischgeht zu Gebiisch­ lich sind und wahrscheinlich erst mit der Schwachung wald iiber. oder dem ganzlichen Aufhoren des Weidedrucks in In dem schon friiher entwickelten Gebiischwald den letzten Beobachtungsjahren ins UG eingewan­ auf dem mittleren Teil des UG (Rodenborg, 1965 S. dert sind, wie Cy stopterisfragi/is, Anemone nemoro­ 439 f.) hat sich Rubus caesius wahrend der Beob- sa, Sorbus aucuparia (siehe Tab. I; in fine).

Acta Phytogeogr. Suec. 68 114 Lars Rodenborg

Pf la nzengesellschaften chis ustulata, die fruher in gewissen Jahren in bedeu­ tender Menge auftrat,kommt jetzt nur vereinzelt vor. In diesem Abschnitt werden die Vedinderungen Dagegen haben sich einige Graminiden der Gattung­ beschrieben, welche die einzelnen Pflanzengesell­ en Festuca, Poa und Phleum stark verbreitet vor al­ schaften des UG wahrend der fr aglichen Zeit betrof­ e Festuca ovina, Poa pratensis ang stifolia fe n haben. Die Gesellschaften werden wie bei Roden­ l m ssp. � Phleum phleoides. borg (1965) angefuhrt. Der deutschen Benennung je­ und Sie bilden an mehreren Stellen der Gesellschaft wird der lateinische Name der ent­ grosse, zusammenhangende Bestande, wo die meis­ sprechenden Einheiten (in der Regel Klasse und Ord­ ten anderen der fruher wachsenden Arten verdrangt Artemisia nung, in einigen Fallen auch Verband oder nur niedri­ werden. Die vegetativ sieh vermehrende oelandica Allium vineale Aster linosyris, gere Einheiten) im pflanzensoziologischen System sowie und alle drei Arten spatbluhend, sind auch vom Aufboren beigefugt (vgl. Rodenborg 1976 S. 15 ff.). Weder die der Beweidung begunstigt, wenigstens bis auf weite­ epiphytische Vegetation noch die Vegetation der Sili­ Adonis vernalis katblocke wird hier berucksichtigt. res. Auch einzelne Schosslinge von wachsen nunmehr in dieser Gesellschaft. Hier und da Vegetation auf erdfreiem Kalkgestein (Verruca­ dringen Saumarten, vor allem Geranium sangui­ 1. neum, rietea calcisedae Wirth 72 ad int., Xeroverrucarieta­ ein. Ausserdem verbreiten sich Rosa-Arten mit Rosa duma­ /ia Hadac 44). vegetativer Vermehrung, hauptsachlich lis R. villa­ Dieser Vegetationstypus war schon vorher nur aber auch niedrigwachsende Formen von sa R. obtusifo/ia, fr agmentarisch vertreten. Mehrere der im allgemei­ und und bilden allmahlich ausge­ dehnte, lichte Bestande. Auch Prunus spinosa breitet nen sehr kleinen Kalkfelsenflachen haben durch die Beschattung der sie umgebenden, nicht beweideten sich vegetativ aus. Fruher wurden diese vegetativ ge­ Prunus spi­ epigilischen Vegetation ihre Flechtenvegetation fast bildeten Schosslinge von Rosa-Arten und nosa ganz verloren. Im ubrigen aber haben die Verande­ von den Weidetieren aufgefressen. Es fm det also U rungen des Weidefaktors wahrend der verhaltniss­ an diesen Stellen ein bergang zu xerophiler Saum­ U massig kurzen Beobachtungszeit keinen Einfluss auf und Gebuschvegetation statt, eine bergangsvegeta­ diese epilithische Vegetation ausgeubt. tion die bei Rodenborg (1965 S. 424 fT.) Nanophane­ rophytensteppe genannt wurde. Th erophyten - Chamaephytensteppe (Kalkfelsen­ 2. Nanophanerophytensteppe (Ubergang zu xerophi­ rasen; Sedo-Sc/eranthetea Br.-Bl. 55 em. Th. Miill. 4. ler Saum- und Gebuschvegetation; Trifolio­ 61, Sedo-Sc/eranthetalia Br.-BI. 55, Alysso-Sedion Geranietea sanguinei Origanetalia Oberd. et Th. Mull. 61, Tortel/o-Sedion Hallberg Th. Muller 61, vulgaris Geranion sanguinei 1971 prov.). Th. Muller 61, Tx. 60, Querco-Fagetea Prunetalia sp i­ Kennzeichnend fur diese Gesellschaftist die krafti­ Br.-Bl. et Vlieger 37, ge Verbreitung von Strauchflechten, vor allem Clado­ nosae Tx. 52). U nia rangiformis und C. fu rcata, bei dem Aufboren Alte Teile dieser bergangsvegetation sind nun­ der Beweidung. Therophyten dagegen sind stark mehr als reine Saum- und Gebuschvegetation (xero­ zuruckgegangen. Chamaephyten, hauptsachlich Se­ philes Gebusch) zu bezeic hnen. Die Sukzession zu dum album, sind noch wie vorher haufig dominant. xerophilem Gebusch durfte jedoch im allgemeinen, Das nur in dieser Gesellschaft angetroffene A Ilium auf dem ziemlich flachgrundigen Boden uberKa lkge­ schoenoprasum (Rodenborg 1965 S. 420) ist nach stein, langsam gehen. Andererseits sehen wir, wie dem Aufborender Beweidung nicht mehr beobachtet. oben erwahnt, auf bedeutenden Teilen der Hemikryp­ Das Areal der Gesellschaft auf dem UG ist ungefahr tophytensteppe immer mehr Anfange von Saum- und dasselbe wie fruher. Gebuschvegetation. Da s Areal der N anophanero­ phytensteppe durfte also nunmehr erheblich grosser sein als fr uher. Die ausgesprochene Saumart Gera­ Hemikryptophytensteppe (Xerophiler Trockenra­ nium sanguineum bildet in dieser Gesellschaftau sge­ 3. sen ; Sedo-Scleranthetea Br.-Bl. 55 em. Th. Miill. dehnte, dichte Bestande und durfte hier das Schwer­ 61, Festuco-Sedetalia Tx. 50 em. Krausch 61, gewicht seiner Verbreitung auf dem VG haben. Auch Festuco-Brometea Br.-BI. et Tx. 43, Festucetalia Poa pratensis ssp. angustifolia, Fragaria viridis, val/esiacae Br.-Bl. et Tx. 43). Galium triandrum und Artemisia oelandica spielen In dieser Gesellschafthaben die Therophyten sehr im Aufbauen der Gesellschaft eine wichtige Rolle. In

kraftig abgenommen, mehr als im Kalkfelsenrasen · den Bestanden von Geranium sanguineum entwick­ stellenweise sind die Therophyten ganz verschwun � eln sich hier und da Gruppen von Orchis mascula. �en. Auch die niedrigwachsende Carex caryophyllea 1st fast ganz aus der Gesellschaftverschwunden. Or- Xerophi/es Gebiisch (Xerophile Sa urn- und Ge- 5.

A eta Phytogeogr. Suec. 68 Vegetationsentwicklung in einem Weidegebiet auf Siid-0/and 115 buschvegetation; Trifolio-Geranietea sanguinei Th. Saisonhygrophile Alvarheide (Festucetum alva­ 7. Muller 61, Origanetalia vulgaris Th. Muller 61, rense tortellosum Albertson 1950 ). Geranion sanguinei Tx. 60, Querco-Fagetea Br.-Bl. Dieser Typus von Vegetation ist, unter den Pflan­ et Vlieger 37, Prunetalia spinosae Tx. 52). zengesellschaften des UG, mit der Hemikryptophy­ Diese komplexe Vegetation (Saum und Gebusch) tensteppe am nachsten verwandt. Ausgesprochen sai­ hat sich wahrend der Beobachtungsjahre in qualitati­ sonhygrophile Alvarheide gibt es nur in einigen klei­ ver Hinsicht wenig verandert. Die Dominanz der we­ nen Niederungen, die ihre Entstehung den Weidetie­ nigen, kleinen Baume von Quercus und Fraxinus ist ren wahrscheinlich verdankt. Erheblich grosseres jetzt mehr ausgesprochen als fr uher. Die Straucher Areal nehmen Ubergange zur Hemikryptophyten­ (Rosa, Rubus, Lonicera u.s.w.) sind seitwarts, zum steppe ein. In den Niederungen, mit dunnen Erdlagen Teil in angrenzende Gesellschaften, vegetativ zuge­ iiber Kalkgestein, hat sich die Vegetation ziemlich wachsen. Die hochwuchsigen Graminiden Arrhena­ unverandert gehalten. Kennarten dieser Vegetation therum elatius und Elytrigia repens haben erheblich sind nunmehr Cerastium subtetrandrum, Sagina no­ zugenommen. Arrhenatherum elatius, Brachypo­ dosa, Herniaria glabra, Myosurus minimus, Barbula dium pinnatum, Geranium sanguineum und Origa­ reflexa, B. fa /lax, Riccia sorocarpa (erst 1979 an­ num vulgare sind wichtige Dominanten in der Feld­ getroffen aber wahrscheinlich schon fr uher vorhan­ schicht. Typische Saumarten sind, ausser Geranium den obgleich ubersehen). Die hauptsachlich auf sanguineum und Origanum vulgare, auch Viola hirta Viehpfaden wachsenden Poa annua und Plantago und Satureja vulgaris (beide haufig). Carex divulsa major sind jetzt verschwunden. Myosotis arvensis und To rilis japonica sind nunmehr wahrscheinlich kommt noch vor aber ist stark zuriickgegangen. ausgegangen, jedenfalls nicht in den letzten Jahren Veronica serpyllifolia, schon fruher nur in kleiner beobachtet. Das Areal dieser Vegetation ist durch die Anzahl, ist jetzt noch sparlicher. Weiterentwicklung von alteren Teilen der Nanopha­ nerophytensteppe zu reiner Saum- und Gebuschvege­ Mesophile Wiesenvegetation (Mesophiler Weide­ 8. tation erweitert worden. rasen; Molinio-A rrhenatheretea Tx. 37, Molinietalia W. Koch 26, Molinion W. Koch 26, Festuco­ Steppenartige Trockenwiese (Mesophiler Trock­ Brometea Br.-Bl. et Tx. 43, Festucetalia vallesiacae 6. enrasen (Halbtrockenrasen); Festucetalia vallesiacae Br.-Bl. et Tx. 43, Cirsio-Brachypodion Had. et Klika Br.-Bl. et Tx. 43, Cirsio-Brachypodion Had. et Klika 44). 44). Diese Wiesenvegetation ist nunmehr zum grossten Die steppenartige Trockenwiese ist im Zusammen­ Teil im Begriff zu entsprechender Saum- und Ge­ hang mit den Veranderungen der Bodennutzung buschvegetation uberzugehen. Am wenigsten verijn­ betrachtlichen Abanderungen in der Artenzusam­ dert erscheint sie im Kontakt mit der steppenartigen mensetzung unterworfen worden. Antennaria dioeca Trockenwiesenvegetation auf dem NO-Teil des UG. ist nach 1975 nicht gesehen. Carex ericetorum und U nter den iiberwiegend in diesem Vegetationstypus Anthoxanthum odoratum sind wesentlich zuruckge­ auf dem UG angetroffenen Arten (Rodenborg 1965 gangen. Orchis morio, ustulata, Polygala comosa S. 433) sind fo lgende in dieser Gesellschaft krrutig 0. und Cirsium acaule dagegen scheinen nicht erheblich zuruckgegangen oder ganz verschwunden: Sieglingia abgenommen zu haben. Haufigist immer noch Trifo­ decumbens, Cynosurus cristatus, Orchis morio lium montanum (Prruerenzart der Gesellschaft). Me­ ( verschwunden), Gymnadenia conopsea, Potentilla sophile Wiesen- oder Saumpflanzen wie A rrhenathe­ cranzii, Polygala amarella (nunmehr sehr sparlich), rum elatius, Dactylis glomerata, Elytrigia repens, P. comosa, Prune/la vulgaris, Chrysanthemum leu­ !nu/a salicina, Serratula tinctoria sind in die Gesell­ canthemum, Hypochoeris maculata (verschwunden). schaft eingedrungen bzw. haben sich da weiter ver­ Auch Linum catharticum (annuelle Art) ist stark breitet. Adonis vernalis und vor allem Geranium zuriickgegangen. Anthoxanthum odoratum ist n�n­ sanguineum haben zugenommen. A us xerophilen Ge­ mehr in dieser Gesellschaft nur sparlich angetroffen. buschen in der Umgebung sind Schosslinge von Pru­ Arten, die zugenommen haben, sind Brachypodium nus spinosa in die Trockenwiese eingedrungen und pinnatum, Agrimonia eupatoria, Vicia cracca, La­ leiten einen allmahlichen Ubergang der Wiese in thyrus pratensis, Inula salicina, Serratula tinctoria, xerophiles Gebusch nebst entsprechender Saumvege­ Potentilla fruticosa, Prunus spinosa, Betula verruca­ tation ein (vgl. Rodenborg 1965 S. 430). Gewisse Tei­ sa, Quercus robur. Den Charakter einer Sesleria­ le ehemaliger Trockenwiese sind schon zu wohlent­ Wiese haben immer noch einige kleinen Rasenfrag­ wickelter Saum- und Gebuschvegetation ubergegang­ mente im NO des UG die Ubergange zur angrenzen­ en. den steppenartigen Trockenwiese bilden.

Acta Phytogeogr. Suec. 68 116 Lars Rodenborg

Mesophiles Gebiisch (Mesophile Saum- und Ge­ Feuchtwiese getreten. Das Feuchtgebusch seinerseits 9. buschvegetation; Trifolio-Geranietea sanguinei Th. geht allmahlich zu Eichenmischwald (Quercetum Muller 61, Origanetalia vulgaris Th. Muller 61, Tri­ mixtum) uber. Fruher hatte dieses Gebusch ein mehr fo lion medii Th. Muller 61, Querco-Fagetea Br.-Bl. ausgesprochen hygrophiles Geprage. Nunmehr ist es et Vlieger 37, Prunetalia spinosae Tx. 52). nur schwach hygrophil und bildet einen Ubergang Wie aus dem Abschnitt uber die mesophile Wie­ zum mesophilen Gebusch. Die Artenzusammenset­ senvegetation hervorgeht, erweitert nunmehr die me­ zung des Feuchtgebusches ist wahrend der Beobach­ sophile Saum- und Gebuschvegetation in hohem tungsjahre bedeutenden Veranderungen unterworfen Masse sein Gebiet auf Kosten der vorigen Gesell­ worden. V on den nur oder uberwiegend in dieser Ge­ schaft. Die mesophile Gebuschvegetation ihrerseits sellschaft angetroffenen Arten (Rodenborg 1965 S. entwickelt sich allmahlich zu Eichenmischwald. Jetzt 437 f.) sind folgende nicht mehr beobachtet: Luzula gibt es in der Gebuschvegetation viele Schosslinge multiflora, Carex pallescens, Epilobium lamyi, Pim­ und Kleinbaume von Quercus. In der artenreichen pinella saxifraga, Scutellaria hastifolia, Galium uli­ Strauchschicht treten Prunus spinosa, Crataegus ginosum; Carex tomentosa und Scorz(i)nera humilis spp., Rosa spp., Potentilla fr uticosa, und Lonicera sind zuruckgegangen (Scorzonera jetzt fast nur vege­ xy losteum mengenmassig hervor. Ausserdem wach­ tativ fortlebend). Molinia coerulea durfte in dieser sen hier grosse Exemplare von Rhamnus cathartica Vegetation verschwunden sein. Prune/la grandiflora und Viburnum opulus. Typische Arten in der Feld­ hat stark abgenommen. Listera ovata dagegen ist schicht sind Agrimonia eupatoria (haufig), Lathyrus noch haufig.In der Strauchschicht ist Potentillafruti­ pratensis (haufig) und Trifolium medium (dom.). cosa immer noch der hauptsachliche Dominant mit Auch in dieser Gesellschaft ist Geranium sangui­ bedeutenden Einschlagen auch von Juniperus, Salix neum ein wichtiger Dominant. rosmarinifolia und Rosa majalis. In der Baumschicht dominieren kleine Baume von Quercus, die in spate­ Feuchtwiesenvegetation (Molinio-A rrhenathe­ ren J ahren erheblich zugenommen ha ben. Vereinzelt 10. retea Tx. 37, Molinietalia W. Koch 26, Molinion W. kommt Rhamnus fr angula vor. Koch 26). Diese Gesellschaft ist nunmehr ganzlich zu Meso-hygrophiler Gebiischwald (Querco­ 12. schwach hygrophiler bis mesophiler Saum- und Ge­ Fagetea Br.-Bl. et Vlieger 37, Prunetalia spinosae buschvegetation ubergegangen. Von den nur oder Tx. 52, Fagetalia sylvaticae Pawl. 28). uberwiegend in Feuchtwiesenvegetation angetroffe ­ Dieser Typus von Vegetation, auf ziemlich niedrig nen Arten sind folgende ausgegangen: Carex pani­ liegendem Boden auf dem mittleren Teil des UG, ist cea, C. hostiana, Platanthera bifo/ia, Primulafarino­ in der Beobachtungszeit keinen grossen Verande­ sa, Plantago maritima, also alle solche Arten der rungen unterworfen worden. Die Gesellschaft ent­ Feldschicht. Ubrig sind·nur Betula pubescens Ex.) wickelt sich nach und nach zu Hainvegetation. In der (I und B. verrucosa. Molinia coerulea ist sehr zuruckge­ Feldschicht kommt dieses u.a. darin zum Ausdruck gangen; nur einige wenige bluhende oder fr uchtende dass Paris quadrifolia nunmehr erheblich haufigerist Ex. sind in den letzten Jahren gesehen (vgl. Roden­ als vorher. Andererseits ist das nur im Gebuschwald borg 1965 S. 436-43 7). Uberhaupt ist also der wachsende Hypericum hirsutum nunmehr ver­ Feuchtwiesencharakter der Gesellschaft verloren schwunden. Potentilla fr uticosa wird von hohen gegangen. In der Feldschicht dominiert nunmehr Strauchern und von Baumen unterdruckt. Quercus haupsachlich Brachypodium pinnatum mit bedeuten­ und Fraxinus wachsen an Anzahl und Grosse zu, den EinschUigen von Listera ovata, Galium boreale, was die Untervegetation in verschiedener Hinsicht lnula salicina, Serratula tinctoria, Hieracium umbel­ beeinflusst (vgl. Potentilla fruticosa und Paris quad­ latum. Auf Kosten der Feuchtwiesenpflanzen der rifolia). Die alten Crataegus-Baume, die in einer offe­ Feldschicht hat Potentillafruticosa kraftig zugenom­ nen Wiesen- und Weidenlandschaft herangewachsen men. Auch Quercus hat sich verbreitet, einen all­ sind (Rodenborg 1967 S. 364 f.), werden allmahlich mahlichen Ubergang zu Eichenmischwald mit zer­ von schneller wachsenden Baumen unterdruckt und streuter Betula verrucosa einleitend. verdrangt. Wegen der dichten Strauchschicht ist die Feldschicht entweder sehr schattengeduldig oder Feuchtgebiisch (Schwach hygrophile Saum- und besteht aus Fruhlingspflanzen wie Ranunculus fica­ 11. Gebuschvegetation; Molinio-A rrhenatheretea Tx. ria (haufig). 37, Molinietalia W. Koch 26, Querco-Fagetea Br.-Bl. et Vlieger 37, Prunetalia spinosae Tx. 52). Hainvegetation (Querco-Fagetea Br.-Bl. et Vlie­ 13. Wie aus dem Abschnitt uber Feuchtwiesenvegeta­ ger 37, Fagetalia sylvaticae Pawl. 28). tion hervorgeht, hat Feuchtgebusch an die Stelle von Wie aus dem Abschnitt uber den Gebuschwald

A eta Phytogeogr. Suec. 68 Vegetationsentwicklung in einem Weidegebiet auf Siid-Oland 117 hervorgeht, erweitert die Hainvegetation allmahlich Physical and chemical soil conditions of Albrunna ihr Gebiet auf Kosten der vorigen Vegetation. Die lund were described by Rodenborg (1976 p. 42 fT. ). Hainvegetation ist jedoch immer noch nur fr agmen­ The temperature climate of Gland is maritime, the tarisch entwickelt. Es fehlt eine Reihe typischer, olan­ precipitation is low (in general 400-500 mm per discher Hainpflanzen: Gagea lutea, Anemone hepati­ year) and from the point of view of humidity Gland is ca, Corydalis cava, Mercurialis perennis, Viola mira­ characterized as a subarid area. bilis, Pulmonaria off icina/is, Adoxa moschatellina. As base for the knowledge of the land use of Al­ Anemone nemorosa wurde erst 1977 entdeckt; sie brunna lund in ancient times lie five land-surveyors' kommt sehr sparlich an nur einer Stelle vor, inmitten maps concerning the village of Albrunna, to which junger Eichen, und ist wahrscheinlich neuerdings Albrunna lund belongs. The maps date fr om the eingewandert. Anemone ranunculoides kommt years 1641, 1682, 1734, 1771 and 1816 respectively. gleichfalls nur an einer Stelle vor, in einem mesophi­ From the maps it appears that Albrunna lund then len Gebiisch,jedenf alls seit 196 1, dem ersten Beob­ was meadow-land but with a certain admixture of achtungsjahr, und wahrscheinlich viel friiher da Quercus and Fraxinus in the northern part. On his wachsend; sie zeigt keine Verbreitungstendenz. Unter journey through Gland 1741 Carl von Linne visited Strauch- und Baumarten, welche wohl entwickelte also Albrunna lund. He characterized it as a beautiful olandische Hainvegetation kennzeichnen, fehlen auf oak grove and, inter alia, recorded some typical grove dem UG Corylus avellana sowie Arten von Ulmus, plants as growing there. Acer und Tilia. Die einzige, ziemlich haufigvorkom­ In the 20th century the UG has not been used as mende, reine Hainpflanze in der Feldschicht ist Paris meadow-land but only as pasture-land, at first fo r quadrifolia. oxen and horses, since 1945 for cows, heifers and calves. Sheep did not graze there; sheep-grazing was confmed to Stora alvaret. Oxen and horses grazed on the UG throughout the summer; the cattle only SUMMARY during two rather short periods, one for 2-3 weeks During the years 1957 - 1 965 the author investigated beginning at midsummer, the other for some weeks in the flora and vegetation of the so-called Albrunna September. In 1957-1962 a part of the UG, a sub­ lund (grove of Albrunna) in the parish of Sodra area of about 3000 m 2, was entirely protected from Mockleby in the S.W. of the island of Gland in the grazing. In 1974-1979 no grazing has taken place Baltic (Rodenborg 1965, 1966, 1967). The investiga­ on any part of the UG. tion consisted partly of a thorough research of a mi­ Vegetational development in 1957- 1979 mainly nor area (the UG = Untersuchungsgebiet) of about depends, directly or indirectly, on changes that have 8000 m2 in the S.E. part of Albrunna lund, and partly taken place these years with regard to grazing, other of a survey of the remainder of the grove. The investi­ ecological factors are of secondary importance. gation of the UG has been continued in the years One of the most striking vegetational changes in 1966- 1 979. Results obtained as compared with the years in question is the strong decrease and in previous ones are accounted for in the present paper. many cases also the disappearance of annual species, The main purpose is to show the changes in the corn­ above all of spring-therophytes (Cerastium, Myoso­ position of the vegetation that have occurred during tis, Hornungia petraea etc.). Also several perennial the time in question against the background of former rosette-plants and most orchids, likewise low­ and present-day land use. growing species in general have diminished or disap­ Albrunna lund borders in the E. on the so-called peared. All these species then are negatively influ­ Stora alvaret, the great limestone plateau of Southern enced by the decrease and the ceasing of the grazing. Gland with mostly a thin earth-layer over limestone On the other hand some species are positively in­ and with a peculiar flora and vegetation, to a great fluenced by that development. To the last-mentioned extent composed of xerothermic species with a group of plants belong some fruticulous lichens (Cla­ ± continental or submediterranean distribution. Al­ donia spp.), vascular species with vegetative pro­ brunna lund is only in part a grove in the usual sense pagation, some perennial xerothermic plants, edge of the word, consisting partly of a steppe- or species ("Saumarten") as Agrimonia eupatoria and meadow-like vegetation with features of spiny thick­ Geranium sanguineum, shrubs as Prunus spinosa, ets. The UG, that in the E. borders on Stora alvaret, Rubus and Rosa spp., trees as Quercus robur, Fraxi­ is in the eastern part alvar-like with a thin earth-layer nus excelsior and Betula verrucosa. over limestone and a steppe-like vegetation. Else­ The changes in the vascular flora of the UG in the where it ha& deep earth-layer with a vegetation main­ years 1957-1979 are collocated, with complementa­ ly consisting of thickets and low-grown brushwood. ry comments, in Table I.

A eta Phytogeogr. Suec. 68 118 Lars Rodenborg

In a section regarding the plant-communities of the torum, Sesleria coerulea, Cirsium acaule. Fri.iher auch in VG the changes undergone by the communities mesophiler Wiesenvegetation. during the research period are briefly described. The Betula pubescens: Nur ein kleiner Baum in schwach plant communities have been delimited and denomi­ hygrophiler Gebi.ischvegetation an der W-Grenze des UG; u.c. Betula verrucosa, Quercus robur, Potentillafruticosa, nated according to Rodenborg (1965) with addition, Salix rosmarinifolia, Listera ovata, Serratula tinctoria. as a rule, of the Iatin names of the corresponding Quercus robur: Die Eiche hat wahrend der letzten 20 major phytosociological units. On the whole the Jahre auf dem UG erheblich zugenommen, hauptsachlich edge- and shrub-communities have strongly expand­ im W-Teil des Gebietes. Im Jahre 1979 sind ziemlich viele ed their area at the cost of the steppe- and meadow­ Schosslinge auch in der steppenartigen Vegetation des 0- like communities. Of the mesophilous meadow­ Teiles beobachteten worden. vegetation only small fragments remain. The hygro­ Urtica dioeca: An einigen Stellen in Gebi.ischwald im philous meadow-vegetation has nowadays complete­ mittleren Teil des UG. ly passed to the corresponding edge- and shrub­ Polygonum convolvulus: Beobachtungsj ahre 1960, vegetation. Mesophilous and hygrophilous shrub­ 1965; zufallig. Chenopodium album: Beobachtungsjahr I960; zufallig. vegetation on their part gradually develop into mixed Stellaria apetala : In xerophiler Gebi.ischvegetation auf oak wood (Quercetum mixtum). Of grove vegetation, dem 0-Teil des UG. however, still only fr agments have developed. The S. media : Beobachtungsjahre 1965, 1968; zufallig. xerophilous, steppe-like communities on a thin ± Cerastium subtetrandrum : I979 nur an I Fundstelle, earth-layer over limestone exhibit the same general einer kleinen Niederung in SO-Teil des UG; u.c. Herniaria successional trends as the remaining communities but glabra, Myosurus minimus, Sagina nodosa (vgl. Roden­ in the former ones the succession in the fo reseeable borg I965 S. 432). future probably will not surpass the stage of a low­ Sagina nodosa : I977, 1978, I979 nur an I Fundstelle, growing, rather thin shrub-vegetation with Rosa spp. einer kleinen Niederung im SO-Teil des UG (cfr. Ceras­ and Prunus spinosa as dominants, together with a tium subtetrandrum). Herniaria glabra : 1977 und 1979 nur an I Fundstelle, number of xerothermic herbaceous plants as Gera­ einer kleinen Niederung im SO-Teil des UG (cfr. Ceras­ nium sanguineum and Artemisia oelandica, a stage tium subtetrandrum). that is already attained on certain parts of the UG. Silene cucubalus: Nur 1 Fundstelle, in xerophiler Saum­ vegetation auf dem SO-Teil des UG. S. nutans: Nur 1 Fundstelle, in xerophiler Saumvegeta­ Ergiinzende Anmerkungen zu Tabelle I (u .c. = una tion auf dem SO-Teil des UG. cum; zusammen mit): Dianthus deltoides : An einer Stelle in steppenartiger Allium schoenoprasum: 1 Fundstelle. Vegetation auf dem 0-Teil des UG. Juncus alpinus: Nur an Viehpfaden. A nemo ne ranunculoides : N ur eine Fundstelle, in einem

Luzula multiflora : I m 1969 22 sprosse, an 3 Fundstel­ Gebi.isch von Lonicera xylosteum und Rhamnus catharti­ len. ca auf dem mittleren Teil des UG. Lolium perenne: Im 1973 an 1 Fundstelle, in steppenar­ Myosurus minimus: 1977, 1978 und 1979 nur an I tiger Vegetation. Fundstelle, einer kleinen Niederung im SO-Teil des UG; Poa annua: Im 1971 zerstreut an Viehpfaden. u.c. Poa bulbosa, Herniaria glabra, Sedum acre, S. album, P. pratensis : Ssp. angustifolia sehr reichlich. Ssp. irri­ Valerianella locusta. gata nur einmal, im Jahre 1966, beobachtet. Capsella bursa-pastoris : An einer Stelle in steppenarti­ Agrostis canina: Von ssp. montana vertreten. ger Vegetation auf dem 0-Teil des UG. Phleum pratense: Ssp. nodosum weitaus haufiger als Draba incana: Beobachtungsjahre 1965, 1968; zufallig. ssp. vulgare das nur vereinzelt vorkommt. Cardamine pratensis: Nur I Fundstelle, in einem Ge­ Carex hostiana: Nur 1 Fundstelle, in Feuchtwiesenfrag­ bi.isch mit Potentilla fr uticosa auf dem mittleren Teil des ment. UG. C. pallescens: Nur 1 Fundstelle. Sisymbrium supinum: Beobachtungsjahr 1965; zufallig. Platanthera bifolia : In meso-hygrophilem Gebi.isch­ Medicago fa lcata saliva: Beobachtungsjahre I96I, x wald mit Quercus robur, Potentillafruticosa, Listera ova­ I962; zufallig. la. Trifolium striatum: Sehr wechselnde Individuenzahl in Gy mnadenia conopsea : Sehr wechselnde Anzahl von verschiedenen Jahren. Haufig 196I, 1964, 197 1, I973 und bli.ihenden Exemplaren in verschiedenen Jahren. Doch 1976 (vgl. Sterner 1938 S. I19). deutliche Abnahme mit dem Aufhoren der Beweidung und T. arvense: I m 197 6 1 Ex. in steppenartiger Vegetation der zunehmenden Verwachsung. auf dem 0-Teil des UG. Orchis morio: Nach I974 nur an 1 Fundstelle, in step­ Geranium pusili um: Beobachtungsjahre 1961, 1969, penartiger Trockenwiesenvegetation; u.c. Orchis ustulata, 1974; zufallig. An einer Stelle in steppenartiger Vegetation Trifolium montanum, Anthyllis vulneraria, Prune/la gran­ nahe der 0-Mauer des UG; u.c. Bromus hordeaceus und diflora, Anemone pratensis, Carex caryophyllea, C. erice- Capsella bursa-pastoris.

A eta Phytogeogr. Suec. 68 Vegetationsentwicklung in einem Weidegebiet auf Siid-Oland 119

Erodium cicutarium : 1976 haufig, mit reichlichem Blu­ hendes oder fruchtendes Ex. gesehen, in mesophiler bis hen in ± steppenartiger Vegetation auf dem NO-Teil des schwach hygrophiler Gebuschvegetation in der SW-Ecke UG. des UG; u.c. Brachypodium pinnatum, Carex tomentosa, Hy pericum hirsutum: An einer Stelle in meso­ Listera ovata, Ranunculus polyanthemus, Inula salicina, hygrophilem Gebuschwald auf dem mittleren Teil des UG; Serratula tinctoria, Salix rosmarinifolia, Potentilla fruti­ u.c. Lonicera xylosteum, A lliaria petiolata, Geranium ro­ cosa. bertianum, Anthriscus silvestris, Serratula tinctoria. Leontodon autumnalis: Im 1966 > 15 Ex., an 10 Stel­ Ep ilobium lamyi: Beobachtungsjahre 1961, 1963; zu­ len. fallig. Nur an einer Stelle, in feuchtem Gebusch mit Poten­ Hieracium auricula : I m 1973 an 2 Stellen in mesophiler tilla fr uticosa. Wiesenvegetation auf dem NW-Teil des UG. Carum carvi: Beobachtungsjahre 1964, 1973; zufrulig. Viola rupestris: Fur diese Art wie fur die ubrigen erst im Im 1973 I Ex. in steppenartiger Trockenwiesenvegetation J ahre 1966 angetroffenen Arten siehe Rodeoborg ( 196 7 S. auf dem N-Teil des UG. 378). Pimpinella saxifraga : Im 1965 4 Fundstellen. V. canina : 1979 an einer Stelle, am Gebusch von Poten­ Heracleum sphondylium : Von ssp. sibiricum vertreten. t ilia fr uticosa auf dem NO-Teil des UG. Daucus carota: Im 1977 4 bluhende Ex. Globularia vulgaris : Im 1971 1 fertiles Ex. in steppenar­ Calluna vulgaris : In mesophiler bis schwach hygrophi­ tiger Trockenwiesenvegetation auf dem NO-Teil des UG. ler Vegetation. Lamium amplexicaule: Nur 1966 angetroffen. Ge­ Primula fa rinosa: Im 1970 1 bluhendes Ex. Nur in wohnlich nur als Unkraut in Garten und auf Ackern (Ster­ Feuchtwiesenvegetation auf dem NW-Teil des UG (vgl. ner 1938 S. 145). Rodenborg 1965 S. 436). Stachys arvensis: Nur 1966 angetroffen (vgl. Roden­ A ndrosace septentrionalis : In den J ahren 1967, 1968, borg 1961 S. 22 7). 1970, 1971 und 1976 ziemlich haufig (vgl. Rodenborg Hieracium peleteranum: Im 1971 an einer Stelle in step­ 1965 s. 420). penartiger Vegetation auf dem 0-Teil des UG. Lithospermum arvense: Beobachtungsjahre 1965, Viola arvensis: Beobachtungsj ahre 1968, 1970, 1976. 1974; zufallig. Im 1974 1 Ex. in steppenartiger Vegetation Im 197 6 15 Ex. in steppenartiger Vegetation auf dem NO­ auf dem NO-Teil des UG, u.c. Bromus hordeaceus, Sedum Teil des UG. acre, S. album. Ta raxacum sekt. Palustria: Nur 1969 angetroffen. Nur Myosotis arvensis: 1973 haufig. an einer Stelle. M. stricta: Im 1976 in steppenartiger Vegetation auf Trifolium hybridum: Beobachtungsjahre 1970, 1971, dem NO-Teil des UG. 1972. Im 1972 an einer Stelle auf dem S-Teil des UG, bei Cynoglossum of ficinale: Kommt nur ± nahe der 0- mesophiler bis schwach xerophiler Saumvegetation. Mauer des UG vor. Anagallis arvensis: Nur 1970 angetroffen. Nur an einer Prune/la vulgaris: I m J ahre 1971 ziemlich haufig. Stelle; 2 Ex. Veronica of ficina/is: Im 1973, in mesophiler bis Sonchus oleraceus: Nur 1970 angetroffen. Nur 1 Ex. schwach xerophiler Wiesenvegetation auf dem W-Teil des Veronica persica : Beobachtungsj ahre 197 1, 1972, UG. 1973, 1975. An einer Stelle auf dem NO-Teil des UG in Odontites verna: Beobachtungsjahre 1960, 1962; zufal­ steppenartiger Vegetation; u.c. Gagea pratensis, Saxifra­ lig. ga granulata, Filipendula vulgaris, Erodium cicutarium. Erigeron acre: Beobachtungsj ahre 1957, 1971; zufallig. Arctium tomentosum: Nur 1971 angetroffe n. Nur ein Im 1971 1 bluhendes Ex. in steppenartiger Vegetation auf steriler Spross in steppenartiger Trockenwiesenvegetation dem 0-Teil des UG. auf dem 0-Teil des UG; u.c. Sedum acre und Artemisia Antennaria dioeca : Im 1975 an 2 Stellen, die eine auf oelandica. dem NW-Teil, die andere auf dem NO-Teil des UG. Cerastium arvense: Beobachtungsjahre 1973, 1975. An Tripleurospermum maritimum: Von var. inodorum ver­ einer Stelle auf dem SO-Teil des UG, in xerophiler Saum­ treten. Beobachtungsjahre 1960, 1961, 1962, 1970. Im vegetation bei Prunus spinosa; u.c. A rrhenatherum elatius, 1970 insgesamt 6 bluhende Ex., an 3 Stellen. Filipendula vulgaris, Agrimonia eupatoria, Geranium Cirsium vulgare: Beobachtungsjahre 1959, 1962, 1963, sanguineum, Artemisia oelandica. 1973. Im 1973 1 fertiles Ex. in meso-hygrophilem Ge­ Viola odorata: An einer Stelle auf dem SO-Teil des UG, buschwald auf dem mittleren Teil des UG. Durfte ur­ in xerophiler Gebuschvegetation aus Juniperus, Prunus sprunglich durch Rodung an dieser Stelle eingekommen spinosa und Lonicera xy losteum; u.c. Ranunculus ficaria, sein (Rodenborg 1965 S. 456). Corydalis pumila und Primula veris. C. arvense: Wahrscheinlich durch Rodung eingekom­ Bellis perennis: 1973 an 3 Stellen, in steppenartiger men (Rodenborg 1965 S. 456). Vegetation auf dem NO-Teil und in schwach xerophiler Centaurea scabiosa : 1978 und 1979 1 bluhendes Ex., in Wiesenvegetation auf dem SO-Teil des UG. 1975 an 2 steppenartiger Trockenwiesenvegetation auf dem SO-Teil Stellen ; die eine Stelle Viehpfad. 1979 auf einem Pfade auf des UG; u.c. Brachypodium pinnatum, Adonis vernalis, dem N-Teil; u.c. Myosotis arvensis (vgl. Sterner 1938 S. Geranium sanguineum, Origanum vulgare, Centaurea 156-157). jacea, Serra tula tinctoria. Cystopterisfragilis: Nur an einer Stelle, in der SO-Ecke Hypochoeris maculata : 1977, 1978 und 1979 nur 1 blu- des UG, an der S-Mauer; schattige Lage; u.c. Arrhenathe-

Acta Phytogeogr. Suec. 68 120 Lars Rodenborg rum elatius, Polygonatum odoratum, Potentilla reptans, Tabe 11 e I. Gefasspfl anzenfl ora des UG wahrend der Jahre 1957-1979. Prunus spinosa, Rosa villosa, Rubus caesius, Viola hirta (ziemlich xerophile Saum- und Gebiischvegetation). Anemone nemorosa: An einer Stelle auf dem SW-Teil des UG, unter einem kleinen Baum von Quercus; u.c. Poly­ gonatum multiflorum, Listera ovata, Potentillafruticosa, Rosa majalis, Crataegus oxyacantha, Fraxinus excelsior, Primula veris. Sorbus aucuparia: Ein kleines Ex. in Gebiischvegeta­ 1957-1965 tion auf dem S-Teil des UG; u.c. Juniperus communis, Juni perus COITilluni s p + + + + x Quercus robur, Potentilla fr uticosa, Prunus spinosa und All ium schoenoprasum p 0 1973 A. oleraceum n + + + Lonicera xy losteum. A. vi nea 1 e n + + + Gage a pratens is + + + + G. minima + Po 1 ygona tu m odora tum n + + P. mu ltiflorum n Paris quadri fol ia n + + x LITERATUR VERZE ICHNIS Juncus a 1 pi nus p 0 1965 Luzu 1 a campes tri s pp x l. multifl ora 1969 Linne, C. von, 1745. Carl Linnaei Gliindska och Gotlands­ Siegl ingia decumbens pp ka Resa ar 1741 (red. av Carl-Otto von Sydow 1962). Mol inia coerulea p Sesleri a coerulea p + + +

.- Stockholm och Uppsala. Festuca ovina p + + + F. rubra + + + + + + Rodenborg, L. 1961. Floristiska anteckningar fran soora F. pratens is + + F. arundinacea Gland. - Svensk bot. Tidskr. 55:225-229. x lo 1 i um perenne 1973 x 1965. Flora och vegetation i och vid Albrunna 1und pa Poa annua p 1971 P. tri vialis sodra Gland. 1. - Svensk bot. Tidskr. 59:411-462. P. bu 1 bosa pp Xp. pra tens is 1966. idem 2. - Svensk bot. Tidskr. 60:255-292. P. compressa 1967. idem 3. - Svensk bot. Tidskr. 61:353-384. Bri za media p Cynosurus cri status p + + + 1976. Bodennutzung, Pflanzenwelt und ihre Verande­ Dactyl is glomerata n + + + Arrhena the rum e 1 a ti us n + + + rungen in einem alten Weidegebiet auf Mittel-Gland, A. pubescens 7:1-210. A. pratensc n Schweden. - Vaxtekol. Stud. Des champs i a caespi to sa Sterner, R. 1938. Flora der lnsel Gland. - Acta phyto­ Calamagrostis epigei os n Agros ti s gi gantea + + + + geogr. suec. 9:1-169. A. tenuis + + + + + x A. canina p + 0 1973 Ph 1 eum ph 1 eoi des + + + Xp. pra tense p + + + + + Anthoxanthum odoratum p + + Bromus hordeaceus pp + + + Brachypodi um pi nnatum nn + + + Elytrigia repens n + + Carex contigua + C. divulsa 0 1965 C. fl acca + + + + + + 0 1965 p 0 1971 C. �hi rta�:� x{ ��i / c. pa 11escens 1973 C. tomentosa + + + C. ea ryophyl l ea pp + + + C. eri cetorum pp + + x Platanthera bi folia x 1976 Gymnadeni a conopsea + + Drchis ustulata p + + x D. moria p + 0. mascula + + Listera ovata + + + Sal i x rosmarinifol ia + + + + + .Betu 1 a verrucas a nn + + s n + e r n + + + + + "Urti;�� ���� ea di���� oeca + Rumex acetosa + + Polygonum aviculare p 1970 Xp. convol vulus p x 1965 chenopodium album p 1960 Stel laria grami nea + + + + xs. media p x 0 1968 s. apetala p + 0 1976 Cerasti um holosteoi des + + xc. subtetrandrum p C. pumi lum pp C. semi decandrum pp xSagina nodosa p Arena ria serpyl lifolia pp xHerniaria glabra p Scleranthus po lycarpus 1974 Viscaria vulgaris x5i ene cucuba 1 us x 1 s. nu tans p x 1968 Di an thus de 1 toi des p 1973 fhalictrum simplex p X Anemone ranuncu 1 oi des + A. pratensis p + + Ranuncu 1 us auri comus + + R. acris + + + +

A eta Phytogeogr. Suec. 68 Vegetationsentwicklung in einem Weidegebiet auf Siid-Oland 121

Tabelle I (Fortsetzung) Tabelle I (Fortsetzung)

""' ""' "' "' a> a> N "' N ·� N "' C VI !.. c ,_ C VI !.. c ,_ tOQJ..CV'I "' a> .<::; re QJ..c;VI "' a> .<::; C:+-'ra.- c "' c "' ·� . V'l',...., tO ...., . ...., E ,_-" ...... VI .;:E � ,_ -�'; -" ...... VI 0 QJ O'd... a.>o.o O'l 0 Q) c::n s... a.>o.o O'l '0"' c '0 a> c 0 o� VI::> 0 o� VI::> 9 5!! OJ+' 9 ;� Q)+' OJQJ E E .<:::11!:1! .<::; �-5Q)"' .<::: .<::; �� "' "' "' "' c c c c ..orn QJcv .Cl:;, ..00'\ QJQJ � :� -g .Cl:;, c:t:-aJ :J: <( x x + + + + + Erodium cicutarium p + + H. auri cu1 a p 0 > 1973 L inum catharticum p + + + + + Polygala comosa p + + 1966 x P. vulgari s p 1973 Viola rupestri s 1970 + + x P. amarella pp v. canina x Euonymus europaeus n + + + + Globularia vulgaris 1971 x Rhamnus catharti ea + + + + + + lamium amplex icaule 1966 x R. frangula n Stachys arvens is 1966 xrlypericum hirsutum l g76 H. perforatum + + + l g67 + + + x Helianthemum nummularium p + + + + + + Hieracium pel eteranum p lg71 Viola hirta n + + + + + + xEpi labium lamyi 0 0 l g63 1968 x Chaerophyl lum temulum n + + + + Viola arvensis p 1976 1\nthriscus silvestri's n + + + + Tori lis japonica p + + 0 1977 1969 x x ca rum ea rvi 0 0 1973 Taraxacum Sekt. Pa1 ustri a p + 1969

Acta Phytogeogr. Suec. 68 122 Lars Rodenborg

Tabelle I. (Fortsetzung) (Tabe1 1e I. Erk1 arungen )

Erst werden die Arten , die wahrend der periode 1957-1965 U) \0 beobachtet warden sind, in taxonomi scher Rei henfo1 ge auf­ 0'> "' gezah1t. Dann werden die erst nach 1965 beobachteten Arten c: V'l s.... c:: tO OJ.r:=.Vl ttl "' aufgezah1t; jede so1che Art wi rd unter ihrem ers ten Beob­ c: ....., ro..- c • ,... Vl',..., tt:l .,...... ,., achtungsjahr aufgenommen . W\0 ""' � o� � p = we i debegiinstigt, pp = stark �te idebegiinsti gt ; [ w w � E E -'=.<= n = wei deempfi nd1ich , nn = sehr we i deempfind1ich (p, pp , ::8��· :-0'> "'"' >� c c n und nn beziehen sich auf a1 1 gemeine Erfahrungen auf .£)0'\ QJQJ .o, c(.-col�l3: = Aretium tomentosum p 1971 x gibt an dass die betreffende Art unter Erganzende AnRler­ 1973 x kungen zu Tabe11e 1 aufgenommen ist. Cerastium arvense p 1975 x Viola odorata n x Be llis perenni s pp

1976 x Cys topteri s fragi lis n

1�77 x Anemone nemorosa n x Sorbus aueupari a n

Gesamtzah l van Arten 265; Arten Arten >: 136 davon 246 bi s zum Jahre 1965: 1979 : <: 27 1965 angetroffen und 19 238 199 weder > ers t in den Jahren 1966- noeh< : -1977. 102

Anmerkung : Der Untersehi ed zwi sehen Gesamtartenzah l im Jahre 1965 und im Jahre 1979 betragt also 39, d.h. 16 % der Gesamtartenzahl im Jahre 1965.

Lars Rodenborg, Lappkarrsvagen 48, S- 1 04 05 Stockholm, Sweden

A eta Phytogeogr. Suec. 68 Plant Cover Regeneration 1n a Mechanically Disturbed Limestone Heath on Oland, Sweden Ej vind Rosen

In late January 1975 the Swedish Company Sydkraft Cambrian and Ordovician strata dip gently towards ESE (electric power distributor) started to renew an old (Konigsson 1968 p. 12). electric power line and to build a new one crossing an In the southern part of the island there is a limestone alvar area (limestone heath) E. of Albrunna village, heath, the so-called Stora Alvaret, which dominates the Oland. At that time the ground normally is frozen, landscape. Along the western borderline of the heath there is an escarpment, "klint" (cf. Martinsson 1958 p. 15). but on this occasion only the surface was frozen at In the area of Stora Alvaret there is a remarkable lack of night and in the mornings. The work started in the deeper soil. The topography is very flat. Shore deposits, morning and when the surface thawed the heavy weathered material and wind-transported soils dominate vehicles sunk in the water saturated soil and made the Quaternary deposits (Konigsson 1968 p. 17). The deep tracks over large parts of the area (Fig. 1 ). This deposits are unevenly distributed and in scattered large was seen by local people on February 2nd and a very areas the bedrock is exposed. As bland was repeatedly critical opinion was created. On February 4th the flooded and left dry in the different transgression stages of local newspaper, Barometern, ran an article on what the Baltic after the latest glaciation, it is difficult to had happened. The company was subjected to severe separate washed moraine areas from litoral deposits (op. criticism from both the public and the mass media, cit., p. 15). Fissures in the bedrock are common and significant in the drainage of the Alvar. The area of the particularly as the area was proposed to be a Nature heath is about 300 km2• Smaller alvar areas are fo und in Reserve. The work was stopped. Predictions were the central and northern parts of the island. For more made of the time needed fo r regeneration of the detailed information about Quaternary geology and vegetation. Periods of 25-50 years were mentioned. history see Konigsson (1968), including his Physiognomic The company promised to restore the area. map of the Great Alvar. Restoration is always included in project plans in case accidents like this occur. In this case they had to be particularly careful as the area was specific. No extra soil was allowed to be added, as it could have introduced seeds changing the composition of the vegetation. The restoration was done manually and when it was completed, in April, the area was inspected by nature conservancy authorities who fo und the work well done. Under the circumstances I fo und it in­ teresting to fo llow the regeneration of the plant cover in the area by means of samples taken annually from permanent plots.

Geology bland is one of the two big Swedish islands in the Baltic. The island is situated dose to the mainland. The bedrock is composed of various strata. On the sub-Cambrian Fig. I. The investigation area from the south, from a point peneplain there are pre-Cambrian formations upon which about 50 m south of the location of transects li and Ill. fo llow Cambrian and Ordovician strata (Konigsson 1968 Serious scars in the vegetation were created when heavy p. 12). The Ordovician strata consists of an older and a vehicles sunk in the alvar soil during work with a new elec­ younger Ortocer-limestone (Lundqvist 1955 p. 18 ff.). The tric power line in late January (photo taken in late March).

Acta Phytogeogr. Suec. 68 124 Ej vind Rosen

Climate lowest precipitation (327 mm) was in 1975 an� 979 h�d the highest. There are periods with low prec1p1�tat1on . m A broad review of climatic conditions in Oland is June-July 1975, in April-May 1978, and in Sep­ presented below. Winds from W or SW predominate dur­ tember-October 1979. More detailed information about ing the year (Ostman 1926 p. 15 f.), but in wintertime these periods were obtained from the Ecological Station of winds from NE are frequent. Precipitation values are lower Uppsala University, Olands Skogsby, but this is not an of­ for Oland than fo r the mainland, as a lot of rain is lost in ficial meteorological station. the high areas of southern Sweden. Data from the Ecological Station :- In the period During the period 1901-1930 the lowest values of 1975. June lst-July 14th only 6 mm precipitation was received. mean yearly precipitation (400-450 mm) were obtained in Then 15 mm during a 10-days period. From July 25th to coastal areas and from the northern and southern part of August 16th, 2 mm was measured and finallyfrom August the island. From the central part values around 500 mm 25th to Septemberllth, 4 mm. April 26th-June were received (Bergsten 1955 p. 36 ff.). - 1978. 9th there was 6 mm. September 22nd-November The following values fo r three meteorological stations in - 1979. 1st only 2 mm. - and Precipitation was more the southern part of the island (Morbylanga, Segerstad and 1976 1977. evenly distributed during the growing period than in 1975, Olands sodra udde) close to the investigation area were ob­ 1978, 1979. tained for the period 1965-1979 from data supplied by Irregular periods of drought occur and affect the vegeta­ SMHI (Swedish Meteorological and Hydrological In­ tion more or less. Especially in areas with thin and exposed stitute). Mean values fo r a period of 15 years and fo r the soils, as on the Alvar, the effects of drought are frequently investigation period (5 years) are shown in Table I. severe. Normally a low precipitation in May-June Separate values fo r each year during the pe iod of in­ � together with high insolation will create drought in vestigation are presented in Table 11. The locat1ons of t e � July-August. area and of the three meteorological stations are shown m There is also a rather high local variation in monthly Fig. 2. Calculations have also been made to show P ssible . ? precipitation in different parts of Oland, clearly presented mean values for the investigation area by combmat1ons of by Widen (1980 p. 102, Fig. 3). the values received fro m the three stations each month and The number of days with snow-cover is less than 40 fo r year during 1975-1 979 (Table Ill). the southern part of the island (Angstrom 1974 p. 70). The Table Ill shows a pronounced variation between years mist frequency during the year is 15-20 % of the days fo r as well as within years during the investigation period. The the southern part and 10-15 % for the northern part of the island (op. cit., p. 120). I. Table Mean yearly precipitation from three Regarding humidity, Oland is classified as subarid (op. me teoro logical stations in the southern part of . . i5land during the periods 196S-79 and 197S-79 . cit., p. 99). The mea.., number of sunshine hours at Olands

station 196S-79 197S-79

Morbylanga 486 4S3 Segerstad 471 440 Hlands sodra udde 46{1 423

Table II. Yearly precipitation for three me teorological stations in the southern part of Hland during 197S-1979.

station 197S 1976 1977 1978 1979

Morbytanga 314 491 47S 478 SOS Segerstad 322 424 4S9 463 S34 Hlands sodra udde 343 42S 461 428 4S7

Table III. Mont hly precipitat ion for southern 01and during the period 1975-79 calculated as mean va lues from three meteorological stations (HOrbyllnga, Segerstad and tilands sodra udde ).

J F H A H J J A S 0 N D year

1975 35 10 25 24 52 6 17 32 57 36 26 8 327

1976 33 8 33 49 41 35 40 25 41 21 33 88 447 1977 48 39 28 40 16 24 67 38 27 24 68 54 465 01. sodro udde 1978 36 43 55 18 9 25 59 35 94 14 23 45 456

1979 51 33 65 )) )) 34 54 59 20 3 65 49 499 Fig. 2. Location of the three official meteorological Hean for stations, the Uppsala University Ecological Sta�ion and the period : 41 27 41 33 30 25 47 38 48 20 43 49 439 . the investigation area at Albrunna in southern Oland.

Acra Phyrogeogr. Suec. 68 Regeneration in a mechanically disturbed limestone heath 125

. ,. Investigations The investigation area is located on the Alvar about 200 m E of the northern part of the deciduous wood "Albrunna lund". Close to the wood there is an in­ termediate zone with a mixture of heath and fo rest vegetation on rather deep soil with several weeds favoured by heavy grazing. East of that zone is a moraine ridge with numerous silicate stone deposits. This ridge is about one meter higher than the sur­ roundings. The area of investigation is located on the Alvar just east of the ridge. The ridge has fairly deep soil with scattered Juniperus communis, Prunus spinosa and some other shrubs. Shrubs are rare in the investigated area. The Albrunna lund and some parts of the Alvar areas close to it are described by Roden­ borg (1965, 1966, 1967). Of the transects mentioned in Methods, transect no. I is located on rather shallow soil (10-20 cm) and the original vegetation is classified as a type of Festucetum community, while transects no. 11 and no. Ill are located on deeper soil (20-40 cm) where Fig. 3. Regeneration of the plant cover started after Festucetum re toration of the area in April 1975. In September the vegetation is intermediate between the ame year permanent sample plots were distributed in the and the A venetum. area. The photo shows the area from the north, showing In the southern part, transect no. I ends close to a tran ect 11 with plot no. 6. Transect no. Ill was placed small depression in the Alvar, where vegetation is ab ut 0 m further to the south. 22.9. 1975. supplied with more water than in the rest of the I - transect. Therefore, e.g., Sesleria coerulea ssp. uli­ sodra udde is 1906, calculated for the period 1961-1977 ginosa is found in reference plot no. 4. (Widen 1980 p. 101). The influence of the Baltic creates a maritime climate for C>land (Angstrom 1974 p. 53). The coldest month is Methods February with mean values between -2°C and -1°C and the warmest month is July with values between 16°C and In the area with the restored surface three transects 1 7°C (Angstrom 1953 p. 3). For further information about (I-III) were distributed. They not only included dis­ climatic conditions see Sjogren ( 1961 p. 9 ff.)and Konigs­ turbed parts but also parts where the plant cover was son 1968 p. 12 ff.). ( not, or only little, disturbed. Three or fo ur sample plots were distributed along these transects, each plot Vegetation andflora with a size of 0.25 m2• The transects were permanent­ ly marked with sticks from which the position of the The Alvar vegetation is generally very mosaic and consists plots could be fixed by means of a measuring tape of several plant communities (associations) with a strong variety from one place to another. A venetum alvarense (Fig. 3). (Aibertson 1950 p. 304 ff.) is the predominant association During 1975-79 in September each year, all taxa on moraine ridges and in marginal parts of the heath. in these plots were recorded and their cover degree Festucetum communities predominate on shallow soil and determined according to the Hult-Sernander-Du Sedetum tortellosum on bare rock. In moist habitats Rietz five degree scale (Du Reitz 1921 p. 225). The Molinietum alvarense is fo und (op. cit., p. 284 ff.). recordings were made on September 22nd 1975, On Oland many foreign taxa occur as outposts more or Sept. 20th 1976, Sept. 17th 1977, Sept. 16th 1978 less isolated from their main distribution. In the Alvar we and Sept. 9th 1979. A few notes on frequent taxa find taxa such as Artemisia laciniata and Potentilla were also made during 1975-78 from the restored fr uticosa (both Siberian), Ranunculus illyricus (Southeast European), Poa alpina (Arctic-montane) and many others. tracks just to the south of transect Ill. As the record­ There are also taxa considered as endemic to Oland, such ings were made at the end of the growing season it as Helianthemum oelandicum which is very complex and was difficult to exactly determine species of some shows many variations, e.g, two flowering strategies genuses such as Cerastium, Potentilla and Ta rax­ (Widen 1980). For information about grazing and Alvar acum. I have chosen to treat them as Cerastium spp., vegetation see Rosen & Sjogren (1973) and Rosen (1980). Potentilla sp. and Ta raxacum spp. The genus

Acta Phytogeogr. Suec. 68 126 Ej vind Rosen

Fig. 4. Sample plot no. 6 was totally dominated by Sisym­ Fig. 5. Sample plot no. 6. After 5 years of regeneration the brium supinum. Bromus hordeaceus, Hornungia petraea, composition of vegetation in the plot was totally changed. Erophila verna and some other species also occurred but The plot is now dominated by Agrostis stolonifera, A. to a very small extent. Totally, 10 taxa were recorded. gigantea, Festuca ovina and Poa compressa. Taxa such as Large parts of the plot had naked soil. - 22.9. 1975. Sedum acre, Sagina nodosa and Euphrasia sp. can be identified fr om the photo. The number of taxa had reached 20. There were still patches of naked soil - 9.9.1979.

Fig. 6. Sample plot no. 7 after one year of regeneration. Fig. 7. Sample plot no. 7 after four years of regeneration. The plot was dominated by Agrostis sp. (probably Graminids have taken over in the plot. Festuca ovina has gigantea), Festuca ovina and Poa compressa. Sisymbrium reached its highest cover degree, accompanied by Agrostis supinum and numerous small annuals were recorded as gigantea and A. stolonifera. The two colonized patches of well as some other taxa (totally 20 species). Note two Prune/la vulgaris have increased from 1976. Totally 18 colonized patches of Prunel/a vulgaris in the left part of the taxa were recorded. - 9.9. 1979. plot. - 20.9. 1976.

A eta Phytogeogr. Suec. 68 Regeneration in a mechanically disturbed limestone heath 127

Euphrasia has also been treated as Euphrasia sp. A appeared. The dominance and retreat of Sisymbrium few other taxa are also mentioned as sp. In sample in plot 6 can be seen in Figs. 4 and 5. In not disturbed plot 7 it was not possible to separate Agrostis plots Potentilla sp., Trifolium repens, Galium verum, gig antea and A. stolonifera until 1979. Before that Artemisia rupestris and Antennaria dioica seem to they were recorded as Agrostis sp. occur more regularly than in the other plots, where Changes of vegetation in some plots of the they are rare. transects were photographically documented (Figs. Prune/la vulgaris and Filipendula vulgaris might 4-7). In the firsttwo years a few photos documented also be mentioned. Prune/la shows an increase in plot the actual situation in some plots, but later all plots 7 (Figs. 6 and 7) and Filipendula vulgaris has resisted were photographed on the recording occasions. the disturbance very well and might even have been Nomenclature of species names is according to favoured by low competition in plot 8. Sesleria Lid (1974); fo r subspecies and varieties Sterner coerulea ssp. uliginosa, Carex flacca and Veronica (1948) was used. Mosses and lichens were not spicata are only fo und in plot no. 4, under slightly recorded. From 1976, however, the total cover of moister conditions than the others. mosses was determined each year. The moss cover increased in disturbed plots and cover got clearly higher there than in the others. This colonization, mainly of small annual mosses, will of Results course sharpen the competition fo r other taxa trying All material from the analysis in the permanent to colonize or increase their cover in these plots. sample plots has been presented in Table IV. A first The material also shows that there is a permanent attempt at evaluation shows that the number of dynamic situation in the Alvar v getation, with taxa � recorded taxa is slightly smaller in transect I than in occurring irregularly. In these cases, correlations to transects 11 and Ill. high amount of precipitation or dry periods ought to In 1975 there were few taxa in all plots in com­ be expected. For other species there is a more or less parison to 1976, not disturbed (ND in Table IV) plots clear trend of increase (or decrease) as for many of included. This must be regarded as a combined effect the graminids. This indicates that a succession is still of the disturbance and the dry conditions in 1975. proceeding, with competition between taxa, leading Then an increase of taxa took place during the eventually towards more stabilized conditions. Year­ fo llowing years, except fo r 1978 when there was a ly or seasonal fluctuations also occur here, making very dry period in the beginning of the growing the trends a little less distinct. period (Table Ill) which is the most probable reason In records (see below) from the disturbed area for the decreased number of taxa in that year. The south of transects 11 and Ill it can also be seen that decrease was most pronounced in the plots of succession is in progress with competition between transect I, located on rather shallow soil. plants, annuals as well as perennials, more or less The number of graminids increased more in dis­ correlated to climatic conditions. In 1978 the situa­ turbed plots than in not disturbed ones, as they had a tion became more stabilized as a consequence of the high competitive ability there in comparison with perennial grasses reaching dominance in the plant more stabilized conditions in the established plant cover. cover in reference plots. Frost-induced soil movements which were fre­ There was an evident colonization and increase in quent in the area during the first years have been cover values from 1976 onward in disturbed plots. fo und to be less pronounced as the cover of plants in­ The species most responsible for this is Festuca creases. ovina, but also Agrostis gig ant ea, A. stolonifera and Records from the disturbed area 25-7 5 m south of Festuca rubra ssp. oelandica are important. This is transect 11 and Ill during 1975-78: especially distinct in plots 2, 3, 7, 8 and 9. The values 5. A frequent lush colonization on the restored tracks are higher and more stable in undisturbed plots as 197 could be seen, in spite of the dry conditions this year. Fre­ they represent a less dynamic situation regarding new quent taxa were Plantago lanceolata, Convolvolus arven­ colonization. sis, Galium verum, Sisymbrium supinum, Filipendula Among herbids (as in Table IV, including Thymus vulgaris, Veronica spicata, Prunella vulgaris, Bromus serphyllum), it can be seen that, e.g., Sagina nodosa, hordeaceus and Satureja acinos. Frequent but not regular­ Arenaria serpyllifolia, Cerastium spp., Sisymbrium ly appearing taxa were Cirsium acaule and Fragaria supinum and Sedum album occur more regularly in viridis. disturbed plots, and sometimes also show increasing 1976. Arenaria serpy llifolia and Cerastium spp. were cover values. Sisymbrium is the most interesting in very frequent locally. Frequent taxa were Linum cathar­ this respect as it invaded quickly and then nearly dis- ticum, Sedum album, Sisymbrium supinum, Agrostis

A eta Phytogeogr. Suec. 68 128 Ej vind Rosen

Table IV. Recordings of plant regeneration in ten permanent sample plots during 1975-79.

Transect II Ill

Sample plot no . 1 2 3 4 6 7 5 8 9 10 D.is turbance 0 _0 ___ LO NO 0 0 NO 0 0 NO Year abcde abcde abcde abcde abcde abcde abcde abcde abcde abcde

Grami ni ds :

Phleum bertoloni i 1 1 11 1 2 1 1 1 1 1 1 1 1 1 1 1 1 A3rostis sp. 1 2 2 2 . A. stoloni fera l 1 1 1 1 2 1 1 1 1 2 1 1 111 1 1 2 2 1 1 1 1 A. gigante3 . 1 1 1 1 1 1 :t :J 1 1 2 2 3 1 :l 1 1 11 111 1 1 �,es leria coerulea ssp.ulig .. 2 2 3 2 Briza med ia 1 1 1 1 Poa alpi na P. con:prE:ssa X 1 1 2 1 111 11 1 1 1 1 2 1 X 1 X 1 1 1 1 1 1 1 1 Festuca rubra ssp. oel . 1 1 1 2 2 1 1 2 F. ovi na 1 1 1 1 1 2 22344 22332 1 2 2 2 11233 33344 1233 11123 5 4 4 4 4 Bromus hordeaceus 11 1 1 1 2 2 1 1 1 1 1 1 l 1 1 1 2 1 1 X • 1 1 1 Ca rex f1 ace a 1 111 1 Number of grami nids: 24546 23343 33333 46455 15566 34445 36655 25677 23346 33444

Herbids:

Al l ium schoen . var. alv. 11 I 1 1 X 1 1 1 Rumex acetose 11 a 1 1 1 1 Herni aria glabra 1 1 1 1 1 X Sag i na nodosa 1 1 1 1 1 1 1 X 1 1 1 1 1 1 1 1 Arenaria serpyl lifolia J X 1 1 1 1 1 1 1 1 1 1 X 1 1 1 1 1 1 Cerastium spp. 1 1 1 1 1 X X 1 1 1 1 X X 1 1 X Ranunculus bulbosus 1 1 1 X 1 1 1 1 1 1 Erophila verna 1 1 1 1 1 1 1 Si symbri um supi num 1 1 11 3 1 1 2 1 X 1 1 1 Arabi s hirsuta 1 1 1 1 1 1 Sedum a 1 bum 1 1 1 1 1 1 1 2 1 1 1 2 2 1 1 I 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 S. acre 1 1 1 1 11 1 1 l 1 l 1 1 1 1 1 1 1 1 Fragaria viridis 1 1 1 1 Potenti lla sp. x 1 1 1 1 1 1 1 1 1 1 1 X J 1 1 1 1 2 1 1 1 1 Fi 1 i pendu l a vulgari s 32233 11111 11l 2 Medicago lupul ina 1 l 1 X 1 1 1 X 1 1 1 1 1 X 1 1 1 1 1 1 Tri fo 1 i um campes tre 1 l 1 1 T. re pens X 1 1 1 1 l 1 1 1 Anthyl lis vul neraria 1 X 1 X 1 Lotus corniculatus 1 1 1 2 1 1 • L i nu m cathart i cum 1 1 X . 1 l l 1 1 1 X 1 1 1 Convo 1 vo 1 us arvens is 1 1 1 X 1 1 1 1 1 X X 1 1 1 1 1 1 1 1 1 1 1 1 1 Prune 11a vu1 gari s 1 1 1 1 1 1 1 1 1 2 1 X Satureja acinos 1 1 1 1 11 1 1 K 1 1 1 1 . X 1 1 1 X 1 1 1 1 1 1 1 1 1 1 1 1 Thymus serphyl l um 1 1 1 2 1 1 2 1 1 1 2 1 1 1 1 1 1 1 111 Veronica spicata 1 1111 Euphrasia sp . 1 1 Plantago lanceolata 1 1 1 1 1 X 1 X 1 1 1 X Sherardia arvensis Ga 1 i um tri and rum 1 1 1 G. boreale 1 1 G . verum 1 1 1 11 1 1 11 1 X 1 1 1 Antennari a di oi ea 1 1 1 1 1 11 Achillea mi l lefol ium 1 1 1 Artemi sia rupestris 1 1 2 2 X 1 l 1 1 1 1 1 X 1 1 Ci rsium sp. (juv.) C. acaule X X 1 Leontodon autumna 1 is 1 1 1 1 Crepi s tee to rum ssp. pum . 1 . 1 X X 1 1 X X • • Taraxacum spp . 1 1 1 . 1 1 1 1 X 1 1.

No . herbi ds: 8 14 13 5 l2 6 10 74 10 8 14 Kl 6 16 l2 16 20 16 20 9 15 12 Kl 14 14 16 15 11 l3Kl161412 11 7 ll ll 14 l3 :0 14129l4 10 16 20 19 16

Tota l no . taxa : lO 18 18 ll 18 8 13 lO 8 l) 11 17 13 9 l3 16 22 24 19 25 lO 20 17 16 20 17 a:J 191518 l3 22 20 17 16 9 16 17 21 :D 12 17 15 ll 20 l319lA23

Moss cover: 3344 .3445 .5555 .2111 .1123 .1133 .1112 1 1 3 4 . 1 1 1 4 . 1 1 1 2

a = 22.9.1975 0 = disturbed x stands for one single small individual with b = 20.9.1976 LO = little disturbed very smal l cover, as juvenils or some annuals c 17.9.1977 NO = not disturbed d = 16.9.1978 e = 9. 9.1979 Fol lowing occasional recordings compl ete Table IV (plot no ., year, cover degree ). Stel laria sp . p. 7.-75 :x Hornungia petraea p. 1.-77:1; p. 6.- 75:1, -79:1 . Saxi fraga tridactyl ites p. 6.- 76 :1 Prunus spi nosa juv. p. 4.-76:1 Potenti lla fruti cosa juv. p. 4.-77:x Geranium pusillum p. 6.-7 7:1; p. 9.-79:x Ga lium pumi lum p. 6.-79: 1; p. 7 . - 75:x Succi sa pratensis p. 4.- 75:1 Hieraci um sp. p. 4.-75:1, -76: 1, -77:1, -7!3:1, -79:1

A eta Phytogeogr. Suec. 68 Regeneration in a mechanically disturbed limestone heath 129 gigantea, Prunella vulgaris and Bromus hordeaceus. On Irregular occurrence of taxa (vascular plants) in the whole, the plant cover was more sparse and not so lush the plots has been noticed and is probably correlated as in 1975, which might be a result of the dry conditions in to variations in precipitation from year to year. 1975. The recovery of vascular plants has been fairly 1977. The most frequent taxa were Plantago lanceolata, rapid. It has included a smaller number of plants Sedum album, Sisymbrium supinum, Prunella vulgaris, fo reign to the original plant communities than ex­ Herniaria glabra, Sagina nodosa and Festuca ovina. Very pected. But, still many years are needed before the few specimens of Arenaria serpyllifolia and Cerastium recovery will be total, i.e., in a stabilized dynamic spp. were fo und. situation similar to that previous to the disturbance in 1978. Along the restored tracks the vegetation has 1975, especially concerning mosses and lichens. recovered. But still there were many naked soil patches left on the tracks and in other disturbed small areas close by. Some of these patches were to some extent probably kept Acknowledgements open by the trampling of heifers (about 20-30) in the area. For information on how the damage to the Alvar occurred, Graminids had become the most dominant plants in the I wish to express my thanks to Mr. Sigge Samuelsson, vegetation. Especially Festuca ovina seemed to have Albrunna, Dr. Arnold Samuelsson, Kalmar, and to the spread since 1977. Herbids were less dominant than Sydkraft Electric Power Company. before. Some of these were probably suffering from the During the fieldwork I stayed at the Ecological Station drought in the beginning of the summer. Frequent taxa of Uppsala University, Olands Skogsby. For help in were Prune !la vulgaris, H erniaria glabra and Artemisia different ways I thank the Head of the Station and his staff. rupestris. A little less frequent were Sagina nodosa and For photographic laboratory work, I am grateful to Mr. Euphrasia sp., both richly flowering. Folke Hellstrom, and for discussions of the manuscript I thank Dr. Erik Sjogren as well as all the staff of the Obviously regeneration of the plant cover in the dis­ Institute of Ecological Botany, Uppsala. turbed area is going to take a long time. A fairly far­ reaching recovery has already taken place for the graminids and to some extent for the her bids. A much longer time is needed fo r mosses and lichens. REFERENCES Recolonization of lichens was only observed in Albertson, N. 1950. Das grosse siidliche Alvar der Insel scattered spots in the transects. However, the Oland. Eine pflanzensoziologische Ubersicht. - regeneration of the plant cover has included a smaller Svensk bot. Tidskr. 44:269-331. number than expected of plants fo reign to the original Angstrom, A. 1953. Temperatur, Humiditet. Atlas over plant communities. The reason for this may be that Sverige. 25-26. Stockholm. the restoration was done very carefully by the com­ - 1974. Sveriges klimat. 3rd ed. - Stockholm. 188 pp. pany, without any extra soil supplied. Therefore the Bergsten, K.E. 1955. Olands klimat. - Sterner, R. and Curry-Lindahl, K. Natur pa Oland. Stockholm. pp. dominance of taxa normally less fr equent did not last 36-4 1. as long as has been seen in other areas on the Alvar. Du Rietz, G.E. 1921. Zur methodologischen Grundlage der modernen Pflanzensoziologie. - Wien. 272 pp. Konigsson, L.-K. 1968. The Holocene history ofthe Great Alvar of Oland. - Acta phytogeogr. suec. 55:1-172.

SUMMARY Lid, J. 197• . . Norsk og Svensk Flora. 4. utg. - Oslo. 808 pp. The plant cover of a limestone heath (alvar) was Lundqvist, G. 1955. Olands jord och grund. - Sterner, R. seriously damaged by heavy vehicles during work on and Curry-Lindahl, K. Natur pa Oland. Stockholm. an electric power line in January 1975. The area was pp. 16-28. restored in April the same year. Permanent sample Martinsson, A. 1958. The submarine morphology of the plots were distributed and the regeneration of the Baltic Cambro-Silurian area. - Bull. geol. Instn. Up­ plant cover has been studied during 1975-79. sala, 38:11-35. A recovery has taken place. Herbids were most Ostman, C.J. 1926. Om stormar vid Svealands och frequent during the first years. Some herbids in­ Gotalands kuster. (Les grandes vents pn!s des cotes du Svealand et du Gotaland.) - Meddn St. Meteor.-Hydr. creased quickly and then decreased again, such as Anst. 3:1-37. Sisymbrium supinum, while others increased all Rodenborg, L. 1965. Flora och vegetation i och vid through the period. Graminids gradually took over Albrunna lund pa Sodra Oland. 1. Det speciellt un­ and taxa such as Festuca ovina, Agrostis gigantea dersokta omradet. - Svensk bot. Tidskr. 59:411 - 462. and A. stolonifera gradually became the most domi­ 1966. Flora och vegetation i och vid Albrunna lund pa nant ones again, as in the reference plots. The moss Sodra Oland. 2. Albrunna lund, utanfor det speciellt cover has increased in disturbed plots. undersokta omradet, samt lundens omgivningar. -

A eta Phytogeogr. Suec. 68 130 Ej vind Rosen

Svensk bot. Tidskr. 60:255-292. Zoon, Suppl. 1:1 37-151. 1967. Flora och vegetation i och vid Albrunna lund pa Sjogren, E. 1961. Epiphytische Moosvegetation in Sodra Oland. 3. Betesgang och vegetationsutveckling, Laubwaldern der Insel Oland (Schweden). - Acta m.m. - Svensk bot. Tidskr. 61:353-384. phytogeogr. suec. 44 :1-149. Rosen, E. 1980. Application of permanent sample plots fo r Sterner, R. 1948. Olands flora. - B. Palm, L. Landin and conservation of vegetation on the Alvar heath of 0. Nordmark, Oland I. Lund. pp. 91-235. Oland. - Phytocoenosis (1978) 7:317-33?.. Widen, B. 1980. Flowering strategies.. in the Helianthemum Rosen, E. & Sjogren, E. 1973. Sheep grazing and changes oelandicum (Cistaceae) complex oi!'Dland,Swede n. - of vegetation on the limestone heath of Oland. - Bot. Notiser 133:99-115.

Ej vind Rosen, Institute of Ecological Botany, Uppsala University, Box 559, S-75 1 22 Uppsala, Sweden

A eta Phytogeogr. Suec. 68 Observations on Two Bushfires in the Western Kalahari, Botswana Christina Skarpe

Fire is a natural factor in and a condition for many all uncontrolled burning. Usually, in the Kalahari en­ savanna ecosystems. It is caused naturally by, fo r ex­ vironment, it is impossible to distinguish between the ample, lightning. For thousands of years in southern effects on the vegetation of the fire itself and the Africa man has also induced fires by accident or detrimental effects of heavy grazing of the regrowth voluntarily fo r hunting purposes or to provide fresh immediately after a fire. grazing fo r livestock. Lately, burning has been used A small grazing ecology project is working in the in range management in order to control encroaching western Kalahari under the supervision of the shrubs or to keep the field layer in a desired condi­ Ministry of Agriculture, Botswana. The two study tion. areas were burnt by accident in the dry season of The most obvious result of burning is the increase 1977 (Fig. 1). As background data on the vegetation in fire-tolerantspecies and life forms and the decrease were available, some effects of the fireswere followed in those sensitive to fire. The reaction by plants and up in order to get an idea of the reaction to fire of soils to fire, however, varies depending on the phe­ vegetation that was not stressed by grazing. nological stage of the plant and on different con­ Serious limitations to the applicability and com­ ditions of the fire, such as amount and character of parability of the results are that the two areas burnt at fuel available and weather conditions such as speed different times, in July and September respectively, and direction of wind and relative humidity (Dauben­ and with different intensity, and that no unburnt mire 1968). Grasses may be damaged or stimulated areas were available as reference. by fire, depending on species and on conditions for the burn. Bush encroachment and thicket fo rmation may be favoured by fire (Story 1952; Waiter & Yolk Study area 1954) but as a rule frequent firesprevent the develop­ The Kalahari is an almost flat inland plateau about ment of woody species (Reynolds & Bohning 1956; 1 000 m above sea level. It forms part of the vast Guilloteau 1957; Komarek 1965). It has frequently geological Kalahari formation, whose uppermost been fo und that savanna areas protected for long member is aeolian sand of presumed Pleistocene age. periods fr om fire are eventually invaded by trees and The surface is almost flat, slightly undulating from shrubs (Guilloteau op. cit.). stabilized vegetated dunes and transected by fo ssil Fires also have indirect effects on vegetation by river beds, remnants from periods with dryer and changing microclimate and soil properties. The wetter climate than the present (Leistner 1967). denudation and blackening of the soil surface in­ The area under consideration is arid to semi-arid creases soil temperature and windrun close to the (Meigs 1960) with increasing aridity towards the ground, thus increasing evaporation. As a rule, soil southwest. Rainfall is erratic, and the variation moisture is reduced after a fire (Phillips 1919; Cook roughly increases with decreasing means. Rains fall 1939). Nitrogen and sulphur are largely lost in almost entirely during summer, from October to gaseo·us form during the burn, and phosphorus, April. The two study areas have an annual precipita­ potassium, calcium and magnesium, etc. may be lost tion (long-term means) of about 300 mm (Dondong) by ash blowing away. and 350 mm (Tshobokwane) (UNDP/FAO, map). Little is known about fireimpa ct on soil organisms The vegetation is largely savanna with perennial such as algae and fungi, and their direct and indirect tufted grasses, a varying element of annuals, shrubs significance for the higher vegetation ( Cohen 1950; and usually scattered low trees. The present study Meiklejohn 1955). was carried out at Dondong, at 23°07'S, 20°32'E, A great part of the savanna area in Botswana is and at Tshobokwane, at 22°04'S, 21°12'E (Fig. 2). burnt every year. At many frequently burnt places Both areas are covered by fairly deep red Kalahari the vegetation shows signs of deterioration. The sand. At Dondong there are small patches of white Botswana Government has, with little effect, banned sand, which were omitted in this study. The vegeta-

A eta Phytogeogr. Suec. 68 132 Christina Skarpe

Fig. 1. Bush fire in the vicinity of Tshobokwane. In the background, fa intly visible, three columns of ashes being carried off the burnt area by whirlwinds.

tion at both places is Short Grass Savanna with ZAMBIA (mainly) Deciduous Shrubs, as defined by Mueller­ ',, ANGOLA _-:_'- .... ------· - - - ... , Dombois & Ellenberg (1974). -----.... _,- -'

Dondong ZIMBABWE

Moun At Dondong the vegetation was classified by Blair NAMIBIA 0 Rains & Yalala (1972) as Central Kalahari Tran­ sitional Savanna. There are scattered trees, mainly Acacia erioloba (Table A. luederizii and Boscia Ill), OGhon zo albitrunca. Shrubs include, apart from young * hhobokwone BOTSWANA specimens of the tree species, tall-growing Acacia mellifera with a lower storey of Grewiaflava, Lycium * Dondong namaquense and Rhus tenuinervis. The field layer is dominated by two perennial grasses, S chmidtia pap­ 0 Tshone pophoroides and Stipagrostis uniplumis, and further includes, e.g., Brachiaria nigropedata, Anthephora pubescens, Eragrostis lehmanniana and the annuals Rhynchelytrum repens and Urochloa brachyura. Before the fire, herbs included many chamaephytes 100 200 km with one species of Monechma and Plinthus sericeus as dominants, and rather few therophytes. Fig. 2. The position of the study areas in Botswana.

Acta Phytogeogr. Suec. 68 Two bushfires in the Western Kalahari 133

Table I. Ra infall in nvn on the two meteorological Methods stations closest to the study areas. Both study areas include an ungrazed reference area 30-year of 50 ha, that was used for the present study. seasonal Station mean 1 976/77 1 977/78 The field layer was sampled in quadrats, 1.0 x 0.5 m, placed in clusters of four, and laid out in a Ghanzi 401 .4 388.6 689 .4 stratified random system. In all, 64 quadrats were Tshane 355.7 335.9 383 .2 sampled every or every second month at Dondong and with somewhat longer intervals at Tshobokwane. No spot was sampled twice. Grass species were The vegetation shows a mosaic pattern with fairly sampled individually, but all other species were taken open areas dominated by Schmidtia pappophoroides as a group. The samples were clipped at ground level and with scattered low shrubs varying with patches and stored air dry until taken to the laboratory, where of large Acacia mellife ra shrubs with a sparse field they were oven dried fo r eight hours at 80°C and layer of, fo r instance, Eragrostis lehmanniana, weighed. Stipagrostis uniplumis, annual grasses and hemi­ From the repeated values of standing crop thus ob­ cryptophytic herbs (or often intermediate between tained, a rough estimation of net production was hemicryptophytes and geophytes, as defined by made by simply adding the positive differences dur­ Leistner & Werger 1973). ing the year (1st July-30st June) fOreach species. Shrubs were sampled once a year at the end of the growing season, using the point centred quarter Tshobokwane method (Cottam & Curtis 1956). Five lines, laid out The Tshobokwane area was included in the northern at random each year, and running perpendicular to Tree and Shrub Savanna by Blair Rains & Yalala the slight gradient, each with ten sampling points, (op. cit.). Here it is without trees and taller shrubs, were used, giving 200 sampled shrubs in all. Shrubs and consists of open grassland with low shrubs, were definedas woody species between 0.2 and 3.0 m mainly Acacia fleckii, Bauhinia petersiana, high, typical dwarf shrubs excluded. Acacia mellife ra Dichrostachys cinerea, Lonchocarpus nelsii and Ter­ was considered a shrub even when higher than 3 m. A minalia sericea. The field layer is dominated by shrub individual, when not single stemmed, was Anthephora pubescens, Aristida meridionalis, defined in the common meaning of stems radiating Braclziaria nigropedata, Eragrostis spp., Schmidtia from about the same spot. This rough definition pappophoroides and Stipagrostis uniplumis. Herbs caused little problem in the present study. It is, are largely geophytes such as Ty losema esculentum. however, appreciated that these "individuals" in The vegetation is fairly homogeneous, with only a many cases, e.g., in Bauhinia petersiana, Grewia slight difference between the very low dune crests flava and Rhus tenuinervis, are interconnected and with their denser shrub cover, and the flat dune are not true individuals. The shrubs were sampled by valleys. species and in three height classes: 0.2-1 .0, 1.0-2.0

Table I!. Estimates of net producti on in kg dry matter per ha and i n % for major grass species at Dondong and Tshobokwane during the seasons 1 976/77 and 1 9 77/78.

Dondong Tshobokwane 1 976/77 1 977/78 1976/77 1 977/78 seeci es kg/ha % kg/ha % kg/ha % kg/ha %

An thephora pubescens 282 25 548 24 Aristida me ridionalis 153 14 90 Brachiaria nigropedata 0 0 364 16 Rhynchelytrum repens 161 Schmi dti a pappophoroi des 302 27 464 25 124 11 233 11 Sti pagrostis unipl umis 736 65 669 35 563 50 1052 46 Uroch 1 oa brachyura 81 596 32

sum 1128 1890 1122 2287

Acta Phytogeogr. Suec. 68 134 Christina Skarpe

Table Ill. Plant species mentioned in the text. Nomenclature fol lows mainly Flora Zambesica, Sal i sbury. Life forms : P - Phanerophytes , Ch - Chamaephytes , H - Hemicryptophytes , C - Geo­ phytes , GH intermediate between Geophytes and Hemi cryptophytes , T - Therophytes . Occurrance in the study areas at Dondong (D) and Tshobokwane (T).

life occur- species auctor form ranee co11111ents

Acacia erioloba E. Mey D = A. giraffae Willd.

Acacia flecki i Schinz p DT infrequent at D Acacia me11 if e ra (Vahl ) Benth. ssp. detinens (Burch. ) Brenan DT

Acacia 1 vederi zi j Engl . DT Anthephora pubescens Nees H DT Aristida meri dionalis (Stapf) Henrard H DT infrequent at D

Bauhinia petersiana Bol le T � B. macrantha Baker Boscia albitrunca (Burch.) Gi lg & Bened . DT infrequent at T Brachiaria nigropedata (Munro ex Hi ern) Stapf H DT Citrul lus lanatus (Thunb. ) Matsum. & Nakai T DT

C1eome rube 11a Burch . T DT Crotalaria spbaerocarpa Perr. ex DC . T DT

Dichrostachys cinerea (L. ) Wight & Arn . p DT Eragrostis lehmanniana Nees H DT x Gisekia africana (Lour. ) Kuntze T o x Gi sekia pharnaceoides L. T o

Grewi a flava DC. p DT lndigofera charl ierana Schims . T D Ipomoea magnusiana Schinz T DT Limeum fenestratum (Fenzl ) Heimerl T DT Limeum sulcatum (Klotzsch) Hutch . T DT

Lonchoca rpus ne 1 si i (Schinz) Schinz ex DT Heeri ng & Gri11111e

Lyci urn namaquense Da11111er p D Plinth us seri ceus Pax Cb D Pol lichia campestris A it, GH DT Raphionacme burkei N.E. Br. G D

Rhus tenui nervi s Engl . p D Rhynchelytrum repens (Wi lld,) C.E. Hubbard T D Schmidtia pappophoroides Steud. H DT Sesamum alatum Thonn. T D Sti pagrostis uniplumi s (Licht. ) De Winter H DT

Terminalia sericea Burch , ex DC. p T Tribulus terrestris L. T D Tribulus zeyheri Sond. T Tylosema esculentum (Burch.) Sbreiber G T Bauhinia esculentba Burcb.

Uroch 1 oa brachyura (Hack.) Stapf T D x One or both Gi sek i as also at Tshobokwane

Acta Phytogeogr. Suec. 68 Two bushfires in the Western Kalahari 135

and 2.0-3.0 m. Two maximal canopy diameters, N-S and E-W were measured, and the crown cover was calculated using the mean diameter and assum­ ing the crown to be circular (The foliage cover is con­ siderably less due to spaces between leaves.) 160 trees were sampled with the same method "once and for all" in April 1977. To avoid double sampling, the sampled trees were marked. Thus, with few exceptions, they could be refound after the fire and their survival noted.

Results of observations on the bushftres

Dondong burnt in September 1977. As the fire was not directly observed, efforts were made afterwards to gather information by questioning people living about five kilometers from the research area and by studying signs in the terrain. The fire seems to have spread against a moderate wind. It was stated that smoke and fire had been seen in the area for two or three days. As it had not been grazed except for a very slight grazing by game, there was a fairly large amount of dry grass, about 2 400 kg/ha, available as fuel. There was also a lot of dry wood, as the area had not burnt for more than four years. September is in the end of the dry season. The time Fig. 3. Acacia erioloba shrub, ea 3 m high, at Dondong fo r the burn as well as the fuel available and the fact with part of the old canopy sprouting in February 1978, that the fire spread against the wind (Komarek op. five months after the burn. cit.; Daubenmire op. cit.), indicates a hot and intense burn. Its long duration must mean that firela y glow­ ing in woody material and, possibly, spread secon­ seen in the area before. Among the most important darily to unburnt grass. There were fe wer unburnt ones were: patches than is often the case after Kalahari grass Citrullus lanatus Ipomoea magnusiana fires. Cleome rubella Limeum fe nestratum In October hardly any regrowth of grasses and no Crotalaria sphaerocarpa Limeum sulcatum seedlings were noted. A few geophytes, such as Gisekia africana Sesamum alatum Raphionacme burkei, were in flower. Many shrubs, Gisekia pharnaceoides Tribulus terrestris especially Boscia albitrunca and Grewia flava, were I ndigofe ra charlierana sprouting from basal parts. Early in December the rains had started, and the Sesamum alatum had not been seen in the area area was an intensive green from sprouting grass, before. Of the perennial grasses Schmidtia pap­ 10-15 cm high. Seedlings of annuals were develop­ pophoroides and Stipagrostis uniplumis obviously ing rapidly, and most shrubs were sprouting had suffered from the fire. Most tufts weredama ged vigorously. By the end of the month Schmidtia pap­ to various degrees and many, especially Stipagrostis, pophoroides was by far the most prominent grass and had been killed. The production by S tipagrostis in the although still low, was in flower. season following the fire was lower than in previous Sampling was started in February 1978 when the seasons. Although even Schmidtia suffered from the vegetation was developed well enough for most burn, its production increased and the flowering was species to be easily identified. The annual grasses exceptionally rich. Rhynchelytrum repens and Urochloa brachyura were Anthephora pubescens and Brachiaria nigro­ then among the dominants. Also annual herbs had in­ pedata are too sparse at Dondong to allow any es­ creased very markedly and included some species not timate of production with the method used. The two

Acta Phytogeogr. Suec. 68 136 Christina Skarpe

CROW N NUMBER COVER PER HA 6 ( ofo ) 300 f'Z 1977 1978 1977 1978 <3m 5 >3m rn 3m rn 2- 4 2-3m 1-2m 200 1-2m 3 0.2-1m 0.2-1m �

r- 100 �(L � � �� �

Fig. 4. Crown cover in % before and after the fire fo r ma­ � jor shrub species at Dondong. >3 m only Acacia mellife ra. (a) Acacia erioloba, (b) A. mellifera, (c) Boscia albitrunca, 0 �,a b d e (d) Grewia flava, (e) Lycium namaquense, (f) Rhus tenui­ nervis. Fig. 5. Number of individuals per ha before and after the fire for major shrub species at Dondong. (a) Acacia erioloba, (b) A. mellifera, (c) Boscia albitrunca, (c) Grewia flava, (e) Lycium namaquense, (f) Rhus tenuinervis.

species were, however, in excellent condition and did rechecked five months after the fire, 39 were dead, 6 not seem to have suffered at all from the fire. were regarded as dying and 5 could not be retraced. Of the perennial her:-bs, the hemicryptophytes sur­ Of those missing, two or three were almost certainly vived the fire to some extent. Especially Pollichia identifiedas dead and fallen, the marking being either campestris proved fire tolerant, and in some cases destroyed or on the under side of the trunk. If the dy­ even survived where large shrubs had been burnt out. ing and missing trees are counted as dead, 31 % of The previously important chamaephytes, Monechma the tree layer was killed directly by the fire. Later sp. and Plintus sericeus, decreased very markedly observations have shown that an additional 5 to 7 after the fire, the latter being all but eradicated from trees have died or are in a very poor condition, the area. probably from disease entering through wounds and Most shrubs, particularly those less than two cracks in the bark caused by the fire. The majority of meters high, were killed above ground, and re­ the trees killed were small, but some big trees suc­ sprouted from basal parts. On many of the taller cumbed and were even fe lled owing to holes and shrubs, a portion of the canopy survived, but es­ cavities burnt out in the trunks, probably in old scars pecially Acacia erioloba and A. mellifera often with dry wood (Fig. 6). Of the trees killed, only a few, sprouted from the base even with part of the old mainly Acacia erioloba sprouted from the base. crown alive (Fig. 3). A number of large Acacia mel/ife ra shrubs were burnt down to the ground and Ts hobokwane burnt in July 1977, in the middle of the completely killed. The total crown cover of shrubs dry season. The fire was said to have spread rapidly decreased from 15.4 % in 1977 to 10.1 % in 1978 with the wind. There was about 3 300 kg grass per ha (and the foliage cover was still more reduced due to available as fuel, but very little dry wood, as the area thinning of the canopies) (Fig. 4). The number of only supports small shrubs and was burnt two years shrubs per ha increased from 625 to 778 (Fig. 5). before. The fire must have been considerably less in­ Of a total of 160 marked trees which were tense than that at Dondong.

A eta Phytogeogr. Suec. 68 Two bushfires in the Western Kalahari 137

CROWN COVER 5 (0/o)

4

3 2

Fig. 7. Crown cover in % before and after the fire for ma­ jor shrub species at Tshobokwane. (a) Acacia fleckii, (b) Bauhinia petersiana, (c) Dichrostachys cineria, (d) Grewia flava, (e) Te rmina/ia cericea.

NUMBER PER HA

Fig. 6. Acacia luederizii tree in the vicinity of Dondong, killed and felled by fire. 2000 1977 1978 �1-2m In October most grasses were sprouting, and some rn geophytes were fully developed, in flower or even in 1500 o0.2-1m fruit. The shrubs were sprouting vigorously from the base. Sampling began in November, when most grasses could be identified. There was no increase in annuals as compared to previous years. Of the perennial grasses Schmidtia 1000 pappophoroides, Stipagrostis uniplumis and to some extent A ristida meriodionalis had damaged tufts, but only the latter showed a decrease in production. Anthephora pubescens and Brachiaria nigropedata were in very good condition and both increased their production as compared to the previous season. 500 Most shrubs, including all Dichrostachys cineria and Grewia flava had their canopies killed, and resprouted from basal parts. A number of Bauhinia petersiana and some Lonchocarpus nelsii and Ter­ minalia sericea retained part of their old canopies and often sprouted from the base as well. The general trend was the same as at Dondong with a decrease in Fig. 8. Number of individuals per ha before and after the crown cover from 15.7 % in 1977 to 11.5%in 1978, fire fo r major shrub species at Tshobokwane. (a) Acacia and an increase in number of individuals per ha from fleckii, (b) Bauhinia petersiana, (c) Dichrostachys cineria, 2 04 1 to 3 636 (Figs. 7 and 8). (d) Grewia fla va, (e) Te rminalia cericea.

A eta Phytogeogr. Suec. 68 138 Christina Skarpe

Discussion vious. They must have been present as seed before the fire, survived it and then have been fa voured by In both areas the total production of grass was higher the temporary set-back in competitive fo rce by the in 1977/78 than in 1976/77. It is not possible to perennials (Fig. 9). The patches that presumably determine the extent to which this was caused by burnt most severely, where large shrubs were burnt higher precipitation in the season 1977/78 or by out completely, remained bare for the first season stimulation by the fire. The change in production in afterthe fire,but in the following year were colonized kg dry matter per ha per year may say less about the by annuals and scattered Eragrostis lehmanniana. fire tolerance of different species than their share in Probably temperatures there had been too high fo r the total production, although that too may be too long and had killed the dormant seed. At affected by the good rains. Tshobokwane, where little damage to perennial The increase in Schmidtia pappophoroides at the grasses was recorded, no increase in ephemerals took expense of Stipagrostis uniplumis at Dondong was place. most likely fire-induced. The initial decrease in Most shrubs survived the firein the sense that they production of Stipagrostis probably resulted more sprouted anew from basal parts. In Bauhinia peter­ from the reduction in number of tufts than in a reduc­ siana, Te rminalia sericea and Loncocarpus nelsii the tion in vigour of the survivors. In the later develop­ canopies survived to some extent. The two former ment, competition from the Schmidtia must have species were the only ones to increase both in number played a major role. Schmidtia pappophoroides is and crown cover after the fire. Too few specimens of said by Yolk (1974) to be favoured by fire, and the Lonchocarpus nelsii were sampled to allow any con­ local cattle-keepers claim the same. If only S chmidtia clusion regarding that species. The general trend at pappophoroides and Stipagrostis uniplumis are taken both places was a decrease in total crown cover and into consideration, their contributions were 29 and an increase in number of individuals. An increase in 71 % respectively for the season preceding the fire, number was caused by many young shrubs for the 41 and 59 % fo r the first and 63 and 3 7 %for the se­ firsttime reaching sampling height (0.2 m) stimulated cond season after the fire, in spite of the lower by the good rains and by the set-back in water uptake precipitation in 1978/79. In the third season, by the larger shrubs and the grass layer. It might also 1979/80, there seems to be a drop in the production to some extent have been caused by root shoots of Schmidtia. Anthephora pubescens and Brachiaria developing at some distance from the parent shrub, nigropedata were in excellent condition after the fire and thus in the present study regarded as separate in­ both at Dondong and at Tshobokwane. In contrast to dividuals, although this was not investigated. In the the S chmidtia, their tufts were never damaged in the ungrazed area at Dondong, the number of shrubs burn. The Schmidtia probably should be regarded as dropped during the fo llowing seasons, and in 1980 being promoted by the fire, rapidly establishing new was back at the same level as before the fire. This plants from seed and/or rooting stolons, rather than reduction may have been caused by increasing com­ as an individual survivor like the Anthephora and the petition for water between the growing shrubs Brachiaria. The two latter are cushion grasses fo rm­ themselves and/or between the shrubs and the grass ing large dense tufts with the perennating buds well layer. The fact that the number of shrubs increased protected by sheaths on the short rhizomes. Schmid­ more and has remained high under heavy grazing in tia pappophoroides and Stipagrostis uniplumis are an adjacent area, may suggest the latter reason as the fairly loosely tufted, the fo rmer of a shrubby most important. appearance, the latter a bunch grass. The fire Shrub encroachment caused by overgrazing is a tolerance of Anthephora pubescens, Brachiaria serious problem in range management in the nigropedata and Schmidtia pappophoroides have Kalahari, and burning is regarded as the only been indicated also from experimental burning economic way to control it. In this case the fires had without grazing in eastern Botswana (D. Pratchett, no persisting adverse effect on the shrub layer. The pers. comm.). These three species are those with the height and size of shrubs was, however, much best forage value in the region (Animal Production reduced, especially at Dondong where large shrubs Research Unit, 1977), and in a grazed area the had been present. It is mainly the large shrubs and the development is likely to have been different. This was trees that have any marked effect on the composition also observed at Dondong, where under heavy graz­ of the field layer around them. One of the most im­ ing the Stipagrostis increased at the expense of the portant encroaching species is Acacia mellifera, and Schmidtia, i.e., the opposite reaction to that in the un­ the killing of a number of the large specimens, while grazed area. the smaller sprouted again from the base, is At Dondong the increase in annuals was very ob- noteworthy. Instances are known, for example cited

A eta Phytogeogr. Suec. 68 Two bushfires in the Western Kalahari 139

Fig. 9. Tribulus zeyheri invading on area where Stipagrostis uni­ plumis have been damaged by fire and grazing. Tshane.

by Werger (1977), where thickets of encroaching ment was made of stem height below the canopy. shrubs suddenly died after some thirty years, due to Bark thickness is closely correlated to stem diameter. senescence of the shrub itself or to insufficient soil In both Acacia species the bark of a large tree is 20 to moisture. In either case it might be presumed that the 30 mm. In Boscia albitrunca it seldom exceeds 20 shrubs get increasingly sensitive to adverse factors mm. such as fire. Nothing is known about the age of the From the present study it seems probable that shrubs killed at Dondong. frequent fires can easily wipe out the tree population The tree species at Dondong seem poorly adapted at Dondong and might have done so at to fire. Even if this was a severe burn, the killing of Tshobokwane. Otherwise, apart from the reduction one-third of the trees must be regarded as high. Both in chamaephytic herbs at Dondong, the fires had little Acacia species show roughly the same mortality, remaining effect on the vegetation. The plant cover at somewhat more than 30 %. Boscia albitrunca seems Tshobokwane, however, seems better adapted to fire to survive better, although the sample is too small for than that at Dondong, possibly as a result of more any accuracy. This is, however, also indicated by the frequent burns in the past. Certainly the firesdid not presence of Boscia in some frequently burnt areas in cause deterioration of the field layer from a grazing the eastern Kalahari. point-of-view, as the best surviving grasses were also That small trees are more vulnerable than larger those with the best fo rage value. The study reveals, ones seems natural, but the data do not permit the however, nothing about the effects of different fire correlation of survival with any special factor such as frequencies, nor of the combined effect of fire and height, stem diameter or bark thickness. No measure- grazing.

Acta Phytogeogr. Suec. 68 140 Christina Skarpe

Acknowledgements Komarek, E.V. Sr. 1965. Fire ecology-grasslands and man. - Tall Timbers Fire Ecol. Conf. 4:169-220. I thank the Animal Production Research Unit, Ministry of Leistner, O.A. 1967. The plant ecology of the southern Agriculture, Botswana, for making the present study poss­ Kalahari. - Mem. bot. Surv. S. Afr. 38. ible and fo r permission to publish this paper. I am also very Leistner, O.A. & Werger, M.J.A. 1973. Southern Kalahari grateful to Mr. R. Bergstrom, Dr. A. Blair Rains and Dr. phytosociology. - Vegetatio 25 :353-399. M.J.A. Werger for their valuable comments on the Meigs, P. 1960. Revision of maps from world distribution manuscript. of arid and semi-arid homoclimates. - Arid Zone Progr. 1 :203-2 10. Meiklejohn, J. 1955. The effect of bush burning on the microflora of a Kenya upland soil. - J. Soil. Sci. 6:1 1 1-1 18. REFERENCES Mueller-Dombois, D. & Ellenberg, H. 1974. Aims and Animal Production Research Unit 1977. An integrated methods of vegetation ecology. - New York, London, program of beef cattle and range research in Botswana Sydney, Toronto. 1970-1977. - Animal Production Research Unit, Phillips, E.P. 1919. A preliminary report on the veld burn­ Ministry of Agriculture, Botswana. ing experiments at Groenkloof, Pretoria. - S. A fr. J. Blair Rains, A. & Yalala, A.M. 1972. The Central and Sci. 16. Southern State Lands, Botswana. - Land Resource Reynolds, H.G. & Bohning, J.W. 1956. Effects of burning Study 11. on a desert grass-shrub range in southern Arizona. - Cohen, C. 1950. The occurrence of fungi in the soil after Ecology 37:769-777. different burning and grazing treatments of the veld in Story, R. 1952. A botanical survey ofthe Keiskammahoek the Transvaal. - S. Afr. J. Sci. 46:245-246. District. - Mem. bot. Surv. S. Afr. No. 27. Cook, L. 1939. A contribution to our information on UNDP/FAO Mean annual rainfall, Botswana. (Map, un­ grassburning. - S. Afr. J. Sci. 36:270-282. dated). Cottam, G. & Curtis, J.T. 1956. The use of distance Yolk, O.H. 1974. Graser des Farmgebietes von measures in phytosociological sampling. - Ecology Si.idwestafrika. - Wiss. Forsch. S.W. Afrika 13. 37:45 1-460. Waiter, H. & Yolk, O.H. 1954. Grundlagen der Daubenmire, R. 1968. Ecology of fire in grasslands. - Weidewirtschaft in Siidwestafrika. - Stuttgart. Adv. ecol. Res. 5. Werger, M.J.A. 1977. Environmental destruction in Guilloteau, J. 19 57. The problem of bushfiresand burns in southern Africa: the role of overgrazing and tramp­ land development and soil conservation in Africa south ling. - Vegetation Science and Environmental Protec­ of the Sahara. - African Soils 4:64- 102. tion.

Christina Skarpe, Institute of Ecological Botany, Uppsala University, Box 559, S-75 1 22 U ppsala, Sweden

A eta Phytogeogr. Suec. 68 Continued Investigations of Epiphytic Lichen Flora around Kvarntorp in Narke Erik Skye

The present work fo llows up earlier investigations of the placing of the above-mentioned composition in alterations in the epiphytic flora composition as a result of relation to important environmental factors. air pollution in K varntorp, in the Swedish province of The first return visit was made by Ingemar Narke (Skye 1958; Skye & Hallberg 1969). The fieldwork Hallberg in the summers of 1967 and 1968. He ex­ 1979 was carried out during the summer of and in this task tended the list of species to include the crustose I was helped by my wife Monica Skye who also assisted lichens and published his results at the Institute of me in classifying the material. Willy Jungskar assisted me with the pH-analysis. Ecological Botany (Hallberg 1969). Hallberg also re­ As a result of the kind cooperation of Lennart Persson marked the trees with small aluminium plates that of the Nature Protection Unit, Provincial Administration, were nailed to the bark. Orebro, I have received valuable information on the I visited the district in the summer of 1979 and magnitude of the S02 discharge in the Kvarntorp region. registered all fructicose and foliose lichens that were He also presented me with a booklet, Miljon i Kvarntorp, present on the marked trees. The occurrence of Kumla kommun, lansstyrelsen i Orebro lan (The Environ­ Bacidia chlorococca, Lecanora conizaeoides and ment at Kvarntorp, Kumla Municipality, Provincial Ad­ Lecidea scalaris was also noted. Notes were also ministration, Orebro). This booklet is discussed below un­ made of any changes in the environment that could der the reference Eriksson 1979. Most of the information be observed. in the historical review has been taken from this publica­ tion. I am most grateful to all those involved for their The present paper reports the gradual changes that valuable assistance. have occurred since 1952/53 by comparison with earlier analyses (Skye 1958), Hallberg's results (Hallberg 1969) and the present notes. Attempts are The object of the research also made to interpret the reasons behind these The first investigation of the lichen flora's composi­ changes. tion within the K varntorp district was done during the summers of 1952 and 1953. Already at that time I was prepared to return and study the changes that could be expected to occur in the future. The Methods "stations" that were analysed were numbered, the An account of the field work,choice of trees, etc., is trees were marked and the location of the place given in Skye (1958 p. 138 ff.).At the two latest in­ described as accurately as possible. However, ex­ vestigations the same "stations" were visited as those perience has shown that a description of this kind in the first investigation (cf. Hallberg 1969 p. 5) and cannot be made sufficiently clear. Nevertheless, the field work has been done in principle in the same despite the deficiencies a very large number of way on all three occasions. However, Hallberg only stations have been re-located. Naturally, the choice included one tree per station but in addition also in­ of stations can also be discussed, but by means of the cluded the crustose lichens. repeated analysis of the same stations it is possible to In all the investigations, material for comparison eliminate some of the weaknesses in the first in­ was taken from the agricultural landscape to the west vestigation. My experience of this type of investiga­ of Europe Route 18 near Kumla. tion was non-existent when the work started in 1952. Of the 58 rich bark stations that were analysed in The contours of the research task could, thus, be 1952/53, Hallberg visited all of them, in most cases distinguished already from the start and can be fo r­ analysing exactly the same tree. In the 1979 in­ mulated as follows: At regular intervals and at vestigation three stations had disappeared, namely, selected places within the K varntorp district to nos. 12, 17 and 70. Of the remaining stations most of analyse the epiphytic lichen flora's composition, the them are the same tree as studied earlier and, conse­ abundance and vitality of the species concerned, and quently, a large number have been visited on all three

Acta Phytogeogr. Suec. 68 142 Erik Skye occasions. This provides a unique occasion to study agriculture, quarrying was early an important in­ alterations in lichen flora. dustry. Reports from the 17th century mention Hallberg collected bark samples fo r pH -analysis quarrying of sandstone and limestone from from 9 places along an east-west line through the area Hallabrottet, but this work probably started much of investigation (Hallberg 1969 p. 38 fT. ). In the pre­ earlier. The stone was used for building. Limestone sent work bark samples have been taken from the was probably also burnt to a certain extent for same trees, possibly with the exception of the eastern­ manufacture of mortar. most sampling place and in addition from station no. Quarrying in the district increased during the 19th 13. century. In 1879 Yxhults Stenhuggeri AB (Yxhult's The pH -analysis was done in the same way as Stone-works Co.) was founded. Lime-burning was described in Skye ( 1968 p. 1 06). also increasing now and alum shale was used as fuel. The railway from Kumla to Hallabrottet was opened in 1883. In 1929 Yxhults Stenhuggeri AB started to Area of comparison manufacture lightweight concrete from shale ash and The requirement fo r the area of comparison (the con­ lime. The burning was done in field ovens at the trol area in Hallberg 1969) was that it should be as quarry. The product is called Ytong. uninfluenced by industrial effluent as possible but The field ovens probably released considerable 9therwise have the same type of nature and a similar amounts of sulphur pollution. Until 195 1, namely, development as the area of investigation around the field ovens were in the open, i.e. the ovens were K varntorp. In all three investigations the area of not covered. However, this did not occur later (from comparison is on the Narke plain to the west of Eriksson 1979). Kumla. The composition of the epiphytic fructicose and foliose lichens in the area of comparison is in good Influence on the environment agreement with that occurring at the most peripheral The influence on the environment before 1941 was stations in the area of investigations, which is inter­ probably limited. The epiphytic flora was probably preted to indicate that the area of investigation is of fairly similar throughout the entire Narke plain but sufficient size. locally the influence from, e.g., the shale burning, can The alterations that could be observed depend on be demonstrated. The changes that have occurred natural variations in diversity and abundance. These, due to the expansion of modern agriculture, increased in turn, can be referred to a number of factors, fo r ex­ road traffic, etc., are reflectedin the conditions within ample, the age and competitive ability of the different the area of comparison studied to the west of Kumla. individuals, changes in the substrate as a result of We consider that a corresponding development also the phorophyte growing older, changes in ex­ would have taken place in the K varntorp area if the posure to light, road dust, etc. Predation also plays a industry had not developed in the way it has. certain role. In the summer of 1979 about ten trees were observed with a mass invasion of a common snail (Cepaea hortensis). On several stems the density 1941-1966 of individuals was about 1 t per dm 2 from the In January 1941 the Svenska Skifferolje AB (Swedish rootneck and up to a height of at least 3 m. It Shale Oil Co.) was formed to produce oil from alum appeared as if Parmelia acetabulum was the most ap­ shale in order to safeguard the needs of the armed preciated lichen but also, fo r example, Anaptychia fo rces, primarily the navy. Work on the factory was ciliaris. It appears without doubt that an invasion of started at K varntorp in the spring and the firstoil was this kind can influencethe epiphytic lichen vegetation produced one year later. fo r several decades in the future. The production of shale oil was based on the dry distillation of alum shale. The shale contains 7 % sul­ phur, one-third of which is transformed to hydrogen sulphide and is removed with the distillation gases, Area of investigation whereas the remainder is retained in the shale ash. The premises of the Shale Oil Co. at K varntorp Pre-1941 were extended during the 1940's and 1950's but Before 1941 the district around K varntorp was pure­ production was never profitable. In June 1961 it was ly agricultural. Many farms in the area are very old decided to close down the shale oil production at and are mentioned in the earliest lists of landed Kvarntorp. At that time two other companies had property from the 16th century. In addition to already become established there, AB Svenska

A eta Phytogeogr. Suec. 68 Epiphytic lichen flora around Kvarntorp 143

plant was ea. 550 000 mj /h. This contained 265 tonnes of sulphur dioxide and about 13 tonnes of dust every day. The Ytong factory also contributed with large discharges. Up to 1946 it was calculated that 50 % of the dis­ charge came from the burning slag heaps. Subse­ ss• \I quently the contribution from this source is estimated s6e " to be reduced to about 20 % of the total amount (Hallberg 1969). Shortly after the start the vegetation around the factory began to show clear signs of influence. The National Plant Protection Institute was consulted and in 1948 they started to investigate the problem. At that time damage to coniferous forest could be observed up to about 10 km from Kvarntorp. It could also be observed that the epiphytic lichens were in­ jured. Following a proposal that they should be Fig. I. Area of investigation. The "stations" visited in mapped in the area the field work was done during 1952/53, 1967/68 and 1979. the summers of 1952 and 1953 (Skye 1958). (See Fig. 1.) In the summers of 1967 and 1968 the investigation in the K varntorp area was re-opened by lngemar Salpeterverken (later known as Supra) and AB Hallberg (Skye & Hallberg 1969) and interesting Atomenergi. The latter company was dissolved in changes could be observed. 1964. In 1961/62 Supra purchased part of the in­ stallations and is still operating at K varntorp. In 1965 the remaining part of the shale oil plant Post- 1966 was bought by Yxhults Stenhuggeri AB, who im­ The activity in the K varntorp district thus did not mediately started to build an Ytong factory on the cease with the closing down of the shale oil produc­ site. tion. The industries in the area now consist of AB Shale oil production ceased completely on 1 Oc­ Ytong, AB Supra, AB Ferriklor, AB Aminkemi, tober 1966. Prior to this the production of bottled gas SAKAB and a number of smaller companies. had been taken over by Svenska Esso. At that time In 1976 Yxhult AB closed down the field oven two other companies started work at K varntorp, the burning on account of its being environmentally un­ environmental protection company Terra Bona and acceptable. At that time the discharge of sulphur and Reci. gases amounted to about 2 400 tonnes per year From the shale oil era there now remains not only despite the ovens only being used for a minor part of the damage to the countryside but also a smouldering the total Ytong production. Since 1 October 1977 the 3 ash heap, 100 m high and containing 40 million m K varntorp complex releases 420 tonnes S02 annual­ shale ash. ly at full capacity. The southern factory at Hynne­ berg emits 60 t/year and the northern factory at Hallabrottet emits 6 t/year. Supra emits 3 420 kg so per day in addition to some hydrogen sulphide, Influence on the environment 2 and also about 4 000 kg NH3 and 240 kg N03 per The discharge into the air from the shale oil factory day. contained, for example, sulphur dioxide, hydrogen According to a letter from Lennart Persson, sulphide and other sulphurous compounds (Skye & Nature Protection Unit, Provincial Administration, O Hallberg 1969 p. 548, Fig. 2). The major discharge rebro, the discharge of so2 in the K varntorp area is was from two chimneys, about 105 m high, but some as follows: fumes came from the burning heaps of ash. In addi­ YTONG before 1977-10-0 1 1 050 tonnes/year tion there were occasional discharges of gases con­ after 1977- 10-0 1 420 tonnes/year taining hydrogen sulphide from the Lj ungstrom field on account of leakage through cracks in the SUPRA before 1977-05-01 3 000 tonnes/year limestone bed-rock. Production here ceased in 1960. 1977-05-0 1 to 10-0 1 1 600 tonnes/year The amount of smoke released from the shale oil after 1977- 10-0 1 1 200 tonnes/year

A eta Phytogeogr. Suec. 68 144 Erik Skye

During August 1979 Supra started a new sulphur does not distinguish between Physica entheroxantha recovery unit which is estimated to reduce the and Ph. fa rrea but calls them Ph. grisea. so2 discharge to less than 5 tonnes/year. During the Newcomers at the investigated stations in 1967/68 production of the shale oil company the discharge of are Cetraria pinastri, Cladonia sp., Parmelia juli­ was about 90 000 tonnes/year. ginosa, Parmelia tubulosa, Physcia dubia and Ph. so Emissions2 from SAKAB to the air are so small nigricans. New species in 1979 are Lecidea scalaris, that no specified control is required. emission, Parmelia saxatilis and P. tiliacea. The fo llowing spe­ so2 fo r example, is reported to be 43 kg/year and HCl to cies were not re-found in 1967/68: Parmelia sub­ be 20 kg/year. Ferriklor AB was founded in the aurifera and Physcia caesia ; in 1979 : Parme/ia beginning of the 1970's. The emission consists of aspidota and Physcia stellaris. On the whole the chlorine gas. The control programme requires daily changes are fairly unimportant. Some species have notes of wind direction, air temperature and disappeared from the area of investigation and others precipitation. AB Aminkemi started at K varntorp in have arrived. 1968. The plant occasionally releases ammonia into A closer analysis shows that the changes in the the air but does not appear to contribute with notable lichen flora was fairly extensive in the central parts of amounts (data obtained from Eriksson 1979). the area of investigation. The following developments The large ash heap is still smouldering. The release was illustrated at station 5: of sulphur dioxide from it is difficultto estimate but is probably fairly considerable. species 1952/53 1967/68 1979

A naptychia ciliaris X Candelaria concolor Other changes within the district X Evernia prunastri X The area has become more densely built since 1941. Lecanora conizaeoides X A new community grew up in connection with the in­ Lecidea scalaris X dustrial area. Local centres are at Yxhult-Halla­ Parmelia exasperatula X brottet, on Kumlaasen and at Skollersta. The railway Parmelia subargentifera X Parmelia sulcata was extended to K varntorp for both goods and X Physcia grisea coli. X passenger traffic. Physcia pulverulenta X The road net has changed considerably. Large sur­ ( =Physconia pulverulacea) faced roads cross the area and the traffic intensity Physcia tenella X has increased considerably. Some roads used earlier are now cut off and are no longer used. Road There are difficulties in knowing how to interpret transports dominate over rail transport and several the changes between 1967/68 and 1979. Both tracks have been pulled up. Changes in the vegetation Lecanora conizaeoides and Lecidea sca/aris were have also occurred. Fairly considerable amounts of reported in Hallberg's material from other stations. It young fo rest have grown up with the result, for ex­ is difficult to say whether their arrival at station 5 in ample, that some of the investigated stations that 1979 indicates an improvement or a deterioration, or were exposed in earlier years are now protected by anything else. The marked increase in pH at almost grown-up trees. all stations that has been registered since Hallberg's investigations must also be considered in this context. Station no. 7 illustrates a gradual deterioration of conditions. In 1967/68 there was only Physcia Results and discussion tenella left as representative of a strongly influenced rich bark flora of 8 species. Moreover, Hallberg The method used in placing the "stations" at the first recorded Hypogymnia physodes and Parmelia su/­ investigation means that they are primarily compar­ cata. These species were present neither in 1952/53 able with themselves. Consequently, in the present nor in 1979. At the latest investigation only Bacidia mapping material the extent of the differentspecies at chlorococca and Lecanora conizaeoides were pre­ all three investigations has been illustrated on one sent. and the same map. Changes in the extent are thereby Checks of the species determinations show that illustrated more clearly. Physcia ascendens and Ph. tenella were not reliably In the investigation from 1952/53 a total of 31 separated earlier. Ph. ascendens in many cases has fructicose and fo liose lichens are listed. Hallberg been poorly developed and lacked typical helmet­ (1969) lists 3 7 species and the present investigation shaped soralia. It should be considerably more com­ 38. In this respect it should be noted that Skye (1958) mon within the area than Ph. tenella. Consequently,

A eta Phytogeogr. Suec. 68 Ep iphytic lichen flora around Kvarntorp 145

the two species are listed together. were found from the earlier investigation. In 1979 Knowledge of the ecology of the different Lecidea scalaris was new on deciduous trees (oak, epiphytic lichens, and thus the natural fluctuations beech, ash, etc.) at two stations, numbers 5 and 25, in within an epiphytic community are still surprisingly the central parts of the area. small. It is clear that all species that may occur at a given time on a certain phorophyte are not present Group (Fig. 3) is an interesting group consisting 2 and that those present do not remain so indefinitely. largely of poor bark species. In all three in­ The presence of a species thus says more than its vestigations we fm d a distribution that leaves the cen­ absence. If a species exists at a certain station con­ tral part of the investigation area completely free. ditions are such that it can exist there. The reasons Station 10 is an exception, it is in a protected site and for the species not being at a certain place at a certain in 1979 had a few small plants of Cetraria pinastri. time may be many. We also know fairly little about This group of species also avoids the peripheral parts how the lichens colonize a new substrate or return to of the area and, consequently, has a relatively narrow an old site, how old they can be, how they compete area of distribution around the source of pollution. with each other, and what order of succession they Species of this group appear to have had their most have. extended distribution in the 1967/68 investigation. Studies of Parmelia sulcata over several years have taught me that the colonization proceeds con­ Group contains rich bark species that have suffered siderably faster and also that the individual lichen 3 grows considerably faster initially than was earlier severely from air pollution. In the first investigation believed. It is not until the individuals have reached a there were still remnants of the rich lichen flora pre­ certain size that the radial growth decreases (cf. sent before the arrival of the shale oil plant. The Topham 1977 p. 40 ff.). following five subgroups have been distinguished. In order to understand this process more clearly it (a) (Fig. 4) The largest consists of no less than 10 is probably better to work with groups of species than species, namely Evernia prunastri, Parme/ia ace­

with individual species. However, there are dif­ tabulum, P. exasperatula, Physcia ascendens + Ph. ficulties in grouping the material so that species with tenella, Ph. fa rrea + Ph. enteroxantha, Ph. pulver­ similar requirements are classified together. ulenta = Physconia pulverulacea), Ramalina frax­ ( Du Rietz (1945) divided the epiphytes into poor inea and Xanthoria parietina. bark species and rich bark species. This is a very The three species Evernia prunastri, Physcia

crude division that includes species with a large pulverulenta = Physconia pulverulacea) and Xan­ ( number of different requirements. Another sub­ thoria parietina appear to be the most commonly oc­ division is found in Skye (1968) which is based on the curring ones in this subgroup. pH of bark. In the present paper the following groups (b) (Fig. 5) A subgroup that has suffered severe set­ have been distinguished : backs is Anaptychia ciliaris, Physcia aipolia and Ph. 1. Species mainly fo und in the central parts of the orbicularis. Of these Ph. orbicularis is the species area. that now goes furthest into the area. Earlier, the most 2. Species that avoid the central and peripheral parts common species was A. ciliaris which, consequently, of the area. has suffered the greatest decimation. 3. Species that in 1979 primarily occur in the (c) (Fig. 6) Species in this subgroup are even more peripheral parts of the area but which earlier had a sensitive, namely Ramalina fa rinacea, Xanthoria considerably greater distribution throughout the candelaria and X. polycarpa. R.farinacea appears to entire area. This group is the largest and an be the most common and has invaded several stations attempt has been made to divide it into sub­ in the 1979 investigation. The least common is groups. Xanthoria candelaria, which also demonstrates a 4. Species present more or less over the entire area, certain recovery. even in 1979. (d) (Fig. 7) Four of the brown species of the genus Parmelia have been classified in one subgroup. These

Group (Fig. 2) comprises Lecanora conizaeoides are Parmelia aspera = P. exasperata), P.fuliginosa 1 ( and Lecidea scalaris. Lecanora conizaeoides, unfor­ = P. glabratula), P. subargentifera and P. sub­ ( tunately not in the 1952/53 investigation, occurred in aurifera. Between the two first investigations great 1967/68 at 7 stations. In 1979 it was found at 12 alterations took place in the distribution of this group. stations, 10 of which being in the most exposed parts Candelaria concolor can also be classified in this of the area. It is interesting to note that no recoveries group with advantage.

A eta Phytogeogr. Suec. 68 146 Erik Skye

Fig. 2. Group 1 : Lecanora conizaeoides and Lecidea scalaris.

i

1 I I I � ---- I � /

Fig. 3. Group 2: A tectoria sp., Cetraria chlorophy/la, C. glauca (= Platismatia glauca), C. pinastri, Par­ melia bitteriana, P. saxa­ tilis, Hypogymnia tubu­ losa, Parmeliopsis ambigua, Pseudevernia fu rfuracea and Usnea sp.

A eta Phytogeogr. Suec. 68 Epiphytic lichen flora around Kvarntorp 147

Fig. 4. Subgroup 3a : Ever­ nia prunastri, Parmelia acetabulum, P. exaspera­ tula, Physcia tenella + ascendens, Ph. grisea (coli.), Ph. pulverulenta (= Physconia pulverulacea), Ramalina fraxinea and Xanthoria parietina. � �ll.....

Fig. 5. Subgroup 3b: Ana­ ptychia ciliaris, Physcia aipolia and Ph. orbicularis.

A eta Phytogeogr. Suec. 68 148 Erik Skye

Fig. 6. Subgroup 3c: Ramalina fa rinacea, Xan­ thoria candelaria and X. \ �. polycarpa. J

•• tF ' I • • ' !. ' ' ' ' ' eei6

Fig. 7. Subgroup 3d: Can­ de/aria concolor, Parmelia aspera (= P. exasperata), P. fu liginosa ( = P. gla­ bratula), P. subargentifera and P. subaurifera.

A eta Phytogeogr. Suec. 68 Epiphytic lichen flora around Kvarntorp 149

Fig. 8. Subgroup 3e: Par­ melia tiliacea, Physcia du­ bia, Ph . nigricans (= Phaeophyscia nigricans) and Xanthoria fa llax.

io�o o�o

- - -�·-­ � ..

Fig. 9. Group 4: Hypo­ gymnia physodes and Par­ melia sulcata.

A eta Phytogeogr. Suec. 68 150 Erik Skye

(e) (Fig. 8) Very sparsely occurring species are the pollution situation is improving. However, the pH

Parmelia tiliacea, Physcia dubia, Ph. nigricans = might increase too rapidly for a development of this ( Phaeophyscia nigricans) and Xanthoriafallax. All of kind. The presence of Cetraria pinastri at station 10 them are fo und in the periphery of the area and are might possibly be interpreted as a sign of the lacking in a large area in the centre. X. fa llax goes described development. Otherwise it is difficult to see furthest into the area. other trends than those already described. Group 3 has been divided into 5 more or less well Group (Fig. 9) consists of Hypogymnia physodes defined sub-groups. The subgroup with most species 4 and Parmelia sulcata. The widest amplitude in the (subgroup a) penetrates furthest into the area. It entire material appears to be found in Hypogymnia appears as if the conditions stabilized fairly quickly. physodes, fo llowed by Parmelia sulcata. In 1952/53 Admittedly, the decline continues between the two H. physodes penetrated negligibly further into the first investigation occasions and to some extent it central parts of the area compared with later in­ continues between the second and the third inventory vestigations. P. sulcata went considerably further in but the group still covers a large part of the area of in­ towards the centre and was more numerous in the vestigation. In order to obtain a more detailed picture periphery than later. This species thus approaches of the development, study must be made, for ex­ group 2 as regards distribution. In 1967/68 H. ample, of the abundance relationships of the species physodes was negligibly further in towards the central included. parts of the area. Candelaria concolor has a distribu­ The group with A naptychia ciliaris etc. tion that is similar to that of Parmelia exasperatula. (subgroup b) has suffered a very strong decline between 1952/53 and 1967/68 but subsequently the changes have been unimportant. - The brown Changes in distribution Parmelia species (subgroup d) declined very strongly Naturally, sub-divisions of the material in another between the investigations in 1952/5 3 and 1967/68 in manner might give very different results. The groups the southern parts of the area of investigation. that have been distinguished consist of lichens with, However, these species have started to invade the in my opinion, similar distributions. I interpret this to northern parts of the area. Following this invasion a imply that the species included have approximately new decline appears to have taken place so that in the same susceptibility to air pollution. Therefore, 1979 these species are only found in the peripheral without emphasizing this too strongly, tendencies for parts. Candelaria concolor also suffered very serious the species either to return or to continue their decline set-backs between the first and secondinvestigatio ns should be reflected in the material. but even here a certain recovery has occurred. Group A common fe ature of all species included in group c is lacking within large parts of the central part of the 1 is that they occur deep in air polluted industrial area of investigation. However, its distribution in areas and cities (cf. fo r example Skye 1968). In most 1979 appears to show some recovery. - Subgroup e cases the conditions do not allow any other lichens to shows fairly large alterations from one investigation live at the stations in these areas. If the air pollution to the next but throughout it is characteristic that all situation is improved these two species will either dis­ finds are in the peripheral parts of the area. One of appear through competition or will become very the two new Parmelia species, namely P. tiliacea, sparse. Lecanora conizaeoides extends furthest from belongs to this group. P. saxatilis has been placed in the industrial area in the direction of the prevailing group 2. wind. Since 1967/68 the species has spread to The total picture of group 4 gives the impression stations 38 and 56, two stations that illustrate by that no particular changes have occurred with regard means of other lichen groups conditions that indicate to the two species included. If both species, however, noticeable influence of air pollution. are examined separately it is found that Hypogymnia The species in group 2 appear to fm d suitable con­ physodes has become more common in the periphery ditions within a zone around the source of pollution. of the area since 1952/53 and that this development ln the central parts of the investigation area the air continued after 1967/68. A certain migration pollution situation is too severe, in the periphery the towards the centre occurred following the second in­ competition from other species is probably too strong vestigation. The change for Parmelia sulcata is, on or else the bark pH is too high for many of these the other hand, small. A certain degree of air pollu­ species. If this reasoning is correct it is possible that tion appears to favour these two species which can the species of group 2 will spread in towards the survive in fairly acid environments. If the acidity source of pollution and will become more sparse in becomes too high they disappear and provide room the periphery of their present distribution now that for Lecanora conizaeoides, for example. On the other

Acta Phytogeogr. Suec. 68 Epiphytic lichen flora around Kvarntorp 151

\

en c:: .Q .2 00.

I

I •

/,'

• . 'J f, \_ L,'

no A naptychia • ®.5 ,' 0 ..r,: poorly represented t� e 0 normally represented ' e ® ecfo richly represented e ® )

Fig. 10. Changes in dis­ tribution and cover degree of A naptychia ciliaris from 1952/53 to 1979. hand, if the acidity is too low at least Hypogymnia but was in better condition and more numerous at the physodes will fm d competition difficultwith the most second investigation. However, in 1979 it was no tolerant species in group 3. longer at this site. In 1967/68 the innermost site was station 17, but unfortunately it is not included in the 1979 material, Changes in abundance when the innermost fm d was at station 20. The map A naptychia ciliaris has been randomly chosen to il­ indicates a continued decline as the species has dis­ lustrate the changes that have taken place in the dis­ appeared from a further number of stations (e.g. tribution and cover degree of the individual species stations 33 and 53) but it is difficultto know whether (Fig. 10). this is a question of normal fluctuations or not. As illustrated by the map, the species has dis­ Stations 30 and 55 showed a return of this species but appeared from the innermost sites (stations 1, 5, 7 this may also be a question of normal fluctuations. and 39) between the first and the second investiga­ The recorded cover degree also appears to support tion. This was an expected development when con­ the impression that the decline is continuing. Ana­ sidering how severely influenced the species was in ptychiaciliaris is now richly abundant only at station 1952/53. It is also poorly represented at station 18, 45 if station 58 is disregarded, where it is not corn-

pH

6 X X 0 0 X X

X 0 0 X X X 0 0 5 0 0

• • • • • Fig. 11. Comparison between the 4 • • years 1967/68 and 1979 with • regard to pH at certain "stations" along an east-west line through the 3 area of investigation. 1967 (•); 1968 (o): 1979 (x); stn station. = 53 39 29 13 9 2 10 22 42 55-56 (stn. no.)

Acta Phytogeogr. Suec. 68 152 Erik Skye

SUMMARY 53 39 29 13 9 2 1 10 22 42 56 55 Anaptychia ciliaris 0 0 0 )( The examination of material from 1979 and the com­ Physcia enteroxantha (&1 0 0 0 (&1 parisons made with earlier investigations does not Parmelia sulcata support Hallberg's optimism with regard to an im­ ['g) (&1 0 0 0 lecanora conizaeoides )( )( provement already in 1967/68. It is primarily among ® 0 ® 1 � � lichens group 2 that a return to the area has taken in I �� place but this may be suspected to depend on the 0 occurrence in Skye (1958). ( lecanora conizaeoides not included. ) occurrence in Hallberg time-factor rather than on an improvement of con­ 0 (1969 p.39l. x present occurrence (1979). ditions. Schematically, the changes in the lichen flora in Fig. 12. Occurrence of certain lichens at "stations" the area could be presented as in Fig. 13. analysed with regard to pH of the bark (Fig. 11).

pletely healthy. On the other hand Hallberg noticed Lichen- free, or group 1 certain indications of improvement in the condition and the map also gives some support fo r this inter­ pretation as the species was abundant at 9 stations in - Slow 1952/53 and only at 5 in 1967/68. changes

Changes in bark pH It is interesting to see whether the reductions in so2 increasing pH emission during recent years are reflected in an in­ group 3 / group 2 crease of the bark pH. In Hallberg's material stations decreasing pH 7 29, 13 and 10, for example, have very low values. As Fig. 13. Changes in the lichen flora's composition on bark the district is lime-rich and the industrial activity when exposed to different types of air pollution. Bark pH should produce fairly large amounts of dust it is indicates the air pollution situation. logical to expect an increase in the pH level, and this indeed has also occurred. Fig. 11 shows that the pH of the bark has in­ creased by 0.5-1.5 units and is approaching the nor­ mal fo r the respective species of tree. It should be observed that the collection of the bark samples REFERENCES probably did not tak place at the same time of year e Du Rietz, G.E. 1945. Om fattigbark- och rikbarksam­ in the two investigations. However, it is hardly prob­ hallen. - Sv. bot. Tidskr. 39:147- 1 50. able that any seasonal fluctuations of bark pH would Eriksson, S. 1979. Miljon i Kvarntorp, Kumla kommun. ­ be so high. The most probable situation is that the pH Lansstyrelsen i Orebro tan. - Orebro. of the bark relatively quickly returns to a fairly nor­ Hallberg, I. 1969. Forandringar i den epifytiska lavfloran mal value and that the return of species to the area kring Kvarntorp. - Vaxtbiologiska institutionen, Upp­ that appears to be in progress will continue. sala (mimeographed). If Hallberg's Fig. 6 (1969) is compared with the Skye, E. 1958. Luftfororeningars inverkan pa busk- och present occurrences and notes from the first in­ bladlavfloran kring skifferoljeverket i Narkes Kvarn­ vestigation a strong decline between the two first in­ torp. - Svensk bot. Tidskr. 52:1 33-190. vestigations can be seen, whereas the changes from 1968. Lichens and air pollution. - Acta phytogeogr. suec. 52:1-123. 1967/68 until 1979 are limited. It should be noted Skye, E. & Hallberg, I. 1969. Changes in the lichen flora Lecanora conizaeoides that is not in the material following air pollution. - Oikos 20:547-552. from 1952/53. The extended distribution that Topham, Pauline, B. 1977. Colonization, Growth, Succes­ Hallberg reported can hardly be confirmed by this sion and Competition. - Lichen Ecology. Seaward ed. comparison. London. pp. 31-68.

Erik Skye, Institute of Ecological Botany, Uppsala University, Box 559, S-751 22 Uppsala, Sweden

A eta Phytogeogr. Suec. 68 Dynamics and Production of Semi-natural Grassland Vegetation in Fennoscandia 1n Relation to Grazing Management Eliel Steen

Under the cold temperate, semi-humid climate of as well as the convergent development of the Scandinavia, natural grasslands can exist only as successions towards a rather simple final stage of low small refuges in marine and lacustrine litoral zones, diversity dominated by Poa pratensis. The highest on delta land, on talus slopes and other spots where diversity other than in some deciduous fo rests is tree and shrubs cannot persist or appear in later fo und in the firstcircle, i.e., grazed but not intensively successions. This is especially the case within areas managed types of grassland. dominated by the acid archaic rocks and glacial till, typical of most of the Fennoscandian shield. The soil In fo rmer times, manure was a limited resource, The semi-natural grass lands preferably used in arable fields. The grasslands Both the taiga and the mixed deciduous-coniferous therefore became gradually depleted of plant fo rests of Northern Europe have been influenced by nutrients. The impoverishment of the grassland soils domestic grazing animals for thousands of years and went on for hundreds of years and did not stop until have given rise to semi-natural grasslands (Steen industrial fertilizers became available some 100 years 1958). In the past, periods of more extensive grazing ago. Even after their introduction, comparatively in the forest commons gave vast areas of open restricted amounts of plant nutrients have been grasslands, especially in South Sweden. Where the grazing ceased, as often happened due to shifting of DRY Pi nus grazing areas, the forest returned (Malmstrom 1938). silvestris Quercus Helic ­ robur If the grazing continued without interruption, a estuc otrichum ovina � pratense treeless or sparsely tree-covered grassland landscape Festuca rubra developed, often in the neighbourhood of the farm village.

Grazing alone cannot keep these semi-natural Poa pratensis grasslands in a steady state. Clearing, cutting and BASIC burning of woody plants are also necessary parts of ) the management fo r their persistence as grasslands.

These practices have been of great importance in Sesleria shaping a very distinct pastoral landscape. This in­ caerulea fluencehas been greatest in the southern part of Fen­ noscandia, where deciduous forests are also fo und Fraxinus excelsior among the predominant coniferous types. The vegetation dynamics as influenced by grazing WET domestic animals can be summarized as in Fig. 1. - Intensified management The circles represent increasing human influence No grazing from the periphery to the centre, as grazing, burning, 2 Grazing and occasional clearing clearing and manuring. The outer zone contains the 3 Grazing, clearing and occasional fertilizing Grazing, clearing , fertilizing different forest types from which the grassland 4 vegetation has developed. The figure also shows the Fig. 1. The dynamics of semi-natural grassland vegetation influence caused by the soil water gradient (ver­ in Fennoscandia. Vegetation represented by quantitatively tically) and the soil nutrient gradient (horizontally), important species.

A eta Phytogeogr. Suec. 68 154 Eliel Steen

DRY Rate of growth 1 1 Kg ha- day -

50

0 April May June July Aug Sept Oct

� = Poa praten1s - type

,_- = Agrost1s - Poa - type

-- = AgrostJs tenu1s-type Fig. 3. Net primary production above-ground over the 1 300 vegetation period in three grassland types.

WET Fig. 2. Net primary production above-ground, tree-less DRY seminatural grasslands in Fennoscandia (see Fig. 1). Dry matter, kg per hectare and vegetation period. spread on natural grassland. Their plant nutrient status is therefore generally rather poor. Most grassland soils on dry land developed from podsols and, although the soil profile is masked by humus down to 25 centimeters, the podsol character can ACID BASIC oftenbe detected, at least chemically. However, when well-fertilized, a development towards brown earth takes place. In dryer regions, such as eastern Den­ mark and eastern Sweden, natural brown earth exists, especially in areas with calcareous subsoil, e.g., on the Baltic islands. In many parts of the world, grazing is correlated with misuse of land. Overgrazing leads to damaged WET plant cover, soil deterioration or even soil erosion. In Metabolizable energy : . . .. = megajoules per kg OM cold humid climates, overgrazing is not very common Dig. crude protein: -- = grams per kg OM and does not cause such severe damage as in warm, 1 M joule = 0. 239 Meal arid regions. Examples of overgra zing leading to soil Fig. 4. Energy and protein in the herbage of different semi­ deterioration are fo und on calcarous rocks with a natural grassland types. thin till layer (Rosen & Sjogren 1973) and on the Calluna heath in western Scandinavia (Damman 195 7). (Steen 1971). It is therefore useful to discuss some of these agronomic studies. Fig. 2 presents the agronomically relevant net primary production from Production and productivity the different grassland types. It should be emphasized Very few determinations have been made of net that the figures refer to grassland areas free of trees primary production above- and below-ground in and shrubs. Woody species may be of some import­ Scandinavian semi-natural grasslands. On the other ance in the outer circle, i.e., grazing and occasional hand, considerable amounts of data have been clearing, but no relevant data are available. collected on the economically relevant production, A typical fe ature of grazed semi-natural i.e., net primary production above-ground, harvested grasslands is the great variation of growth rate or at a certain stubble height. The production below­ productivity during the growing period from early ground and the uneaten or uncut residue above­ May to late October. The integrated growth rate ground can be roughly estimated from these data curve (Fig. 3) is typical of normal grassland. It has a

Acta Phytogeogr. Suec. 68 Dynamics and production of grassland vegetation in relation to grazing 155

1 1 Head ha- Live weight gain, kg ha-

7 - =cattle , months, graz ing days -- = cattle, 8 months, 130 grazing days 8 130 . . / " 20 " ' 130 " " " '20 .. ,130 " / . - = sheep, ewes with lamb, gr days --- = sheep, ewes with lamb, gr days / - - 160 6 160 / 300 - = suckling cow w1th calf , " -- - sucklmg cows w1th c;:�lf / 180 " . 180 " / / / 5 / / / / / / 200 / / / / 3 / / / _...... - / - / - / - 2 - 100 / - / ------

0 � 0 +------.------,------,------.-- 0 1000 2000 3000 4000 0 1000 2000 3000 4000 1 1 1 1 Net primary production above- ground , OM kg ha- yr- Net primary production above -ground , OM kg ha- yr- Fig. 5. Relation between stocking rate and net primary Fig. 6. Relation between net production and net primary production. production.

peak in June. In dryer types the peak is more often > 50) per cent of the net primary production pronounced than in moister ones. above-ground. Besides the importance of moisture and pH of the There is a positive correlation between manage­ soils, the management of the semi-natural grasslands ment (notably fertilization) of semi-natural grassland, is decisive for their productivity. With grazing alone and the quality of the herbage. There is also a tenden­ they remain low productive. Fertilization will not cy to find the best quality in intermediate (i.e., not too only change the composition of the plant cover but wet and not too dry) grassland types (Fig. 4 ) . also increases the production. With a good supply of phosphorus and potassium, each kilogram of Secondary production nitrogen added per hectare will increase the produc­ tion by about 15 kilograms on a DM basis. This Under the assumption that 25-60 per cent of the her­ amount is a little less on dryer and little more on bage produced is eaten, a certain number of animals wetter types (Giobel & Steen 1960). can be stocked on any area. Their production can be measured as weight increase and/or milk yield. They can also be kept on maintenance alone, which means Grazing intensity and herbage consumption that there is no net production. Figs. 5 and 6 sum­ The number of grazing animals per unit area is very marize results from south and central Sweden on important for the effect of grazing on the vegetation. carrying capacities and net secondary production. Several experiments with different stocking rates have been performed in Scandinavia. For each type of grassland, an optimal rate can be determined or es­ The stability of the semi-natural state timated, which will maintain the vegetation and the Grazing with domestic animals and management soil in its semi-natural labile state of equilibrium. practice will keep the semi-natural grasslands in a However, management in the fo rm of clearing the labile state (Ryberg 197 1). However, during the last thorny shrubs at some intervals is also necessary. 25 years the number of domestic grazing animals has In most natural grasslands the vertebrate her­ decreased considerably throughout the whole of Fen­ bivores consume only a small part of the net primary noscandia. Grazing is becoming increasingly con­ production. A minor fraction is taken by the in­ fm ed to sown and intensively managed longterm vertebrate grazers and the rest goes to the decom­ pastures or short rotation leys. Only the best semi­ posers (van Dyne 1973; Marshall 1977). In semi­ natural grasslands are still grazed. The grazing natural grasslands the grazing primary consumers animals are sheep, young cattle, saddle-horses and play a much greater role and they are dominated by special beef-cattle breeds. the domestic mammals, which take at least 25 (and This means that considerable areas in Fennoscan-

Acta Phytogeogr. Suec. 68 156 Eliel Steen dia, approximately 2 million hectares of semi-natural REFERENCES grasslands, have been abandoned during this period. Damman, A.W.H. 1957. The South Swedish Calluna­ In their place have appeared a number of successions heath and its relation to the Calluneto-Genistetum. - towards closed forest. These dynamic processes are Bot. Notiser 110:363-398. going on in all the Fennoscandian countries and have Dyne, van G. 1973. Analysis of structure, function and been documented in several reports (Ekstam & utilization of grassland ecosystems. - N at. Resource Sj6gren 1973; Elveland 1979; Koponen 1967; Solbu Ecology Lab., Colorado State University. A progress 1976; Rodenborg 1976; Zackrisson 1976). report. However, about a million hectares of semi-natural Ekstam, U. & Sjogren, E. 1973. Studies of past and pre­ grasslands still persist within this part of Europe. sent changes in deciduous forest vegetation on bland. - Zoon, Suppl. 1:123-1 35. They increase the species diversity in a landscape Elveland, J. 1979. Dammangar, silangar och raningar - dominated by coniferous forests and also provide norrlandska naturvardsobjekt. Summary : Irrigated variation in the decidous forest zone, as well as in in­ and naturally flooded hay-meadows in North Sweden dustrialized areas and areas with intensive agriculture - a nature conservancy problem. - Statens natur­ (Petrini & Jacobsson 1975). vardsverk, PM 1174:1-124. About 300 species of grasses, herbs, mosses and Giobel, G. & Steen, E. 1960. Godslingsforsok i naturlig leafy shrubs are fo und in these grasslands. The soil betesmark. Summary : Fertilizer experiments on types are associated with different grassland types natural pastures. - Statens Jordbruksforsok, Medd. Nr and contain a rather specific fauna. Several bird 109:1-59. species are found. Koponen, T. 1967. On the dynamics of vegetation and It is essential to continue the grazing with the flora in Karkali Nature Reserve, southern Finland. - Annals bot. fenn. 4:121-218. specific animals at proper stocking rates on the Malmstrom, C. 1938. Hallands skogar under de senaste remaining semi-natural grasslands if their floristic 300 aren. Zusammenfassung: Die Walder Hallands and faunistic diversity and other biological values are wahrend der letzten 300 J ahre. - Meddn Stat. to be conserved. Continued grazing also means con­ Skogsforsoksanstalt 31:171-300. servation of an attractive landscape, which otherwise Marsh all, J .K. (ed.) 1977. The belowground ecosystem. A would be lost. synthesis of plant-associated processes. - Range Sci. Dept., Sci. Ser. 26. Colorado State University. Petrini, F. & Jacobsson, L. 1975. Idle land in Sweden: problems and solutions. - Swed. J. Agric. Res. 5:91-99. Rodenborg, L. 1976. Bodennutzung, Pflanzenwelt und ihre Veranderungen in einem alten Weidegebiet auf Mittel-Oland, Schweden. - V axtekol. Stud. 7:1-210. Rosen, E. & Sjogren, E. 1973. Sheep grazing and changes SUMMARY of vegetation on the limestone heath of Oland. - Zoon, The close relation between semi-natural grasslands in Suppl. 1:137-151. Ryberg, M. 1971. The deciduous woods on the N aset pe­ Fennoscandia and agricultural practises, especially ninsula at Tullgarn, Province of Sodermanland. - K. grazing, is discussed. The role of different manage­ Vetenskapsakad. Handl., 4 Ser., 14( 1): 1-80. ment, as grazing, cutting, burning, clearing and Solbu, I. 1976. Landskapsendringar pa nedlagte manuring, is put in relation to soil water and plant heiegarder i Vest-Agder med hovedvekt pa vegetasjon. nutrient status of different soils. Summary : Landscape changes in derelict hill-farm Primary and secondary production is discussed. areas in Vest-Agder. - Report Inst. fo r landskaps­ The connection between production and manage­ arkitektur, N orges landbrukshogskole. 120 pp. ment as well as quality of grass and management is il­ Steen, E. 1958. Betesinflytelser i svensk vegetation. Sum­ lustrated. mary: Effects of grazing on Swedish vegetation. - The stability of the semi-natural state is threatened Statens Jordbruksforsok, Medd. 89:1-82. 1971. Lantbruk och landskapsvard. Summary: The by decreased grazing and cutting. Many areas are role of agriculture in rural preservation. - Skogs­ even abandoned as agricultural land and a number of Lantbruksakad. Tidskr. 1 10:181-196. successions towards closed fo rest have developed. Zackrisson, 0. 1976. Vegetation dynamics and land use in This means losses of attractive landscapes and biotic the lower reaches of the river Umealven. - Early diversity. Norrland 9:7-74.

Eliel Steen, The Swedish University of Agricultural Sciences, Department of Ecology and

Environmental Research, S- 7 50 07 Uppsala, Sweden

A eta Phytogeogr. Suec. 68 Plankton of Lake Kariba Re-examined Kuno Thomasson

Shortly after the dam across Kariba Gorge on the samples are rather sparse, and it is difficult to make Zambezi River was completed a number of samples any comparisons between them and the results from collected in May 1959 from Lake Kariba-then 1959. filling-were examined (Thomasson 1965). The composition of plankton as reflected by the The lake had been filled by 1963, by which time it samples from 1968-70 is summarized in Table I. had assumed the fo llowing dimensions: Length 300 The sampling stations are scattered all over the lake, km; Maximum width 40 km; Area 5,250 km2; and their location has been indicated in the table by a Storage capacity at 484 m above m.s.l. 160 km3• number which refers to the grid in the map in Coche Today it is the fourth largest man-made lake in the (1968 Fig. or the one in Balon & Coche (1974 Fig. I) world, coming after Lake Volta, Lake Nasser, and 23). Lake Bratsk. Sample 162 from the Devil's Gorge is a scanty Thorough descriptions and discussions of physico­ one, probably reflecting riverine conditions. Samples chemical aspects of Lake Kariba have been published 175, 186, 187, 189, and 190 are also poor ones. All by Coche (1968) and Balon & Coche (1974). the interesting samples, viz., 173, 178, 179, 189, 192, The epilimnic waters of Lake Kariba are slightly and 193, were collected from the central basin of the alkaline, the pH varies between 7.5 and 8.5. It is a lake, which has been designated as Basin Ill by warm monomictic reservoir, the epilimnic tem­ Coche 968), stretching from the Chete Gorge area (I peratures varying between 20°C and 31°C . Of to Sibilobilo Narrows. It consists of large areas of course, in such a large lake with a complicated open water and several islands. The phytoplankton in pattern, the transparency figures show a wide range, that area is characterized by the abundance of from 50 cm to 1060 cm. The depth of the euphotic Aulacosira. In addition, Microcystis aeruginosa and zone in the limnetic region varies greatly, having a Microcystis wesenbergii have a high frequency; they range from 200 cm to 2400 cm. Following an overall might possibly fo rm water-bloom. maximum value fo the conductivity in 1959, due to the valley soils and vegetation bringing into solution Comments to Ta ble vast quantities of organic and inorganic matters, a I certain chemical equilibrium now seems to have been Anabaena sp. listed above has rather broad cells, about 10 reached. The mean conductivity for the period ,urn. The trichomes are straight, and the heterocysts are 1966-1973 ranges from 80 to 87 micromhos. spherical. Microcystis aeruginosa (Kiitz.) Kiitz., a common The initial studies on the composition of plankton plankter which until recently has been called M.flos-aquae in newly created lakes have been fo llowed up only on (Wittr.) Kirchn. em. Wesenb.-Lund. a few occasions. We know very little about the Microcystis elachista (W. & W.) Starmach has been successive stages in the development towards a stable called Aphanocapsa elachista W. & W. plankton community. Of course, as almost every lake Microcystis wesenbergii Komarek is the valid name fo r has a character of its own, generalizations should be M. aeruginosa Kiitz. em. Wesenb.-Lund. Margins of avoided. Therefore it was with great interest that I colonical mucilage are very distinct and highly refractive. studied the samples from Lake Kariba which were Scenedesmus arcuatus Lemm. forma which has been placed at my disposal through the courtesy of Andre listed in Table I is identical with the plants figured in G. Coche (Fishery Limnologist). They were collected Uherkovich (1966 Figs. 153-1 55). Micrasterias americana (Ehrenb.) Ralfs resembles the from different stations between 30 March 1968 and 3 plants which have been discussed in Thomasson (1965 p. April 1970. 14). Also Compere (1967 Fig. 215) has depicted a similar All samples were taken with No. 20 net, the aper­ one. However, the regular ornamentation of the cell wall, ture size of the mesh being about 63 ,urn. The net had and the occurrence of the supraisthmal papillae indicates a diameter of 30 cm. The sampling was made by a some relationship between this forma and M. vertical tow of 10 meters. Accordingly, many mahabulesh warensis v. dichotoma G.M. Smith. In M.

A eta Phytogeogr. Suec. 68 158 Kuno Th omasson

americana the cell wall is ornamented with scattered the changes might to some extent also have been granules or small spines, these are not arranged in rows created by the change in nutrient status, the like those in M. mahabuleshwarensis. meiotrophication. Pleurotaenium cy lindricum (Turn.) W. & W. The plants observed were all rather long, 600-6 15 .urn. Pleurotaenium elatum (Turn.) Borge. Length 690 .urn, breadth 60 .urn. In contrast to the statement in Krieger REFERENCES (1935 p. 425), these plants occurred also in chains, like P. subcoronulatum (Turn.) W. & W. Balon, E.K. & Coche, A.G. 1974. Lake Kariba, a man­ Staurastrum gracile v. protractum G. S. West described made tropical ecosystem in Central Africa. - Monogr. in 1907 (p. 128, Fig. 6:1 1) from Lake Nyasa belongs Bioi. 24. 767 pp. probably to the plants which have been grouped with S. Beadle, L.C. 1974. The inland waters of tropical Africa. ­ jloriferum W. & W., e.g., the similar but smaller S. London. 365 pp. jloriferum v. variabile Rich (1935 p. 145, Figs. 17E-G). Begg, G.W. 1974. The influence of thermal and oxygen Staurastrum johnsonii v. altius Fritsch & Rich. The stratificationon the vertical distribution of zoopla nkton plants observed are similar to those figured in Thomasson at the mouth of the Sanyati Gorge, Lake Kariba. - (1965 Fig. 11:1). Kariba Studies, 4:60-67. Staurastrum schuurmanii Rich is a plant which has Bourrelly, P. 1957. Algues d'eau douce du Soudan been seldom recorded. The measurements of Lake Kariba Franc;ais, region du Macina (A.O.F.). - Bull. Inst. plants are: length without processes 18-2 1 .um, breadth franc;. A fr. noire, ser. A, 19:104 7-1 102. with processes 63.5 .urn. 1969. Quelques remarques sur la classification des Staurastrum sebaldi v. ornatum Nordst. forma is iden­ algues bleues: 11. Stigonematales et Oscillatoriales. - tical with the one figured in Bourrelly 9 57 p. 108 7, Fig. (I Verh. int. Ver. Limnol. 17:757-760. 140). Coche, A.G. 1968. Description of physico-chemical Staurodesmus subulatus (Kiitz.) Thorn. The plants aspects of Lake Kariba, an impoundment in Zambia­ observed are identical with those figured in Thomasson Rhodesia. - Fish. Res. Bull. Zambia 5:200-267. ( 1966 p. 29, Figs. 1 :4-5). Compere, P. 1967. Algues du Sahara et de la region du lac Staurodesmus validus v. subvalidus (Gronbl.) Teil. It Tchad. - Bull. Jard. Bot. Nat. Belg. 37:109-288. should be pointed out that Lake Kariba plants have only Harding, D. 1966. Lake Kariba. The hydrology and one pyrenoid in each semicell. development of fisheries. - Symp. Inst. Bioi. 15:7-19. Keratella cochlearis pachyacantha Thorn. (Thomasson Krieger, W. 1933-39. Die Desmidiaceen Europas mit 1980) is a new member of the plankton of Lake Kariba. It Beriicksichtigung der aussereuropaischen Arten. - is rather frequent; hitherto I have seen it only in the Kryptogamenflora 13: 1. 712 pp. plankton of Lake Shiwa Ngandu. Rich, F. 1935. Contributions to our knowledge of the freshwater algae of Africa. 11. - Trans. roy. Soc. S. Afr. 23:107-160. Thomasson, K. 1965. Notes on algal vegetation of Lake SUMMARY Kariba. - Nova Acta Soc. Sci. ups al., Ser. 4, 19:1-34. In samples from 1959 there were quite a number of 1966. Phytoplankton of Lake Shiwa Ngandu. - Expl. tychoplankters and also benthic organisms. They hydro bioi. Bangweolo-Luapula, 4:2. 91 pp. were probably favoured by the special environmental 1980. Eine neue tropische Unterart von Keratella cochlearis (Rotatoria). - Int. Revue ges. Hydrobiol. conditions: submerged woodlands substituting to 65:745-747. some extent aquatic vegetation, the explosive Uherkovich, G. 1966. Die Scenedesmus-Arten Ungarns. - development of the water fern Salvinia auriculata, Budapest. 173 pp. and development of sudd communities. The situation West, G.S. 1907. Report on the freshwater algae, including was reflected in the number of desmid and rotifer phytoplankton, of the Third Tanganyika Expediton, taxa in samples from 1959. In samples from conducted by Dr. W.A. Cunnington, 1904-1905. - J. 1968-70 the numbers have decreased but, of course, Linn. Soc. (Bot.) 38:81-197.

A eta Phytogeogr. Suec. 68 Plankton of Lake Kariba re-examined 159

Table I. Plankton of Lake Kariba

CO CO CO a- a- 0 0 0 10 10 10 10 10 1"-- 1"-- 1"-- a- a- a- a-� a- a-� a-� a- � � � """: ....; .....: .....: CO CO N � o:t" o:t" Date CO 10 a- 0'1 0 0 M0 .0 a:iN Na:i � M M

o:t" 10 10 0 0 10 10 10 10 0 M N o:t" o:t" N N N N Station 0 MN Ln \D Ln 10 a- CO a- � o:t" �o:t" N N CO a- N N N N N CO \D 1"-- a- 0 N ..., Sample � � 1"-- � CO a- � � � � � CYANOPHYCEAE

Anabaena fl os-aquae Ki rchn. + +

A. spiroides Kl eb . +

Anabaena sp. + +

Chroococcus dispersus (Keissl .) Lenm . +

c. limneticus Lenm. + Gomphosphaeria lacustris Chod. +

Meri smopedia glauca {Ehr. ) Nag . +

Microcyst s aeruginosa (KUtz.) KUtz . + + + + + + + � M. elachista (W. & W. ) Starm. +

M. flos-aquae Breb. + + + + + + +

M. wesenbergi i Kom. + + + + + +

M. viridis (A. Br. ) Lemm. + +

Osci llatoria borneti i Zukal + 0. irrigua KUtz. + Pseudanabaena mucicola (Hub . -Pest. & Naum. ) Bourr. + +

XANTHOPHYCEAE

Gl oeobotrys sp. +

Trachydiscus el lipsoideus Ettl +

BACILLARI OPHYCEAE

Aulacosira agassizii (Ostenf. ) Si m. + + + + + + +

A. ambigua (Grun .) Sim. +

A. distans {Ehr.) Sim. + +

A. granulata (Ehr. ) Sim. + + + + + + + A. granulata v. angustissima ( 0. MU 11. ) Si m. + + A. nyassensis (0. MU l l.) Sim. + A. nyassensis v. victoriae (0. MUll.) Si m. + + + Cyclotel la meneghiniana KUtz. +

Cyma topleura solea (Breb.) W. Sm. + c. solea v. clavata 0. MUl l. + Diploneis subovalis Cleve + Eunoti a pecti na 1 is {KUtz. ) Rabenh. + Fragi laria crotonensis Ki tt. +

Suri rella elegans Ehr. + s. engleri f. sublaevi s 0. MUl l. + s. tenera f. subconstricta Hust. + Synedra acus KUtz. + + + + + s. bero 1 i nens is Lenm. + + s. ulna (Nitzsch ) Ehr. + + + + + + + s. ulna f. danica (KUtz. ) Grun. +

DINOPHYCEAE

Cerati um hirundinel la f. brachyceroi des Schroder + + +

Peri di ni um cunni ngtoni i ( Lenvn. ) Lerrm . + +

P. wi llei Huitf. -Kaas + + P. volzi i v. cincti forme Lef. + Peri dinium sp. +

Acta Phytogeogr. Suec. 68 Kuno Thomasson 160 Table ( conti n.)

Sample N ...., t.n CO 0\ \0 ,..... 01 0 N ...., ,..... ,...., ,..... 0\ 0\ � !::: � � � � EUGLENOPHYCEAE

Euglena oxyuri s Schmarda + Phacus unguis Pochm. + Trachelom6nas oblonga Lemm . + + T. volvocina Ehr. +

CHLOROPHY CEAE

Eudori na e 1 egans Ehr. + + + + + + Volvox sp. + + + + + + Actinastrum hantzschii Lagerh . + + + + + Ankistrodesmus fa l catus (Corda ) Ralfs + + + Botryococcus brauni i KUtz . + + + B. protuberans W. & W. + Coelastrum asteroideum De-Not. + + + c. cambricum v.intermedium (Bohl .} G.S. West + + + c. mi croporum Nag . + + proboscideum Bohl . c. + ·� c. reticulatum (Dang.} Lemm . + + + + + + + Crucigenia mi nima Brunnth. + + C. tetrapedia v. apiculata Fritsch & Rich + Francei a tenuispi.na Korsch. + Ki rchneriella arcuata G. M. Smith + Nephrocytium limneticum (G.M. Smi th} Skvja + Oocystis pusilla Hansg. + Pediastrum duplex Meyen + + P. duplex v. rugulosum Racib. + P. gracillimum (W. & w. ) Thunm. + P. simplex Meyen + + + + + + + + + P. tetras (Ehr. ) Ralfs + P. tetras v. tetraodon (Corda l Hansg. + Scenedesmus acumi natus ( Lagerh . ) Chod. + s. arcuatus Lemm . f. + s. armatus Chod . + s. denticulatus Lagerh . + + s. denticulatus v. 1 inearis Hansg. + s. ecornis (Ralfs) Chod. + + + + s. ecornis v. concavus Hortob. + s. oahuensis (Lemm. ) G.M. Smi th + s. quadricauda (Turp. ) Breb. + s. quadri cauda v. longispina (Chod. ) G.M. Smith + s. serratus f. minor Chod. + Sorastrum spinulosum v. hatoris (Cohn) Lemm. + Sphaerocystis schroeteri Chod. + + Tetraedron arthrodesmi forme v. i rregul aris Wolosz. + + T. arthrodesmiforme v. 1obulata Wolosz, + +

CONJUGATOPHYCEAE

Mougeoti a pseudoope 1 ousens is Gauth.-Liev. + Closterium aciculare T. West + c. acutum v. variabile (Lemm. ) Krieg . + c. dianae v. pseudodianae (Roy ) Krieg. + c. ehr.enbergii Menegh. + c. kuetzi ngi i Breb. + + + c. moniliferum (Bory) Ehr. + c. subulatum (KUtz.) Breb. +

Acta Phytogeogr. Suec. 68 Plankton of Lake Kariba re-examined 161

Table I (contin.)

Sample

Cosmarium contractum f. jacobsenii (Roy & Biss.) W.& W. + C. lundel lii v. nyassae (Schmidle) Krieg. & Gerl . + C. margari tatum (Lund.) Roy & Biss . + C. quadrum Lund . + Desmi di urn aptogonum Breb . + + Euastrum denticulatum v. rectangulare W. & W. + E. engleri v. madagascariense Bourr. + E. turneri W. West + Hyalotheca dissi liens (J.E. Smi th ) Breb . + Micrasterias ameri cana (Ehr. ) Ralfs + M. americana (Ehr. ) Ral fs f. + M. mahabuleshwarensis ssp. europaea (Nordst. ) Teil. + + M. novae-terrae (Cushm. ) Krieg. + M. papillifera v. glabra Nordst. + M. radians Turn . + M. radiata v. brasiliensis Gronbl . + M. radiosa v. ornata f. elegantior G.S. West + Pleurotaenium cyl i ndricum (Turn.) W. & W. + n P. ehre bergii (Breb.) De-Bary + t P. ela um (Turn.) Borge + P. ovatum v. tumidum (Mask. ) W. West + P. trabecula (Ehr. ) Nag . + Spondylosium planum (Wol le) W. & W. + Staurastrum anati num f. curtum (G.M. Smi th ) Brook + S. anatinum f. denticulatum (G.M. Smi th ) Brook + + S. anatinum v. floriferum W. & W. + S. cyclacanthum W. & W. + S. fuel leborni i Schmidle + S. furcatum (Ehr.) Breb . + S. graci le v. protractum G.S. West + + S. hexacerum (Ehr. ) Wi ttr. + S. johnsoni i v. altius Fri tsch & Rich + S. leptocladum v. africanum G.S. West + + S. leptocladum v. africanum G.S. West f. + S. leptocladum v. cornutum Wi lle + + + + S. leptocladum v. insigne W. & W. + + + s. longispinum (Ba il.) Arch . + S. messikommeri f. planctica Thorn . + S. playfairi anum Thorn. + S. pseudotetracerum (Nordst. ) W. & W. + S. schuurmani i Rich + + S. sebaldi v. ornatum Nordst. f. + S. sebaldi v. ornatum f. planctonicum Teil. + + S. setigerum v. subvi llosum Granbl . + S. stelliferum v. africanum Fri tsch & Rich f. + S. stel liferum v. corpulentum (Thorn. } Forst. + S. subanchora Gronbl . + S. subrichianum Thorn. + S. tetracerum Ral fs + + S. triundulatum f. brevibrachiatum Thorn. + S. wildemanii Gutw. + S. wildemanii v. ma ius (W. & W. ) Scott & Presc. + Staurodesmus connatus (Lund.) Thorn. + S. convergens (Ehr. ) Tei l. + S. convergens v. ralfsii f. curtus (Turn.) Tei l. +

Acta Phytogeogr. Suec. 68 162 Kuno Thomasson

Table (eontin. )

N Lt"l I.D ...... 0'> 0 N M Sample � ...... � � 00 00 00 0'> 0'> 0'> :£ s. eurva tus V. latus (Seott & Prese.l Thorn . + s. subulatus (Kutz. ) Thorn . + s. subulatus v. nordstedtii (G.M. Smi th ) Thorn. + s. tri angul ari s (Lagerh. l Tei l. + s. va1 idus v. subval idus (Gri:inbl .) Teil. +

HYPHOMYCETES

Aeti nospora megalospora Ingol d + +

Clavi ri opsi s aquatiea De-Wi ld. +

PROTOZOA Bertrami a asperospora (Fri tseh) +

Cothurnia sp. +

Lesquereus i a spira 1 is (Ehr. ) + Paradi leptus eoni e·us Wenrieh +

Peri aei neta lingui fera Col lin. +

ROTATOR lA

Anuraeops is fiss a fiss a (Gosse) + + +

Asplanehna sp. + Braehionus ealyeifl orus ealyeifl orus Pallas + +

B. fal eatus Zaeh. + +

B. patulus patulus (Mul ler) + +

B. quadridentatus quadridentatus Herm. + B. sess·i lis Varga +

Conoehi 1 oi des natans (Sel igo) + Di peuehlanis propatula (Gosse) +

Euehl ani s di l atata dilatata Ehr. + +

Hexarthra mi ra (Hudson) +

Hexarthra sp. + +

Keratel la eoehlearis eoehlearis (Gosse) + +

K. eoehlearis paehyaeantha Thorn. + + + +

K. tropi ea tropi ea (Apstein) + + + + + + +

Leeane bul la bul la (Gosse) + + +

L. leontina (Turn. ) +

Ploeosoma sp. +

Po lya rthra vul ga ri s earl in + + + + +

Synehaeta s p. +

Tetramasti x opol iensis Zaeh . + +

Tri ehoeerea eh a ttoni (Beaueh.) + +

T. eyl indri ea (Imhof) +

T. elongata (Gosse) +

T. longi seta (Sehrank) + + T. simi 1 is (Wierz.) +

CRUSTACEA

Bosmi na longirostri s (O.F .M. ) + + + + + + +

Ceri odaphni a eornuta Sars + +

Daphnia lumhol tzi Sars + +

Diaphanosoma exei sum stingel ini Jenkin + + + + + Mesoeye 1 ops 1 eueka rt i (Claus ) + + + + + + + +

Kuno Thomasson, Institute of Ecological Botany, Uppsala University, Box 559, S-75 1 22 Uppsala, Sweden

Acta Phytogeogr. Suec. 68 Rhodomonas minuta and Rhodomonas lens (Cryptophyceae) - Aspects on Form-variation and Ecology in Lakes Ma.laren and Vattern, Central Sweden Eva Willen, Mats Oke Frank Gonzalez &

Rhodomonas and especially the species minuta, by adding a phosphorus reduction step (with an effect described by Skuja (1948 p. 346) is a very frequently of ea 90 %) to sewage plants. The development in the occurring nannoplanktic organism fo und in a wide four largest lakes of Sweden has been fo llowed by variety of fresh-water ecosystems all over the world, comprehensive and continuous limnological in­ ranging from ultraoligotrophic high alpine lakes vestigations before and after the measures under­ (Nauwerck 1966 pp. 14-15) through oligotrophic taken. The most evident results have been recorded in and eutrophic temperate lakes (Willen, T., 1959 p. the Lakes Malaren and V attern, central Sweden 262; Nauwerck 1963 p. 97; Ahlgren 1970 p. 360; (Fig. 1). Munawar & Munawar 1975 p. 719; Olsen & Willen Miilaren, the third largest lake of the country, is composed 1980 p. 174) to subtropic (Pollingher 1978 p. 237) and tropic waters (Lewis & Weibezahn 1976 p. 178; of many basins of different, mainly eutrophic, character Lewis 1978 p. 219). and here the principal interest will be fo cused on the north­ Skuja originally found Rhodomonas minuta in easternmost part - the Ekoln Bay (Fig. 2). This bay is stratified in summer and winter, has a surface area of 30 mesotrophic-eutrophic central Swedish lakes es­ km\ a maximum depth of 50 m and a mean depth of 15 m. 1948 pecially fr equent during the cold seasons (Skuja The theoretical residence time of the water is calculated to p. 34 7) while Rhodomonas lens described by Pascher be about one year. A relation of ea 50:1 between drainage & Ruttner (Pascher 1913 p. 103) originally was area and lake area suggests a heavy load from the environ­ found in waters of more oligotrophic character. Later ment, which is densely populated and has considerable fm dings of both species have, however, revealed a agriculture. Ekoln Bay increased its concentration of total broad trophic affinity. The question of relevant phosphorus five-fold from 1934 to 1964 due to influence of nomenclature of species and varieties within the sewage water from the gradually expanding town of Upp­ genus Rhodomonas has recently been discussed by sala-a town which nowadays is the fo urth largest in the Javornicky (1976) who proposes a fusion of the country. The concentration of total phosphorus reached a growing season mean value of 140 ,ug t1 during the 1960's Rhodomonas species minuta, lens and lacustris un­ with maximum values of 250 ,ug 1- 1• Thirty years earlier the der the name with nomenclatoric pnonty - same concentrations were 40 ,ug t 1 and 70 ,ug 1 1, respec­ -Rhodomonas lacustris Pascher & Ruttner. The tively. In the 1970's the concentration started to decrease reason is the large variation in forms which obviously due to a series of years with extremely low water flow and exists among these three species and which makes it a successive introduction of chemical treatment of waste difficultto separate them in a reliable way. Evidently water. From 1973 when the final stage of the phosphorus the flagellates depicted from central European lakes removal programme in the sewage plants was completed (see Pavoni 1963 p. 341) under the name Rhodo­ there was a further decline in the phosphorus concen­ monas lacustris are very similar to some of Skuja's trations of the Bay down to values of ea 50 ,ug t 1• fo rms of Rhodomonas minuta. Viittern is one of the largest oligotrophic water-bodies of The aim of this paper is to discuss the form and Europe (Fig. 2). It has a surface area of ea 2 000 km2 and a size variation of Rhodomonas minuta and Rhodo­ maximum depth of 128 m. The mean depth is 39 m. Due to monas lens considering their annual and long-term its shape the lake is easily affected by winds which often development in two large Swedish lakes of quite break the stratification. The watermasses are well oxy­ different character, and try to estimate the relevance genated from surface to bottom. Owing to a small drainage area (relation 2: 1) and a large lake volume the residence of their fusion under one single species name. time of the water is extremely long, 58 years. A unique background material on Lake Vattern available from the The studied lakes 1880's enables comparisons to be made of changes in water quality. It has been possible to trace negative effects In many Swedish lakes it has been possible to break due to discharge of sewage water from the 1940's when the an accelerated eutrophication due to human activities phosphorus load was calculated to be 50 tons. In the late

A eta Phytogeogr. Suec. 68 164 Eva Willen, Mats Oke Frank Gonzdlez & 1960's the corresponding load was 200 tons. The con­ centration of total phosphorus in the lake has, however, not increased concurrently with the loading due to the long retention time of the water. Maximum values of 11 f.Lg I- ' and a mean concentration during the growing season of 7 f.lg t' were measured in 1938 while corresponding measurements in 1967 were 19 and 15 j.lg t 1 in the trophogenic layer. The conditions in Lake V attern in the late 1960's were still oligotrophic but the tendency towards deterioration was clear. A water conservancy plan was adopted in 1966-196 7 and since then far-reaching measures fo r improvements of the water condition have Fig. l. Lakes Vattern and Malaren, Central Sweden. been undertaken. Most of these measures were completed in 1972- 1973. The phosphorus load entering the lake decreased drastically to a minimum value of 64 tons in 1976 (Fig. 3). Depending on differences in the water flow this figure may rise to ea 80 tons which still will be enough to retain the highly oligotrophic character of the water. The average concentration of total phosphorus dur­ ing the late 1970's has been 7 J.Lg I- '. These nutrient reductions have affected the phyto­ plankton flora in many ways and especially on the species level (Olsen & Willen 1980) but the discussion here will concentrate solely on the genus Rhodomonas.

Sampling and methods Regular phytoplankton studies in Malaren and the Ekoln Bay have been in progress since 1965 with sampling mostly every month or every third week during the growing season May to October. The samples have been taken in the most productive up­ per 0- 1 m layer with a Ruttner sampler and im­ mediately preserved with a 12IK solution sup­ plemented with acetic acid (Hobro & Willen 1977 p. 809). All sampling has been performed at a central site in the Bay. In Lake Vattern regular phyto­ Lake Vdttern plankton studies started in 1967. After a two-year

tonnes

Lake Mdlaren

0 10 20 km

Fig. 2. Lake Vattern with the two sampling sites, and Lake Fig. 3. Supply of phosphorus to Lake Vattern 1967-1 976 Malaren with the Ekoln Bay shaded. with reference to diffe rent sources (from Olsen & Willen 1980).

A eta Phytogeogr. Suec. 68 Rhodomonas minuta and R. lens -form-variation and ecology 165

break the sampling continued in 1970 with col­ appeared in such large numbers that they could be lections every month from April to October, here counted reliably (Willen, E. 1976 p. 268). From very representing the growing season. These samples have rare occurrence they suddenly increased their cell been taken as mixed samples from the trophogenic number to a growing season mean value of 10 000 layer (0-20 m). The sampling site that has been con­ cells l-1 in 197 5 and in the two following years the tinuously investigated since 1967 is situated in the population rose to 20 cells (Fig. 4). The 000 1-• southern part of the lake but experiences from a number of Rhodomonas minuta cells has slowly and dense sampling network indicate large uniformity all successively diminished in concurrence with the over the lake. nutrient load from a growing season mean cell To get information about the dynamics as well as number of 190 000-200 in 1967 and 1970 to ea 000 the fo rm and size variations of different species, 120 000 after 1972. samples were collected more frequently in 1975 in the To obtain information on the seasonal develop­ Ekoln Bay and in 1978 in the northern central part of ment, these species were counted in the more fre­ Lake Vattern. Samples taken represent mainly the quently sampled material from 1978, which revealed upper wind-mixed layer 0-2 m in Ekoln and 0-5 m two maxima fo r Rhodomonas minuta during the in Vattern. In the fo rmer, samples were collected growing season, one in April and another one in every week until the end of October and in the latter August (Fig. 5). These results coincide well with every fortnight until the decline of the springbloom those attained during earlier years although the and after that once a month throughout the growing sampling then was less frequent. The real spring max­ season. Due to the large number of samples, all imum might possibly have occurred earlier because drawings and analyses have been made on preserved the Rhodomonas population was steadily declining material-a matter of inconvenience especially when from the April sampling occasion. Rhodomonas lens studying flagellates which may change forms and has a long spring development with a population of loose their flagellas. The Rhodomonas species seem, uniform number from the end of April to the end of however, to retain their shape well and are easily May and after that another peak in September. recognizable in iodine-fixed material. Besides, the Whereas Rhodomonas minuta is present all through single pyrenoid in the cells is often fully visible in this the sampling period, cells of Rhodomonas lens are preservation fluid. The rough outer contour of some of the depicted organisms is hence to be interpreted as a preservation artefact. The cell measurements

were made in 400x magnification and the error of the Rhodomonas minuta micrometer was ea 1 ,urn. The volume calculation of the species was performed after adjustment to suit­ able stereometric shapes and the density of the cells b 0,1 L­ QJ was approximated to 1. Rhodomonas were regarded .D E as cone + sphere/2. In the large-scaled continuous in­ :J c Rhodomonas lens vestigations, median volumes were calculated for - - ..------r--, these organisms based on a large number of 0 - -,------.---, ------.--�--- measurements spread throughout the sampling 1967 70 71 72 73 74 75 76 77 period. The organisms were counted according to the Fig. 4. Development of Rhodomonas minuta and Utermohl technique in settling chambers of 10, 50 of Rhodomonas lens in Lake Vattern based on mean cell 100 ml size depending on plankton density (Utermohl numbers during the growing seasons (April-October) 1958). The objectives used were of 40x or 100x 1967- 1 977. magnification.

Rhodomonas mmuta Results �

.....� 0,10 Dynamics 0 O,OS During the long-sequential phytoplankton investiga­ ll Rhodomonas lens - · - - ·- ---.A - --·...... - · ------tion of Lake Vattern the species Rhodomonas minuta c� 0 +----.------��==��-.------,-- has been a characteristic and frequent organism, es­ A M A s 0 pecially in summer. When all measures for decrease Fig. 5. Growing season development (April-October) of of the nutrient load to the lake had been completed, Rhodomonas minuta and Rhodomonas lens in Lake another Rhodomonas species determined as R. lens Vattern.

Acta Phytogeogr. Suec. 68 166 Eva Wilten, Mats Oke Frank Gonzalez & rare in July. The April plankton in Lake Vattern Rhodomoras largely consists of small forms such as Rhodomonas, minuta other flagellates and tiny centric diatoms. During this 1965 period the total phosphorus concentration is com­ paratively high (10-15 f..Lg 1-1) but the water tem­ perature is below 2.5°C. In the beginning of April, 1970 when especially Rhodomonas minuta is well developed, the water temperature is just 0.5-1°C . As the temperature rises to more than 3°C in the middle 1975 of May the spring diatom bloom of Melosira has its maximum and the two Rhodomonas species decline. x 0.210 s 0.213 Diatoms are dominant until June and after their mineralization the nutrients are available for other Fig. 6. Development of Rhodomonas minuta in the Ekoln Bay of Lake Malaren based on mean biomasses (volumes) organisms-in Lake Vattern being Rhodomonas during the growing seasons 1965-1977. minuta, Uroglena americana and other flagellates in = mean value fo r a ten-years-period 1965-1974 0 (x) July. In August when temperature is around 16°C in (1965-1 974) )';=(normaliz. ed growmg. season values) x;-x the circulating surface watermass, Rhodomonas s( l965_1974) minuta and Rhodomonas lens dominate the algal .:if = growing season mean volume, s = standard deviation flora, although Rhodomonas lens is slightly more numerous in September when the water temperature has falien to 1 0°C. In the long-term investigation of the phyto­ plankton dynamics of the eutrophic Ekoln Bay in Lake Malaren the biomass of Rhodomonas minuta markedly increased when the phosphorus concentra­ tion in the water decreased (Fig. 6). The mean grow­ ing season biomass for the period 1965-1971 is 0. 105 mm3 l-1 while the corresponding value fo r the period 1972-1977 is 0.403 mm3 l-1• Unfortu­ nately, the development of Rhodomonas lens was not fo llowed here in the same way because of identifica­ tion problems. This organism has probably been determined as a small Cryptomonas sp. due to dif­ ficulties sometimes in distinguishing the pyrenoid. This species has not been recognized here (Fig. 7) un­ a til recently when a more accurate separation has been possible. The growing season development studied by week­ ly samplings, from May to October, indicate that Rhodomonas minuta has one maximum in May and another, much larger one, in June-July. Subsequent­ ly, the population decreases drastically but a small in­ crease is noticed at the end of August. Concerning Rhodomonas lens, one maximum appears at the end of May roughly coincident with that of R. minuta, and another in late June-early July. As also in Lake Vattern, Rhodomonas minuta occurs all through the growing season while Rhodomonas lens is absent in the late summer (August-September). This period coincides with a maximum of the blue-green algal bloom and a water temperature well in excess of 16°C. Both Rhodomonas species also have an early prevernal peak in the Ekoln Bay-unfortunately not Fig. 7. Rhodomonas minuta (a) and Rhodomonas lens (b) recorded in 1975 when the weekly investigation taken at magnification (oil immersion) after sedi­ IOOOx started after the ice-break. However, in February mentation in a settling chamber.

A eta Phytogeogr. Suec. 68 Rhodomonas minuta and R. lens -form-variation and ecology 167

1980, 1 cells of Rhodomonas minuta and 000 000 70 cells of Rhodomonas lens were counted just 000 under the ice where the temperature was about 0.5°C. Afterthe ice-break in the Ekoln Bay, generally in late April or early May, small centric diatoms develop in water temperatures of 4-8°C together with the two 5 Rhodomonas species. A drastic decrease of the

Rhodomonas populations occurs during the heavy Rhodomonas minuta spring bloom of Melosira in late May to early June. 4 The time between the spring bloom and the onset of a � blue-green algal bloom is again occupied by .._� Rhodomonas minuta and R. lens together with some 0 L.. 3 large Cryptomonas spp. The water temperatures dur­ C1J Ll ing this period may rise to ea 16°C. Nutrients in E ::J terms of nitrogen and phosphorus are available c

throughout the whole year but a drop in the PO 4 P­ 2 concentration occurs in June-July down to ea 5 !J.g l-1• During this period small easily movable flagellates

like Rhodomonas and Cryptomonas efficiently can Rhodomonas compete with large motionless algae. Owing to the 1•\ lens I I much higher nutrient status of the Ekoln Bay com­ I I I \ pared to Lake V attern, the populations of both /J ' \ Rhodomonas minuta and R. lens are very large here, \-' '•, with maximum numbers of 8 800 cells l-1 of 000 M A s 0 Rhodomonas minuta and 1 000 000 cells l-1 of Fig. 8. Growing season development (May-October) of Rhodomonas lens (Fig. 8). Rhodomonas minuta and Rhodomonas lens in the Ekoln Bay of Lake Malaren. Form and size variation A large form and size variation exists within both apex. Viewed from the side, the cells are curved. The species on almost every sampling occasion, il­ posterior end of the cell may be filled by a refractive lustrated by drawings from the Ekoln Bay (Figs. 9 body to a greater or lesser extent. This body is es­ and 10). The largest cells mainly develop in the begin­ pecially well-developed during the summer period ning of the growing season but a considerable spread from mid-June to the end of August (Fig. 9). A of the cell-length occurs which is found on most pyrenoid surrounded by starch plates is easily detect­ sampling occasions, here illustrated by mea­ able near the centre of the cell. surements of Rhodomonas minuta (Fig. 11). The The cells of Rhodomonas lens are pointed at both mean length and breadth measurements of Rhodo­ ends and a pyrenoid is visible in the interior, although monas minuta and Rhodomonas lens with maxi­ it is not as distinctive as that of R. minuta (Figs. 7 and mum-minimum values in the Ekoln Bay and Lake 1 0). In the material from Lakes Malaren and V attern Vattern have been tabulated (Table I). refractive basal bodies are not so evident. This The relation between length and breadth of the species is also a little curved when viewed from the Rhodomonas cells has been tested statistically on side. material from the Ekoln Bay where a correlation coefficient (r) of 0. 72 was attained for Rhodomonas minuta (Fig. 12) and a still weaker correlation of0.6 1 for Rhodomonas lens (Fig. 13). A relationship between length and breadth (urn) calculated from Discussion Figs. 12 and 13 gives an equation for Rhodomonas Lately, some authors have been dealing with iden­

minuta of y = 2.393 + 1.189 x and for Rhodomonas tification problems of Rhodomonas and a few have lens of y = 2.627 + 1.285 x where y is length and x tried to use the scanning electron microscope in order breadth. to get useful separation characters. The separation of The cell-form of Rhodomonas minuta these Rhodomonas from the genus Chroomonas has also in lakes is characterized by a fairly rounded apex and been an object of discussion where pigmentation and an oviform shape with a more or less pointed ant- periplast structure have been suggested as tools of

A eta Phytogeogr. Suec. 68 168 Eva Wil/en, Mats Oke Frank Gonzdlez &

Lake Malaren - Ekoln Bay May 11 May 18

May 25 June 1 June 8 0��� DGOQ� 0 ��@� June 15 June 22 June 29 � 0� I ����JPI ��® n: � �� ��r·O Q O

July July 13 July 20 July 27 6 1 I I

ee� ·. 9���Ei Aug . • 9 ��Aug.10 Aug. 17 Aug. 24 I �@1I a� 3 o. I

.. ·. �. ·- \ · . , n �. D!� · �(@ �\·@�\,r· / c::,o \ 1 v . . on) \ / .. . � � \J_/ V Q � � \ Y · �·/�� ,N�V ��� . . \V i Sept . 7 Sep\tr . 14 <\ Sept . 21 AJg. �1 I I I

Acta Phytogeogr. Suec. 68 Rhodomonas minuta and R. lens -form-variation and ecology 169

May 18 Lake Malaren - Ekoln Bay May 11 1

) I

May June 1 25 0 June 8 June 15 June 22 0®00 June 29 July & July 13 July 20

July 27 Sept. 28 Oct . Oct. 19 I I 5 I

_ 10 ..____._ __, J.lm Fig. 10. Form and size variation of Rhodomonas lens (12IK-preserved) in the Ekoln Bay of Lake Malaren, May-October (Drawing, Frank Gonzalez).

Fig. 9. Form and size variaton of Rhodomonas minuta (12IK-preserved) in the Ekoln Bay of Lake Malaren, May-September (Drawing, Frank Gonzalez).

Acta Phytogeogr. Suec. 68 1 70 Eva Willen, Mats Oke Frank Gonzdlez & differentiation (Butcher 1967 p. 7; Hickel 1975 p. ,urn 221; Javornicky 1976 p. 99). As a hexagonal or rec­ 14 tangular periplast plate shape has been revealed in both genera its taxonomical value needs a cautious 12 interpretation (Gantt 1971 p. 183; Hausmann & 10 Walz 1979 pp. 350-352; Munawar & Bistricki 1979 .r:...... en 8 pp. 248-25 1). c In this work, where all determinations were made � 6 on preserved material in a reversed light microscope, 4 the cells have been referred to Rhodomonas because of their shape, the occurrence of basal refractive 2 bodies, and to some extent because of their size. An 0 organism leading to confusion may in this respect es­ M A s 0 pecially be Chroomonas acuta Utermohl, the dividing Fig. 11. Minimum, maximum and mean length values of line to Cryptomonas being the occurrence of one 10 randomly measured cells of Rhodomonas minuta in single pyrenoid and also here the refractive bodies. weekly samples, May-October, from the Ekoln Bay of When Skuja described Rhodomonas minuta he Lake Malaren. (!lsoseparated a variety called nannoplanctica (Skuja 1948 pp. 346-347) . In his description the species minuta had a size of 9-14 ,urn x 4-7 ,urn and the each sampling occasion is, however, so large that no variety nannoplanctica 8-9 ,urn x 5-6 ,urn. He drew confidence can be given to the existence of a separate the variety with a more pointed antapical part than variety but instead it should be interpreted as a nor­ the main species. The validity of this character is, mal variation within one and the same species. The however, already rejected by Javornicky (1976 p. breadth of the cells does not vary as much as the 100) who considers the form of the antapical part in­ length, as also mentioned in the original descriptions fluenced bythe refractive bodies. Skuja recognized R. and here made clear by the poor relationship il­ minuta especially during the colder periods of the lustrated in Fig. 12. year while he fo und a frequent occurrence of v. nan­ The larger cells observed in the beginning of the noplanctica in summer. In studies of the occurrence growing season in both Lake Malaren and Lake and fo rm variation of Rhodomonas minuta in V attern are explained by a proneness of the cells to England, Lund (1962) also found large summer store photosynthetic products at low temperatures populations of what he called the variety nannoplanc­ when cell division rate is inhibited (Morgan & Kalff tica. Concerning the occurrence of Rhodomonas 1979 p. 132). The cell-size may be affected by large lacustris, R. minuta and v. nannoplanctica it is worth differences in carbon content, thus making the use of noting that Skuja fo und them all in one and the same narrow size-limits doubtful. lake. When large Rhodomonas minuta cells appear later There is good sense in Javornicky's recombination in the growing season they generally seem vacuolized of the above-mentioned species under the heading of and provided with a conspicuous basal refractive the species described first-Rhodomonas lacustris. In body (i.e. the large forms from June and July in Fig. Swedish algological work a fu sion of Rhodomonas 9). These cells are interpreted as entering a decaying minuta and R. lacustris has already been made in stage. practice but under the name of R. minuta. Similarly a When summing up the above discussion, it seems fu sion has been made in central Europe where the likely that Skuja's records of Rhodomonas lacustris, same organism is called Rhodomonas lacustris. In R. minuta and its variety nannoplanctica in the same the original description of R. lacustris a size of lake are only different size-forms of one and the same

10-13 ,urn x 5-8 ,urn was given, further illustrating alga. This combination under one species name is the lack of dividing lines between the two species also suggested by Javornicky (1976 p. but he lOO) when considering a large form and size variation. also separates the variety nannoplanctica, which The growing season development of Rhodomonas does not seem valid according to the results discussed minuta in the Swedish Lake Maiaren indicates that above (cf. also Kristiansen 1959 p. 22; Lund 1962 p. according to cell-length the variety nannoplanctica 135; Ettl 1980 p. 195). Javornicky also proposes that should be a valid name for the populations in early Rhodomonas lens is included as a Rhodomonas June, early July, almost all August and October (Fig. lacustris form, mainly due to lack of knowledge of 11). The rest is composed of mixed populations of R. self-existing populations. When occurring, Rhodo­ minuta and v. nannoplanctica. The size variation on monas lens seems to follow the development of R.

A eta Phytogeogr. Suec. 68 Rhodomonas minuta and R. lens -form-variation and ecology 171

)l13m Jlm18 12 17 1• 1• 11 16 1· 1• 2• 1• 15 1· 2· 2• 10 14 1• 1·5· 3· 9 3· 4· 13 1• 7· 3• •4 1• 8 1• 8· •1 1• .c 7 12 ...... 6 ·5 •1 C7l c 11 1• 6 2•9 3• • 1• ' . 2 � 10 (3·3· 2• 5 ..c ...... 4 C7l 3· 1• 1• c 9 3 � 1· 2 78 1• 6 5 0 0 2 3 4 6 7 4 5 8 )Jm 3 breadth Fig. 12. Relationship between length and breadth of 2 Rhodomonas minuta in the Ekoln Bay of Lake Malaren. The figures refer to numbers of organisms of the same size (n = 250). 0 0 2 3 5 6 7 10 11 4 8 9 }Jm minuta or R. lacustris, which also have been noticed in the Ekoln Bay of Lake Malaren as well as in Lake breadth Fig. 13. Relationship between length and breadth of V attern (Figs. 5 and 8). The cells here determined as R hodomonas lens in the Ekoln Bay of Lake Malaren. The Rhodomonas lens are, however, separated from R. figuresrefer to numbers of organisms of the same size (n = minuta by the pointed ends and the asymmetrical 105). form very similar to the depiction made by Ruttner (Pascher 1913 Figs. 159 and 160). The size of a Rhodomonas lens cell is not given in the original lens appeared after sewage reduction and a decrease description but Huber-Pestalozzi (1950 p. 21) gives of the phosphorus load-a measure which, however, the dimension 12-17 ,urn x 9-11 ,urn, thus indicating affected Rhodomonas minuta in the opposite way a somewhat larger cell-size than that of R. lacustris ­ (Fig. 4). minuta. Also in the measurements of these two algae As in many lakes Rhodomonas is an efficientcom­ in Lakes Vattern and Malaren Rhodomonas lens was petitor and occupies a large part of the total biomass, the largest. The mean length of this organism fitswell thus meaning a lot to the primary production, its into the above size range but the breadth is generally ecological significance cannot be overlooked. The smaller. There is, however, a large size variation on all sampling occasions with a considerable propor­ tion of the large cells occurring in the beginning of the Table I. Measurements of Rhodomonas mi nuta and R. lens . growing season (Fig. 10). Although Rhodomonas maximum minimum mean lens appears at the same time as R. minuta there is Ekol n Vattern Ekol n Vat tern Ekoln Vat tern some doubt about putting them together as fo rms un­ Rhodomonas mi nuta der the same species-name, the morphological reason 1 ength , u m 12.6 15.4 5.4 6.6 7.8 10.5 7.2 6.6 4 6 5.4 being the generally larger cells of R. lens and the breadth , um 3. 3 3. 3 0 different cell-shape. The experiences from Lake Rhodomonas 1 ens 1 17.1 17.6 7 7 8.5 12.2 V attern suggest that an ecological reason may also ength , um 0 13.3 breadth , um 10.8 8.8 5.5 5.1 7. 5 7.0 exist as a considerable population of Rhodomonas

Acta Phytogeogr. Suec. 68 1 72 Eva Wilten, Mats Oke Frank Gonzdlez & value of good and reliable species-limits within this Hobro, R. & Willen, E. 1977. Phytoplankton countings. genus is highly desirable, including a discussion on Intercalibration results and recommendations fo r fo rm and size variation. In the studies reported here routine work. Int. Revue ges. Hydrobiol. the emphasis is placed on demonstration of a clear 62(6):805-8 11. Huber-Pestalozzi, G. 1950. Das Phytoplankton des size and form variation during the year, a size Siisswassers. 3. - Binnengewasser 16(3): 1-310. difference between different ecosystems, and, finally, Javornicky, P. 1976. Minute species of the genus in demonstrating that changes in environmental con­ Rhodomonas Karsten (Cryptophyceae). - Arch. ditions favour Rhodomonas species to different ex­ Protistenk. 118:98-106. tents. These studies are but a minor contribution to Kristiansen, J. 1959. Flagellates from some Danish lakes the discussion of the whole complex. Further in­ and ponds. - Dansk bot. Ark. 18:9-55. vestigations with more delicate tools and using living Lewis, W. 1978. A compositional, phytogeographical and material are also necessary. elementary structural analysis of the phytoplankton in a tropical lake: Lake Lanao, Philippines. - J. Ecol. 66: 213-226. Remarks Lewis, W. & Weibezahn, F. 1976. Chemistry, energy flow, and community structure in some Venezuelan fresh This work has been supported by the National Swedish waters. - Arch. Hydrobiol. Suppl. 50:145-207. Environment Protection Board (Swedish abb. SNV) in its Lund, J.G.W. 1962. A rarely recorded but very common Limnological Survey Organization later reorganized as British alga, Rhodomonas minuta Skuja. - Br. phycol. Water Quality Laboratory of Uppsala. The weekly Bull. 2:133-139. samples from the Ekoln Bay in 1975 have kindly been put Morgan, K.C. & Kalff, J. 1979. Effect of light and at our disposal by the working group Research on temperature interactions on growth of Cryptomonas Recovery of Polluted Lakes, also sponsored by SNV. erosa (Cryptophyceae). - J. Phycol. 15:127-1 34. The responsibilities of the authors are as fo llows : Eva Munawar, M. & Bistricki, T. 1979. Scanning electron Willt�n-author and responsible for the project; Mats microscopy of some nannoplankton cryptomonads. - Oke performed the countings and measurements and Scanning electron micro se. 3:24 7-260. prepared the figures and photographs; Frank Gonzalez Munawar, M. & Munawar, J.F. 1975. The abundance and made all the drawings of the organisms. significance of phytoflagellates and nannoplankton in The results given above were presented and discussed in the St. Lawrence Great Lakes. 1. Phytoflagellates. - May 1980 at an international workshop in Seeon/Munich Verh. internaL Verein. theor. angew. Limnol. arranged by the International Association for Phyto­ 19:705-723. plankton Taxonomy and Ecology. Nauwerck, A. 1963. Die Beziehungen zwischen Correspondence should be addressed to: Zooplankton und Phytoplankton im See Erken. - Eva Willen, National Swedish Environment Protection Symb. bot. upsal. 17(5):1-163. Board, Water Quality Laboratory, Box 8043, S-750 08 1966. Beobachtungen iiber das Phytoplankton klarer Uppsala. Hochgebirgsseen. - Schweiz. Z. Hydro!. 28( 1):4-28. Olsen, P. & Willen, E. 1980. Phytoplankton response to REFERENCES sewage reduction in V attern, a large oligotrophic lake in central Sweden. - Arch. Hydrobiol. 89 :171-188. Ahlgren, G. 1970. Limnological studies of Lake Norr­ Pascher, A. 1913. Cryptomonadinae. - Siisswass.-Flora viken, a eutrophicated Swedish lake. 2. Phytoplankton Dtl. Ost. Schweiz 2:96-1 14. and its production. - Schweiz. Z. Hydro!. 32(2): Pavoni, M. 1963. Die Bedeutung des Nannoplanktons in 353-396. Vergleich zum Netzplankton. Qualitative und quan­ Butcher, R.W. 1967. An introductory account of the titative Untersuchungen im Ziirichsee, Pfaffikersee und smaller algae of British coastal waters. 4. - Fishery In­ anderen Seen. - Schweiz. Z. Hydro!. 25(1):219-34 1. vest. Lond. 4(4):1-54. Pollingher, U. 1978. Algae fo und in the plankton of Lake Ettl, H. 1980. Beitrag zur Kenntnis der Siisswasseralgen Kinneret. Monographiae bioi. 32:236-240. Danemarks. - Bot. Tidsskr. 74:179-223. Skuja, H. 1948. Taxonomie des phytoplanktons einiger Gantt, E. 1971. Micromorphology of the periplast of Seen in Uppland, Schweden. - Symb. bot. upsal. 9(3): Chroomonas sp. (Cryptophyceae). - J. Phycol. 1-399. 7:177-1 84. Utermohl, H. 1958. Zur Vervollkommnung der quan­ Hausmann, K. & Walz, B. 1979. Periplaststruktur und titativen Phytoplankton-Methodik. - Mitt. int. Verein. Organisation der Plasmamembran von Rhodomonas theor. angew. Limnol. 9:1-38. spec. (Cryptophyceae). - Protoplasma 101:349-354. Willen, E. 1976. A simplified method of phytoplankton Hickel, B. 1975. The application of the scanning electron counting. - Br. phycol. J. 11:265-278. microscope to freshwater phytoplankton taxonomy Willen, T. 1959. The phytoplankton of Gorvaln, a bay of and morphology. - Arch. Hydrobiol. 76:21 8-228. Lake Malaren. - Oikos 10(2):24 1-274.

Acta Phytogeogr. Suec. 68 Phytoplankton fr om Lakes and Ponds on V estspitsbergen Torbjorn Willen

In 1911 Barge published a survey of earlier freshwater 10 % acetic acid. The samples were sedimented ( 1, phycological records from Spitsbergen. During recent 10, 25 and 50 ml) according to the Utermohl tech­ decades plankton collections have been made by several nique and counted by means of an inverted plankton scientific expeditions. Some papers have been published microscope (Utermohl 1958). Volume values were containing plankton lists from different waters, e.g., by calculated, some of them being published by Amren Thomasson (1958, 1961). Foged has published a mono­ (1964b, c) in discussing the relationship between graph of diatoms from Spitsbergen. The papers contain qualitative lists of phytoplankton but no quantitative zooplankton and small monads and some other analyses seem to have been made. Some aspects of conser­ phytoplankton organisms. In this paper all total vation and research on Spitsbergen ("Man and the biomass values are given in Tables I-III. Biosphere, MAB") have recently been discussed by Gj rerevoll (1975). In the summers of 1959 and 1962 Dr. H. Amren, Upp­ sala, made scientificexpediti ons to Vestspitsbergen, main­ ly to study the ecology and taxonomy of zooplankton (Amren 1964a, b, c). In 1962 phytoplankton samples were also collected from some waters. This material, totally 144 samples, was analysed as to species composition and quantities of predominating taxa by the author in 1963 and 1964.

Material and methods The material was collected from different types of waters, all situated on the western coast of Vestspitsbergen near the Isfjord Radio Station. The investigated area is shown on maps (Figs. 1-3) slightly modified from Amren (1964c). A description of the localities is given by Amren ( op. cit., pp. 213-223). In this paper the different types of waters have been referred to three groups: 1. lakes, 2. ponds, and 3. localities influenced by sea water. However, the borderline, especially with regard to the last group, is sometimes vague. Two of the ponds (Loc. 23 and 68) may be characterized as puddles according to R0en (1962). Some of the ponds in the area investigated are regarded as extreme arctic ones (Amren 1964c p. 223); from two of them (Loc. 60 and 83) phytoplankton samples were collected. The phytoplankton samples were taken during the period 25.VI-13.VIII.1 962. All samples were stored in 100 ml glass bottles. They were immediately Fig. 1. Map of Vestspitsbergen. The main area of in­ preserved with Lugol's solution supplemented with vestigation is marked by a rectangle (from Amren 1964c).

Acta Phytogeogr. Suec. 68 174 To rbjorn Wil!en

Fig. 2. The main area of investiga­ tion. Magnified part of Fig. 1 (from Amren 1964c).

11

Fig. 3. Part of the area of investigation. Magnified part of Fig. 2 (from Amren 1964c).

Acta Phytogeogr. Suec. 68 Phytoplankton fr om lakes and ponds on Vestspitsbergen 175

Physico-chemical data single trichomes 1· ' on 1l.VII and on 16.VII it had dis­ appeared. In Loc. 6 ea. 600 000 trichomes 1-1 were counted All analytical data included in Tables I-III have on 3.VII. kindly been put at my disposal by Dr. Amren. The values of total nitrogen and total phosphorus were CRYPTOPHYCEAE not published by Amren ( 1964c). As to methods, see Cryptomonas spp. - Loc. 1, 2, 12, 21, 32, 59, 77, 84. It Amren (op. cit., p. 212). was remarkable that this genus played such an unimport­ It may be mentioned that the ice-free period was ant role both as to number of species and of individuals in 2-3 months and that the water temperature normally all investigated waters. varied between 5 and 9°C with a maximum value of Cyathomonas truncata (Fres.) From. - Loc. 4, 21, 6 7. 12.2°C recorded. Rhodomonas minuta Skuja - Loc. 1, 4, 8, 18-22, 32, 43, 51, 59, 60, 62, 70, 77, 78, 84, 85. Highest values: ea. Some localities situated near the coast either were 250 000 cells in Loc. 78 on 31.VII, ea. 2 400 000 cells connected to the sea or were exposed to salt spray by t' 1- 1 in Loc. 22 on 29.VII, but, as a rule, considerably less the winds. They showed relatively high chloride con­ frequent. centrations (30-540 mg often with considerable r1), variation during the year. DINOPHYCEAE The values of conductivity in this paper are given Amphidinium spp. - Loc. 4, 5, 70. as mS m- 1 (20°C) (which means, fo r example, that Glenodinium pulvisculus (Ehr.) Stein - Loc. 1, 3, 4, I2, 26, 67. K20 • 106 250 is expressed as 25.0 mS m -1). Gymnodinium spp. - Loc. 1, 2, 6, 22, 77, 78, 98, etc. Gyrodinium pascheri (Suchlandt) Schiller - In most List of species localities in moderate quantities. Peridinium spp. - Loc. 4, 14, 18, 77, 79. BACTERIOPHYTA Chromatium cf. minus Winogr. - Loc. 85. 28.VI at 40 m (5 million cells 1-1) and at 45 m (3.6 million cells 1 - 1) CHRYSOPHYCEAE together with fragments of diatoms (Diatoma and Bicosoeca cylindrica (Lackey) Bourr. - Loc. 1, 6. Fig. 4a. Synedra). Length 19-28 ,urn, breadth 6.50- 10 ,urn, protoplast 3.8-4.5 x 7.6 ,urn.The opening of the lorica is dilated to a CYANOPHYCEAE diameter of 11.4-13.5 ,urn. The species thus resembles B. Anabaena cf. aequalis Borge - Loc. 15. ainikkiae Jarnefelt, lacking, however, any visible concen­ A. planctonica Brunnth. - Loc. 8, 10, 67. tric rings. Specimens with slightly dilated openings are A. spp. - Loc. 1, 2, 4, 16, 67, 69, 98. earlier depicted (Willen 1963) and a similar specimen is Aphanothece clathrata West & West - Loc. 4, 20, 67. also drawn by Lackey (1942 p. 198, Fig. 2a) under the Chroococcus limneticus Lemm. - Loc. 1, 2, 41, 46, 53. name of Domatomonas sp. C. minutus (Ki.itz.) Nag. - Loc. 2, 4, 18, 2I, 32. Bitrichia ollula Fott - Loc. 77, 78. C. turgidus (Ki.itz.) Nag. - Loc. 41. Cercobodo cf. varians Skuja - Loc. 6, 22. Gomphosphaeria lacustris Chod. var. compacta Lemm. - Chrysococcus minutus (Fritsch) Nyg. - Loc. 84. Loc. 2, 7, 8, 15, 20, 41, 67. In Loc. 4I ea. 25 000 col. Chrysoikos angulatus Willen - Loc. I, 3-6, 8, 12, 18, 21, I - 1 were counted on 5.VIII; otherwise only single specimens 22, 24, 25, 68, 69, 84, 85. Fig. 4b. The species seems to be were observed. frequently occurring in the investigated area, never Lyngbya limnetica Lemm. - Loc. 21, 26. recorded, however, in large quantities. L. vacuolifera Skuja - Loc. 4, 21. C. bicornis Willen - Loc. 1, 8, 21, 22. Fig. 4c. The species Microcystis incerta (Lemm.) Lemm. - Loc. 4, 7, 20, 41, appeared together with C. angulatus in some localities, but 51, 53. was only recorded in single specimens. M. pulverea (Wood) Forti - Loc. 7. 21. The genus Chrysoikos with two new species from Oscillatoria limnetica Lemm. - Loc. 20, 26, 67. Very Spitsbergen and Bjornoya was described by the author abundant, especially in Loc. 20 where ea. 1700 cm 1 - 1 (1967). Nauwerck (1979) revised the genus Chrysolykos were counted on I.VII, ea. 1 150 cm 1-1 on 4.VII and ea. 70 and included Chrysoikos in this genus. The two species cm 1- 1 on I LVII. mentioned above are now named Chrysolykos angulatus 0. splendida Grev. - Loc. 21, 67. (Willen) Nauwerck and Chrysolykos angulatus f. bicornis 0. spp. - Loc. 2, 7, 11, 22, 26, 69. (Willen) Nauwerck, respectively. Pseudanabaena catenata Laut. - Loc. 1, 4, 7, 18, 21, 51, Chrysolykos planctonicus Mack - Loc. 77, 78. Only in 67. small quantities in the lakes. Synechococcus leopoliensis (Racib.) Kom. - Loc. 1, 6, 8, C. skujae (Nauwerck) Bourr. - Loc. 4, 67, 70, 78, 84, 85. 22. The species occurred with solitary cells (ea. 1 ,urn x Dinobryon cylindricum Imhof - Loc. 6, 41. Single 3-5 ,urn), in short trichomes or usually in short spiral­ specimens in Loc. 41. In Loc. 6 650 000, 260 000 and coiled trichomes. It was very frequent from late June but 380 000 cells t1 were counted on 12.VII, 23.VII and had quite disappeared in mid-July. In Loc. 1, fo r example, 3.VIII, respective! y. 800 000 short trichomes t1 were counted on 30.VI, only D. divergens Imhof - Loc. 6, 41.

Acta Phytogeogr. Suec. 68 176 To rbjiirn Wi/Len

./

.--···

. . b '�. ..

...... ::; .. . '··.��·� a · ...�",

Fig. 4. (a) Bicosoeca cy lindrica, (b) Chrysoikos angulatus (Chrysolykos angulatus), (c) Chrysoikos bicornis (Chrysolykos angulatus f. bicornis), (d) and (e) Ankyra judayi with a phycomycete, (f)-(h) Golenkiniopsis parvula, (i) Oocystis parva. Scale = 10 ,urn (a-e), 5 ,urn (f-i).

D. divergens var. schauinslandii (Lemm.) Brunnth. - Loc. 77-79, 85. One of the most frequent plankton algae in the 12, 41. In Loc. 41 ea. 7 million cells per litre were counted investigated area. It was, however, never observed in of D. divergens and its variety on 13.VII and 5.VIII. ponds influenced by sea water. D. sociale Ehrnb. var. americanum (Brunnth.) Bachm. - Epipyxis condensata (Mack) Hilliard & Asmund - Loc. Loc. 2-6, 8, 12, 14, 16, 18, 40, 42, 44, 53, 59, 67, 70, 12, 15, 18, 67. The specimens measured 8-10 25-27.5 x

A eta Phytogeogr. Suec. 68 Phytoplankton fr om lakes and ponds on Vestspitsbergen 177

,urn, thus having somewhat larger dimensions than those at 33 m. Especially large numbers of the species were given by Hilliard & Asmund (1963). The species was to be counted in the ponds influenced by sea water. In some found planktonic. localities the seasonal distribution of the species could be E. polymorpha (Lund) Hilliard & Asmund - Loc. 3, 7, 10, followed, fo r example in Loc. 1. 19, 70, 77. D. vulgare Bory var. linearis Grun. - Loc. 20 22, 27, 32, E. tubulosa (Mack) Hilliard & Asmund - Loc. 18. 85, etc. Dominant in Loc. 20, where more than 13 million

E. spp. Loc. 4, 6, 49, 69, cells 1-1 were recorded on 1.VII. Kephyrion spp. - Loc. 1, 12. Eunotia praerupta Ehr. - Loc. 12. Mallomonas spp. - Loc. 3, 12, 14, 21, 67. Fragilaria capucina Desmaz. - Loc. 14. Monas spp. - In most localities. F. construens (Ehr.) Grun. - Loc. 4: var. venter (Ehr.) Monochrysis aphanaster Skuja - Loc. 1, 4, 70. Grun., 12. M. cf. parva Skuja - Loc. 22. This small monad, ea. 2-3 x Hantzschia amphioxys (Ehr.) Grun. - Loc. 27. 3.5 ,urn, corresponds closest to M. parva described by Meridian circulare Ag. - Loc. 83. Skuja (1956 p. 262); all details, however, were difficult to Navicula amphibola Cleve - Loc. 32. distinguish. It was very abundant in Loc. 22 on 29.VII and N. tuscula (Ehr) Grun. - Loc. 2. 13.VIII: ea. 190 million cells per litre. S tauroneis anceps Ehr. - Lac. 6 7. Ochromonas spp. - Loc. 22, etc. S. phoenicenteron Ehr. - Loc. 4, 21. Pleuromonas sp. - Loc. 18, etc. Stephanodiscus hantzschii Grun. var. pusillus Grun. Pseudokephyrion conicum (Schiller) Schmid. - Loc. 3, 4. Loc. 84, 85. Only in moderate quantities, 20 000-70 000 P. entzii Conrad - Loc. 12, 18, 21, 44. cells r1• Stylopyxis libera Fott - Loc. 21. Synedra acus Ki.itz. - Loc. 1, 84, etc. Synura sp. - Loc. 16, 21, 41, 67. Only observed in single S. ulna (Nitzsch) Ehr. - Loc. 1, 4, 79, 84. specimens. S. spp. - Loc. 2, 7, 12, 67, 69, 78, 79, 85, etc. Uroglena americana Calkins - Observed in most localities Tabellaria jlocculosa (Roth) Ki.itz. - Loc. 18, etc. forming the major part of the total phytoplankton biomass, especially in small ponds. XANTHOPHYC EAE U. volvox Ehr. - Loc. 18, 85, etc. Mischococcus confervicola Nag. - Loc. 21, 28. Fig. 5a. Volvochrysis globosa Schiller - Loc. 4, 7. Globose cells, 3.8-4.2 ,urn, at the ends of branched gelatinous stalks, breadth ea. 5 ,urn.The filaments normally HAPTOPHYCEAE attached to aquatic plants or filamentous algae. Chrysochromulina parva Lackey - Loc. 5, 59, 77, 78, 84. Mainly occurring in lakes, but never observed in great EUGLENOPHYCEAE quantities : maximum value from Loc. 77 on 3l.VII, 0 m, Euglena gracilis Klebs - Loc. 84. with ea. 290 000 cells 1-1• E. spp. - Loc. 28, 84. Petalomonas sp. - Loc. 21. CRASPEDOPHYC EAE Strombomonas acuminata (Schmarda) Defl. - Loc. 67. Aulomonas purdyi Lackey - Loc. 79. The anterior portion Trachelomonas hispida (Perty) Stein - Loc. 2, 16, 32. of the tube was either enlarged or had the same diameter as the tube (cf. Willen 1963 p. 52). PRASINOPHYCEAE Desmarella moniliformis Kent - Loc. 24. Scourfieldia complanata G. S. West - Loc. 21. Lagenoeca globulosa France - Loc. 4, 5, 8, 79. Te tramitus pyriformis Klebs - Loc. 40. Trichloris paradoxa Scherffel & Pascher - Loc. 21. BACILLARIOPHYC EAE Cyclotella antiqua W. Sm. - Loc. 8, 18, 19. CHLOROPHYCEAE C. comta (Ehr.) Ki.itz. - Loc. 18, etc. Ankistrodesmusfalcatus (Corda) Ralfs - Loc. 1, 4, 6, 7, C. kuetzingiana Thwait. - Loc. 84. 18, 20, 21, 24, 25, 67, 69, etc. C. kuetzingiana var. planetophora Fricke - Loc. 84. Ankyra judayi (G. M. Smith) Fott - Loc. 20, 22, 25-27, Diatoma elongatum (Lyngb.) Ag. - Loc. 1-3, 5-8, 10, 51. Abundant: Loc. 20 ea. 380 000 cells 1-1 on 13.VIII; 1 12, 14, 16, 19-23, 25, 26, 32, 53, 69, 70, 83-85. This Lac. 22 ea 900 000 cells t on 13. VIII; Lac. 2 7 ea. species and the following D. vulgare var. linearis were the 450 000 cells 1-1 on 13.VIII. The species also occurred dur­ most frequently occurring diatoms in the investigated ing July but maximum values were recorded in samples waters. In Kongressvatnet (Loc. 85) the number of D. from August. Several specimens were attacked by a elongatum increased considerably at the levels of 12-30 m phycomycete (Fig. 4d, e), with the dimensions 1.5-2.0 x on 28.VI (ea. 300 000 cells l-1, 30 m), while only fragments 5.5-6.0 ,urn. Abundant in Loc. 20 and 26. were observed at the levels of 40 and 45 m, where the Botryococcus braunii Ki.itz. - Loc. 7, 21, 53. water was free of oxygen. On 9.VIII there was a rather Carteria spp. - In most localities but always in moderate even distribution from the surface down to 7 m with ea. quantities. 400 000 cells l-1• At the levels of 12-30 m the number in­ Chlamydocapsa bacillus (Teiling) Fott. - Loc. 16. creased to ea 900 000 cells l-1• No plankton samples were C. planctonica (West & West) Fott - Loc. 53. taken below the level of 30 m; the oxygen concentration, Chlamydomonas spp. - In most localities. A small species however, decreased from 10.7 mg at 30 m to 1.8 mg rt (diam. 4.8 ,urn) was very abundant in localities influenced 1- l Acta Phytogeogr. Suec. 68 !::: ) ·::::: : .... · . . · · ... . ·

· . .. .::· :·. .. :- · . .. : · : . · . .. · ...... · ... .

, . � · .·. .. · · . · (€; . . :::: ::: ::::.· ·.- ·' ({E) :: :'·.

. . · ·. .. ·

b

. · .. ··

. · · ::�� : · ,••"" :·. . ·

· . · .· . · . . · \�: . ... · _ (:: .)

a

. .· · }Jm · · ·-····· Fig. 5 (a) Mischococcus confervicola, (b) Protoderma viride, different stages. by sea water. and cup-shaped protoplast measured 2.9-4.8 ,urn and the Chlorogonium cf. peiforatum Skuja - Loc. 4. The species spines 5-7 ,urn. According to Korshikow (1953) the coincides with that described by Skuja (1956 p. 147) but diameter is 6.4 ,urn and the spines 4- 10 ,urn. the dimensions are somewhat larger. Conium sociale (Duj.) Warming - Loc. 24. Coelastrum cambricum Archer - Loc. 41. Dimensions of Lagerheimia subsalsa Lemm. - Loc. 2, 4. the cells ea. 6.6-9.5 ,urn with broad projections connecting Monomastix cf. pyrenigera Skuja - Loc. 4, 25. the cells. Abundant with ea. 40 000 coen. t1 on 5.VIII. Monoraphidium contortum (Thur.) Kom.-Legn. - Loc. C. microporum Nag. - Loc. 21, 41. 22, etc. Dictyosphaerium primarium Skuja - Loc. 4, 22, 25, 85. M. convolutum (Corda) Kom.-Legn. - Loc. 4, 21. The species was abundant especially in Loc. 85, where ea. M. griffithii (Berkel) Kom.-Legn. - Loc. 1, 24, 46, 79, 84, 200 000 cells t1 were recorded on 9.VIII (0 and 1 m). etc. D. pulchellum Wood - Loc. 7. M. minutum (Nag.) Kom.-Legn. - Loc. 21, 22. Dimorphococcus lunatus A. Br. - Loc. 51. Oocystis parva West & West - Loc. 6, 67, 84, 85. Fig. 4i. Diplostauron sp. - Loc. 21. Cell ovoid, tapering toward The specimens, 2.9-5.0 x 4.7-1 1.5 ,urn,occu rred singly as the top, 5.5-8.4 x (6.6-)8.4 ,urn,in an envelope with slight a rule, or very seldom with 2 cells in envelope. The projecting and rounded angles, 17.0- 19.6 x 19.6-22.5 cylindrical-ovoid cells have rounded poles only showing a ,urn ; the two flag ell a ea. 22 ,urn.Nu cleus in the upper part of very slight thickening. The species was rather abundant in the cell, pyrenoid basal. Envelopes with 4 daughter cells, the lakes (Loc. 84 and 85) with ea. 50 000 cells 1- 1 • ea. 4.2 ,urn, observed. The species does not correspond to 0. solitaria Wittr. - Loc. 2, 4, 7, 8, 16, 19, 21, 25, 41. taxa described earlier. The small preserved material does Paulschulzia pseudovolvox (Schulz) Skuja - Loc. 21. not permit a final evaluation. Pediastrum boryanum (Turp.) Menegh. - Loc. 4, 7, 19, Eudorina elegans Ehr. - Loc. 24. Ca. 14 000 col. 1- 1 were 22, 25, 41 (in Loc. 22 and 25 var. granulatum (Kiitz.) Al. counted on 11. VII. Braun). Single specimens except in Loc. 41, where ea. Franceia droescheri (Lemm.) Smith - Loc. 7, 10. 15 000 coenobia 1- 1 were recorded on 5.VIII. The coenobia Geminella interrupta Turp. - Loc. 26. were often abnormal in shape. Golenkiniopsis parvula (Woronich.) Korsch. - Loc. 20, P. duplex Meyen - Loc. 19, 41. 25. Fig. 4f-h. The spherical cells with normal pyrenoid P. integrum Nag. f. granulatum Racib. - Loc. 7.

A eta Phytogeogr. S1.4ec. 68 Phytoplankton fr om lakes and ponds on Vestspitsbergen 179

P. muticum Kiitz. - Loc. 7, 51. The low conductivity value, 7.0 mS m-1, is explained P. braunii Wartm. - Loc. 41. by influence of melt water. From Loc. 59 only a Protoderma viride Kiitz. - Loc. 51. Fig. 5b. The cushion­ limited number of taxa was recorded: Dinobryon like thallus, 40-65 in diameter, consists of short ,urn sociale var. americanum, Chrysochromulina parva, filaments ; cell dimension 7.5-10.0 7.5-12.5 x ,urn. Rhodomonas minuta, Cryptomonas spp., etc. The Presumably loosens from aquatic plants, etc. sample from Loc. 62 was very rich in detritus and Scenedesmus acutus Meyen - Loc. 22. S. ecornis (Ralfs) Chod. - Loc. 22. only some rhodomonads and small flagellates were S. quadricauda (Turp.) Breb. - Loc. 22, 32, 41. observed. S. spinosus Chod. - Loc. 22. Loc. and (Fig. 3). These lakes, situated Stichococcus bacillaris Nag. - Loc. 98, etc. 77, 78 79 Te trae'dron caudatum (Corda) Hansg. - Loc. 22. N .E. of Lake Linnevatnet, are all small and T. caudatum var. incisum Lagerh. - Loc. 22. moderately deep lakes (5.5-7.0 m). "The transparen­ T. minimum (A. Br.) Hansg. - Loc. 4. cy in Loc. 77 and 78 exceeds the depth of the two Ulothrix sp. - A species, diam. 2-4 ,urn, occurred in localities", and from Loc. 79 a value of 3.5 m is given several samples. (Amren 1964c p. 219). From Loc. 77 Dinobryon Willea irregularis Wille - Loc. 6, 7, 53. sociale var. americanum was recorded in 500 000-650 cells t1 on July 6, 27 and 31. CHLOROPHYC EAE/ZYGNEMATALES 000 Closterium parvulum Nag. - Loc. 21, 41. Chrysochromulinaparva was relatively frequent and Cosmarium spp. - In several samples. among other taxa the following may be mentioned: Staurodesmus cuspidatus (Breb.) Teiling - Loc. 14, 67. Bitrichia ollula (with rhizopodial processes), Chry­ S. dejectus (Breb.) Teiling - Loc. 41, 67. solykos planctonicus, skujae, Ep ipyxis poly­ C. morpha, Uroglena americana, Rhodomonas minuta, MYCOPHYTA Gyrodinium pascheri, etc. To the list of phytoplankton two hyphomycetes may be In Loc. 78 Dinobryon sociale var. americanum added, both recorded in single specimens, viz., predominated and was counted in ea. 2 million cells Clavatospora longibrachiata (Ingold) S. Nilsson - Loc. 1- 1 on 31. VII. 1 962. Moreover, about the same taxa 15. occurred in this lake as in Loc. 77. Tricladium angulatum Ingold - Loc. 40. Dinobryonsociale var. americanum was also com­ mon in Loc. 79 but in moderate quantities (max. ea. 150 000 cells l-1). In this lake Lagenoeca globulosa Quantitative studies occurred, and Aulomonas purdyi was only recorded from this locality occurring at various depths. Lakes Material from seven lakes has been examined. From Lake Linnevatnet, Loc. 84 (Fig. 2). The lake is one of one of them, Loc. 62, only one sample was collected; the largest on Spitsbergen with a maximum depth of from the others samplin.gwas done two or three times 35 m. On 25.VI. 1962 the lake was still ice-covered, during the period of investigation. Most of the lakes transparency 8.0 m; on 1l.VIII the ice was broken, are shallow. Lake Linnevatnet (Loc. 84) and Lake the circulation was complete and the transparency Kongressvatnet (Loc. 85), however, are 35 and 46 m only 2.2 m due to the occurrence of suspended deep, respectively. The analytical data are shown in material from the glaciers. Table I. Relatively low water temperatures were Thomasson ( 1961) gives a list of phytoplankton measured in all these waters (maximum value in Loc. 79, 7.4°C). The chloride content was low. As to con­ ductivity, oxygen concentration and content of • nitrogen and phosphorus as well as total biomass of phytoplankton marked stratifications were recorded 10 • in the deeper ·lakes. In general, the biomass of • • • I phytoplankton was very low in all lakes (Fig. 6) and • • • • real mass development of any species was never • • 01 observed. • • Loc. and (Fig. 2). Loc. 59 (Amren 1964c p. 59 62 219), situated 56 m above sea level, was free from ice 001 _L._--,------:------' on July 18, 1962. A Secchi disc transparency value of Fig. 6. Total phytoplankton biomass, mm3 1- 1 , in the in­ 7.5 m is given. Loc. 62 is situated in the same region. vestigated lakes.

Acta Phytogeogr. Suec. 68 180 To rbjorn Willen

Table I. Analyti ca l data from investigated lakes .

Max. . Samp1 e Temp • Conducti.vi ty Tota1 N Total P Phytopl ankton Loc . Date - - depth from ° -1 -1 1 3 l No . 1962 m m °C 20 , mS m pg 1 pg 1- biomass , 11111 1-

59 18. VII 8.0 0 6.9 10 12.0 0.17

l.VIII 0 6.4 12 12.0 0.13

3 6.3 12 12.3 186 2 0.09 8 5.4 13 12.6 0.11

62 1. VIII 3.0 0 5.3 . 7 3 228 4 0.06

77 6. VII 5.5 0 4.5 12 12.6 0.22

27. VII 0 6.4 32 0.34

31 .VII 0 5.7 32 12.3 0.37

5 5.7 32 12.5 0.44 78 27.VII 7. 0 0 6.1 24 0.42 " 31 .VII 0 6.7 24 12.2 0.83

7 6.5 24 12.4 1.27 79 27.VII 7.0 0 6.5 11 0.12

31 .VII 0 7.4 12 12.1 0.19

3 7.3 12 12.0 0.14

7 6.8 12 12.1 0.14

84 25.VI . 35.0 0 0.4 22 14.2 233 7 0.02

3 0.4 28 15.5 5 21 3 6 0.01 12 0.3 27 14.4 0.02

20 0.3 28 14.3 5 200 3 0.01

27 0.4 31 14.2 0.01

11.VII 0 3.3 24 13.2 0,04 1 331 2 0.06

2 . 0.04 3 0.05

4 245 0.05 7 0.04

12 3.3 24 13.5 5 286 2 0.04 20 245 2 0.12

33 3.3 24 13.7 5 180 16 0.04 85 28 . VI >46.0 0 0.4 25 13.0 0.26

1 188 4 0.26 2 0.21

3 0.5 81 13.6 0.11

4 0.05

0.5 87 12.2 5 . 138 0.08 12 0.5 89 11.4 0.12

20 165 0.15

30 1.1 6.0 5 252 2 0.17 40 3000 111 (0.10)

4.5 1.9 176 0.0 6 2920 106 (0.07) . 9. VIII 0 3.7 67 12.6 0.40

1 261 2 0.46 2 0.44

4 3.7 64 0.62

7 3.7 67 12.8 4 182 0.29 12 2.0 83 12.7 0.48

20 1.2 89 12.4 5 200 2 0.50

30 1.0 93 10.7 5 201 2 0.40

Acta Phytogeogr. Suec. 68 Phytoplankton fr om lakes and ponds on Vestspitsbergen 181

(net plankton) comprising 12 taxa from Lake Peridinium, Chrysolykos planctonicus, C. skujae, Linnevatnet. Dinobryon sociale var. americanum, Uroglena, In the samples from June 1962 only a limited Diatoma and small centric diatoms. All taxa were number of taxa were recorded: Rhodomonas minuta, counted in small quantities. Chrysochromulina parva, Rhizochrysis sp., Ste­ The total biomass of phytoplankton was low in phanodiscus hantzschii var. pusillus, Chlamydo­ Lake Kongressvatnet but somewhat higher than cor­ monas sp., Oocystis parva etc. In August about the responding values from Lake Linnevatnet. In June same species occurred but in somewhat larger quan­ about the same values (0.21-0.26 mm3 1-1) were tities. Gymnodinium sp., Chrysolykos skujae, recorded under the ice down to 3 m. At the levels Euglena sp. and Ankistrodesmus fa lcatus were also between 4 and 12 m lower values were obtained. At recorded. 12 m, however, Synedra spp. developed in greater The total biomass of phytoplankton was very low quantities, increasing down to the level of 30 m where in June and only slightly higher in August. The these taxa were counted in ea. 300 000 cells 1-1• At highest value was obtained at 20 m LVIII), mainly this level Oocystis and small flagellates also were (I caused by the occurrence of Rhodomonas minuta recorded. At 40 and 45 m, as mentioned above, only and Synedra spp. bacteriophytes developed causing relatively high biomass values. In August nearly the same values Lake Kongressvatnet, Loc. 85 (Fig. 2). The lake is the (about 0.4-0.5 mm3 1-1) were obtained from the sur­ firstexam ple of a meromictic lake (probably a case of face down to 30 m, with predominance of biogenic meromixis) from Spitsbergen (Amren 1964c Gyrodinium, small flagellates and Diatoma pp. 221 -222). The lake is deep, probably more than elongatum. 46 m (measured by Amren at a sampling site), and it has a sheltered position. Beyum & Kjensmo (1970) give a depth value of 52 m. At the sampling dates in Ponds 1962 transparency values of 11.5 and 6.0 m, respec­ Most samples-about 75-were collected in small tively, were recorded. On 28.VI the lake was ice­ ponds or shallow waters with depths never exceeding covered. As to details of the physico-chemical 2 m. All data are given in Table II and in Figs. 7-9. In stratification, reference should be made to Amren some of these ponds it was possible to follow the (1964c Figs. 4, 5), to Table I in this paper, and to seasonal development from the end of June to the Beyum & Kjensmo (1970). An oxygen deficit was middle of August (Loc. 1-6 and 12). Rather high recorded at depths below 30-35 m at all sampling water temperatures, 1 0- 1 2°C, were observed in dates and the lake seems to have a permanent stag­ some ponds in early July and in August. With one ex­ nant bottom water. ception (Loc. 83), the conductivity varied between 6 The two vertical sampling series from this lake, and 26 mS m- 1 ; most of them showed values between 26.VI and 9.VIII.1962, comprise 11 and 8 samples, 11 and 17 mS m-1 (Fig. 7). The chloride concentra­ respectively. tion varied between 3 and 16 mg t 1; all waters, to On 28.VI only a limited number of taxa were recorded: Rhodomonas minuta, Glenodinium sp., Gyrodinium pascheri, Chrysoikos angulatus, Sy­ 20 nedra spp., Dictyosphaerium primarium, Oocystis parva, etc. Very small flagellates were counted in ea. 1 5 1.2 million cells 1-1 at 1.0 m and in 200 000-400000 VI \:) c 0 cells 1-1 in the samples from levels between 2 and 30 Cl. m. The samples from 40 and 45 m, however, only 0 10 contained fragments of diatoms (Synedra spp.) but .0� held immense populations of bacteriophytes, mainly :JE z 5 Chromatium cf. minus-the only organisms living in the upper oxygen-free stratum of the lake. The phytoplankton biomass values from these levels have (7) (7) (7) (7) (7) (7) (7) (7) (7) (7) (7) 0 been put in brackets in Table I. They represent only

Acta Phytogeogr. Suec. 68 To rbjom Willen 182

Table 11. Analytical data from _ ponds (samples from a depth of 0.3 m or less) .

Loc . Date Max. Conducti vity Cl Total-N Total·P Phytop 1 ank ton depth No . 1962 ° l l l l 3 l m 20 , mS m- m9 1 - 119 1 - 119 l · biomass, mm 1 -

30. VI 1.1 1.4 13.2 1.41

4. VII 5. 7 12.6 420 14 3.58

8. VII 7.6 12 12.9 5.65

ll.VJI 7.0 ,. 12 2.14

16.VII 7.8 13 12.3 1.80

29. . VII 6. 7 14 11.9 264 0.80

12. VIII 9.2 15 0.17

2. VII 0.9 7.1 10 0.44

4. VII 12.0 11 11.4 503 0.60

11.VII 7.4 13 0.83 ' 16. VII 8.0 1 3 12.0 0. 70

29. VII 6.3 15 11.8 375 0.24

13. VIII 9.8 16 0.18

30. VI 1.5 4.1 11 0.06

8.VII 3. 7 12 12.7 0.20

16.VII 7.4 12 6.11

29. VII 6.4 15 11.9 277 0.30

13.VIII 9.8 19 0.37

30 . VI 1.2 2.1 11 13.1 0.09

8.VJI 7.9 14 12.1 0.08

16.VII 8.1 15 12.0 0.12

29 . VII 6. 7 18 11.7 245 0.15

13. VIII 9. 7 20 0.36

30. VI 0. 7 3.8 12 14.4 0.04

8. VII 8.2 16 12.3 0.07

16. VII 8.3 16 12.0 0.17

29 . VII 7.0 18 11.6 273 0.14

3. VI I 1.3 6.1 12.8 0.20

12. VII 8.6 13 0.29

23 . VII 6.6 15 0.92

3. VIII 5.2 17 12.2 10 427 1.25

13.VIII 1.1 10.0 10 890 18 0.26

6. VII 1.0 9.8 12 0.30

27. VII 6.5 15 0.10

10 12.VII 0.8 8.7 8 1.31

11 12. VI I 0.3 7. 7 12 0.05

12 3.VJI 1.1 7. 5 13 12.3 0.13

12. VII 9.0 16 0.16

23. VII 5.8 16 0.36

3. VIII 4.5 18 12.4 315 0.27 14 3.VJI 0.4 9.8 13 ' - 0.09 15 3.VII 0.3 9. 7 17 0.04

16 3.VJI 0.4 10.8 15 0.10

18 3.VJI 0.9 7.5 15 12.2 0.28

12.VIII 10.0 26 16 0.34

19 3. VII 0.6 7.1 13 12.3 0.01

12.VIII 10.1 17 0.14 23 1. VII 0.5 5.1 22 0.09

11.VII 8.9 26 0. 58

24 l.VJI 0. 7 2.8 0.20

ll.VJI 7.6 0.59

25 1.VII 0.8 5. 7 16 11.4 0.59

10. VII 8.3 18 0.53

32 13.VII 1.4 7. 7 10 12.6 0.19

5.VIII 6.9 12 0.22

40 13.VII 0. 7 9.8 16 12.4 0.45

41 13.VII 0.6 11.8 13 11.9 0.78

5. VIII 8.3 16 1.33

42 13. VII 0.6 12.1 23 12.2 0.77

43 13. VII 0.5 12.2 24 12.0 0.10

44 6.VJII 1.2 8.4 26 0.10

46 6.VIII 0.5 9.1 15 0.12

49 6.VIII 0.4 10.2 20 0.08

50 6.VIII 0.8 8.9 19 272 0.33

53 6.VIII 0.6 8. 7 15 537 10 0. 76

A eta Phytogeogr.. Suec. 68 Phytoplankton fr om lakes and ponds on Vestspitsbergen 183

Table II (conti n.)

Loc . Date Ma x. Temp. Conductivi ty 0 Cl Total N Total P 2 - - Phytoplankton depth o No . 1962 c ° 1 -1 - 1 -1 3 l m 20 , mS m -l ms 1 - m9 1 �9 1 � 1 • biomass , mm 1 -

60 18. VII 1.9 2.4 0.02 1.VIII 2.7 13 149 0.03 67 6.VII 0.6 10.0 8 11.5 0.42 27. VII 7.4 11 0.16 68 6. VI I 0. 7 10.4 11.0 0.11 69 6. VII 1.1 10.3 9 11.2 0.15 27. VII 7.2 11 0.25 70 27. VII 0. 7 6.4 13 0.16 83 9. VIII 0.6 3.4 62 0.07 98 6.VII 0.3 10.3 0.17

some degree influenced by sea water, have been dis­ mm3 ,-l �0,------, cussed as a separate group (see below). • Loc. 83 as well as Loc. 60, situated ea 100 m • above sea level (Fig. 2) are characterized as extreme • • • arctic ponds with low water temperature and a short 10 • • • • • • • • ice-free period (Amren 1964c). Loc. 83 showed a • • • • • •• • • deviating value of conductivity, 62 mS m-1• It has a • • • • • • • • • • • • • • • • size of 10x20 m and "it is of a nature of a well" • • • • • • 01 • • • • •• • • (Amren op. cit., p. 222). • • • • • I As mentioned above, Loc. 23 and 68 are • • • • characterized as puddles, viz. bodies of water which • • dry up fo r varying periods of the year (R0en 1962). • Some ponds manured by birds deviated from all Fig. 8. Total phytoplankton biomass, mm3 1· 1, in the in­ others as to water chemical conditions and phyto­ vestigated ponds. plankton composition (for example, Loc. 7 and 53). As a rule, the ponds are characterized by the oc­ currence of a very limited number of taxa and of low 100 .,. 1 § total phytoplankton biomasses (Fig. 8). Oscillatoria, � �Fo= ICYANOPHYCEAE § 0 � 0 Gymnodinium, Gyrodinium, small flagellates, some 0 § � �0 0 0 0 0 diatoms, Carteria and Chlamydomonas were often 0 0 0 CHRYSOPHYCEAE� 0 recorded. In some localities the genus Uroglena oc­ 0 0 0 0 curred in immense quantities, for example in Loc. 10 0 0 0 �< 0 0 DBACILLA"IOP'KYCEAE (total biomass of Uroglena and small flagellates 1.3 0 50 0 0 0 0 mm3 l-1), 32, 67, and 69 as well as in Loc. 40-44 all 0 0 0 >0 0 0 0 situated in the same region (Fig. 3). Chrysophyceans 0 0 ICRYPTOPHVCfAE 0 0 0 0 0 0 (Dinobryon, Epipyxis), chlorophytes (Coelastrum, 0 0 0 � 0 Eudorina, Oocystis and Pediastrum) and desmids 0 0 OINOPHYCEAE� 0 (Cosmarium) were recorded less frequently. 0 0 0 0 0 0 :"0 0 0 0 In the fo llowing the composition of phytoplankton 0 0 0 0 0 0 0 0 and the biomass values in some of the investigated 0 0 1. 41 358 5 65 214 110 lrrn0.80 0. 17 mml i·l

29.VII ponds will be discussed. 30.VI 4.VII 8.VII !l VII IS VII 12.VIII

Loc. and As mentioned above, it was pos­ Fig. 9. Vestspitsbergen, Loc. 1. Phytoplankton biomass 1-6 12. sible to fo llow the seasonal phytoplankton develop­ and percentage distribution, 30.Vl-12. VIII. l 962. ment in some ponds. One of them, Loc. 1, had con­ siderably higher biomass values than the other: on 8.VII ea. 5.6 mm3 t1 was recorded (Fig. 9). In the biomass values were observed in early August with other ponds values higher than 0.5 mm3 l-I were very Dinobryon spp. and Uroglena together with small seldom obtained. In general, the highest values were flagellates, respectively, predominating. recorded in the middle of July followed by a marked In all these ponds about the same taxa were decrease. In Loc. 4 and 6, however, the highest observed and the number of species was low. The

A eta Phytogeogr. Suec. 68 184 Torbjiirn Wilten seasonal development, however, was not identical. In high, 0.58 mm3 J-1 on ll.VII, with predominance of Loc. 1 Glenodinium and Gyrodinium predominated Gyrodinium, small flagellates and Diatoma elong­ in June and at the beginning of July. These genera atum together with single specimens of Anabaena, were followed by Diatoma elongatum, but this Oscillatoria and Chlamydomonas. In Loc. 68, species, however, was only seen in single specimens however, the biomass was only 0. 1 1 mm3 l-1 on 6.VII, in August. Rhodomonas, Kephyrion and Monora­ consisting of Gy rodinium, small flagellates and single phidium occurred in small quantities and single specimens of Chrysoikos angulatus. specimens of Anabaena, Cryptomonas, Chrysoikos, Loc. and Amren (1964c p. 216) enumerates six Synedra, Chlamydomonas and Cosmarium were 7 53. localities situated on the coastal plain which are es­ observed. pecially manured by birds. Among these Loc. 20, 26, In Loc. 2 Dinobryonsociale var. americanum and 27 and 51 in this paper are referred to the group of small flagellates predominated and only single ponds which are influenced by sea water. From Loc. specimens of the taxa just mentioned were recorded. 7 and 53 some observations will be mentioned here. Maximum values of Dinobryon were recorded in the middle of July. In Loc. 3 small flagellates From Tables and Ill it is obvious that these 11 all predominated throughout the period; the total localities showed high concentrations of nitrogen and biomass was very low. The same applied to Loc. 4, phosphorus. As a rule they also had high numbers of where, however, a large number of species developed phytoplankton species, thus differing from most in August: Aphanothece, Gomphosphaeria, Mic­ other ponds. rocystis, Gyrodinium, Peridinium, Chrysoikos, Loc. 7 showed a low biomass value on 13.VII, Kephyrion, Carteria and Pediastrum spp. 0.26 mm3 1- 1, but the pond was very rich in species. Loc. 5 was characterized by low total biomass Small flagellates, including Uroglena, dominated with small flagellates predominating. In Loc. 6 a volumetrically. All other taxa enumerated below only rather high number of taxa was recorded in July, occurred in small quantities. The number of when, however, the biomass values were low. In cyanophytes and chlorophytes was high in relation to August Uroglena, small flagellates and Dinobryon all their moderate occurrence in other ponds. Loc. 53 were frequent; the highest total phytoplankton had a somewhat higher biomass value, 0.76 mm3 1-1 volume of the locality was observed on 3.VIII. In on 6.VIII , caused by a predominance of small Loc. 12 finally-the last pond from which seasonal flagellates, mainly Uroglena americana (ea. 0.7 samples were collected-small flagellates, Dinobryon mm3 J-1), Dinobryon sociale var. americanum and and Diatoma (on 3.VII) dominated volumetrically. Diatoma elongatum. Some cyanophytes and The highest biomass value, 0.36 mm3 l-1, was chlorophytes were also recorded, but the total recorded on 23. VII. number of taxa was smaller than that of Loc. 7. The following taxa were observed in Loc. 7 and Loc. The Loc. is an example of an extreme arctic 53: 60. pond (water temperature 2.4 and 2.7°C on 18.VII Chroococcus limneticus, C. minutus, Gomphosphaeria and LVIII, respectively). The biomass of lacustris var. compacta, Microcystis incerta, M. pulverea, phytoplankton was very low, 0.02-0.03 mm3 J-1 and Oscillatoria sp., Pseudanabaena catenata. Dinobryon only small flagellates, Rhodomonas minuta and sociale var. americanum, Epipyxis polymorpha, Uroglena single specimens of Gyrodinium were recorded. americana, Volvochrysis globosa. Diatoma elongatum, Synedra spp. A nkistrodesmus fa lcatus, Botryococcus braunii, Loc. Also this extreme arctic pond had a low 83. Chlamydomonas spp., Chlamydocapsa planctonica, biomass characterized by the occurrence of small Cosmarium Dictyosphaerium pulchellum, Franceia flagellates, Diatoma elongatum and Meridion cir­ sp., droescheri, Oocystis spp., Pediastrum boryanum, P. in­ culare. The latter species was only seen in this tegrum f. granulatum, P. muticum, Willea irregularis. sample; according to Foged (1964 p. 55), however, it seems to be rather frequently occurring: "Alkali­ Loc. and . Two ponds rather rich in species. 41 67 philous. Mainly attached to running water, but also From Loc. 41 samples were taken on 13.VII (water fairly numerous in lakes and ponds." temperature 11.8°C) and on 5. VIII (8.3° C). In July Dinobryon divergens, Synura and Uroglena Loc. and The capability of cyst formation is a predominated together with single coenobia of 23 68. condition for the survival of organisms in most Coelastrum and Pediastrum; biomass 0.78 mm3 J-1• waters of Spitsbergen. This is especially important in In August the biomass had increased to 1.33 mm3 l-1• the puddles; from these three samples have been Among the cyanophytes Chroococcus, Gompho­ analysed. In Loc. 23 the biomass value was rather sphaeria and Microcystis were frequent. Dinobryon

A eta Phytogeogr. Suec. 68 Phytoplankton fr om lakes and ponds on Vestspitsbergen 185

Table II I. Analytical data from local ities infl uenced by sea water (samples from a depth of 0.3 m or less).

Max Loc . Date Temp . Conducti vl'ty 0 Cl Total-N Total�P Phytopl ankton depth 2 o No . 1 962 m c 1 l 1 1 3 -l 20° , mS m- mg 1- mg 1-l l- 1- biomass, mm 1 �9 �9

20 l.VII 0.5 5.6 109 12.5 10.26 4.VII 11.5 107 11.6 320 860 63 5. 29

ll.VI I 7.4 107 120 7 1.68 29 .VII 6.2 116 12.1 349 306 3 1 0.49 13. VII I 8.8 128 0. 27 21 l .VII 0.3 5. 3 17 ll.9 l.3 5 ll.VI I 8.5 16 11.7 0.99 13. VI I I 24 50 0.20 22 l.VI I 1.5 2. 2 92 5. 52 4.VII 7. 9 104 11.1 405 1330 64 6.76

11.VI I 7 0 3 97 14.4 13.17 29 . VII 6.9 163 12.3 474 676 34 4.50 13. VIII 9 .2 182 6.83 26 1.VI I 0.8 5. 7 118 12.8 0.85 lO.VII 8.4 117 0.67 13. VII I 10.6 134 386 1030 20 1.04 27 10. VII 0.8 8.4 145 0.62 13. VIII 9 .6 177 538 2080 81 3.30 28 10. VI I 0.4 8.0 115 0.21 51 6. VIII l.O 8.9 21 32 560 50 1.62

divergens and Uroglena were still abundant and position and deviated from all other localities as to Dinobryon cylindricum was also recorded. Among total phytoplankton volumes, they are here discussed the chlorophytes Coelastrum cambricum and C. as a separate ecological group. microporum were counted in ea. 4 7 000 small From Fig. 10 it is quite evident that the phyto­ coenobia per litre and Pediastrum boryanum and P. plankton biomass values in these waters were con­ duplex in ea. 15 000 coenobiaper litre. Loc. 41 was siderably higher than corresponding values from the the only locality where Pediastrum braunii was lakes and ponds discussed above. While values higher observed. This species was very frequent in a small than 1 mm3 very seldom were recorded in the 1-1 lake, Krillvatnet, on Bjornoya (Willen 1970), which is freshwater localities, most values varied between 1 manured by birds. and 13 mm3 in the localities influenced by sea r1 From Loc. 67 samples were taken on 6. VII water. Some of these waters (Loc. 20, 26, 2 7 and 51), (10.0° C) and on 27.VII (7.4° C). In this pond the were also manured by birds, a condition increasing biomass values were considerably lower, 0.42 and the effect of the influence of sea water. 0. 16 mm3 respectively, and there was another r1, species composition. Gyrodinium, Dinobryonsociale Loc. In Loc. 20 the development was fo llowed 20. var. americanum, Uroglena and Synedra were during July and the firsthalf of August (Fig. 1 1). The predominating and among other genera the following may be mentioned: Anabaena, Gomphosphaeria, mm3 1-l

Oscillatoria, Pseudanabaena, Glenodinium, Epi­ • pyxis, Mallomonas, Diatoma, Carteria Stau­ 100 • and • • • • rodesmus. • •

• • • • • Localities influenced by sea water 10 • • Phytoplankton samples were also collected from • • seven localities which all showed high conductivity • • values and high chloride concentration (Table Ill). 01 -l....,_---r------.,...--�---__,....,------,-,------' They are all situated near the coast (Figs. 2 and 3). Fig. 10. Total phytoplankton biomass, mm3 1-1, in the Since they also held a specific phytoplankton corn- localities influenced by sea water.

Acta Phytogeogr. Suec. 68 186 Torbjiirn Wi/!en

. •t. . r- - 100 � - - ,- 100 , ,- - r- ' I ICYANOPHYCEAE DCHLOROPHYCEAE � ' § � ' D �CHRYSOPHYCEAE CHLOROPHYCEAf ' ' ' � �CHRYSOPHYCfAE 0BACILLARIOPHYCEAE 50 50 '

0 ICRYPTOPHYCEAE IIACILLAIIIOPHYCEAE � � � � ICRYPTOPHYCEAE DINOPHYCEAE

�0 0 ;-0 0 0 0 tmr 0 �0 0 0.49 552 6 76 7 4 50 iiTIT6 1Cl26 529 1 se 0 27 mm3 1-1 13 1 � 83 mm3 r1 29VII l VII 29V11 13VIII 1 VII 4 VII 11 VII 13VIII LVII IIVII Fig., 11. Vestspitsbergen. Loc. 20. Phytoplankton biomass Fig. 12. Vestspitsbergen, Loc. 22. Phytoplankton biomass and percentage distribution, l.VII-13.Vlll.l 962. and percentage distribution, 1.VII- 13. Vlll.l 962.

biomass values were highest in early July with million cells l-1• In August single chlorophytes also dominance of Diatoma elongatum and Carteria sp. occurred. In addition to the taxa enumerated above, Diatoma was counted in more than 13 million cells t1 Ankistrodesmus fa lcatus, Ankyra judayi and Pedi­ on 1.VII, about 9 million cells t1 on 4.VII but only in astrum boryanum were recorded. 0. 7 million cells 1-1 on 11.VII. In August ea. 0.3 As to the percentage distribution (Fig. 12), the

million cells 1-1 were recorded. Small flagellates chlorophytes predominated only in early July predominated from 11. VII and from this date Ankyra followed by diatoms and chrysophyceans. The judayi also was rather frequent. Among the succession of genera was Chlamydomonas - Diatoma cyanophytes Aphanothece, Microcystis and - Monochrysis. Oscillatoria were represented by some species in small quantities. The locality was characterized by Loc. and These three ponds were also 26, 27 51. the occurrence of fe w taxa developed in great quan­ manured by birds (Amren 1964c p. 216) and showed tities. rather high amounts of nitrogen and phosphorus. The biomass values were high but the number of species Lake Fy rsjiien, Loc. (Fig. 3). The outlet of Lake was not comparable to the high numbers recorded in 22 Fyrsjoen was blocked in June "by a mixture of ice, Loc. 7 and 53. Small flagellates (mainly Uroglena), gravel and sand" (Amren 1964c p. 217) fo rming a Glenodinium, Diatoma and A nkyra predominated reservoir for melting water (conductivity 26 mS m-1). together with single specimens of Anabaena, Micro­ When this barrier was broken, however, sea water cystis, Oscil/atoria, Pseudanabaena, Rhodomonas, could penetrate, changing the physico-chemical con­ Chlamydomonas and Pediastrum. ditions of the water (conductivity 97-182 mS m-1 during July and August). Very high phytoplankton biomasses were recorded during the whole period with a maximum on 11.VII, ea. 13 mm3 l-1• Com­ SUMMARY pared with Loc. 20, correspondingly high values of chlorophytes were obtained in early July. In Lake The present phytoplankton material from Spits­ Fyrsjoen Chlamydomonas spp. occurred together bergen is collected from lakes and ponds of dif­ with Monoraphidium minutum, Scenedesmus spp., ferent types. In some cases samples were taken Tetrai/dron caudatum and its var. incisum. Certainly 5-7 times during the period of investigation, late Diatoma elongatum was frequent on LVII (ea. 1.5 June to mid-August, 1962, but in many cases only million cells t1), but a peak value of more than 18 one or two samples represent the actual locality. The million cells 1-1 was recorded on 11.VII-about one preservation of the samples (acid Lugol's solution) week later than the corresponding peak value from sometimes restricts the possibility for exact identifica­ Loc. 20. At the end of July and in August tion of species. Most material earlier collected by Monochrysis cf. parva developed in enormous quan­ scientific expeditions, however, was preserved with tities: on 29.VII it was counted in more than 96 formaldehyde, resulting in the destruction of many

A eta Phytogeogr. Suec. 68 Phytoplankton fr om lakes and ponds on Vestspitsbergen 187 delicate organisms, and thus the plankton lists only vironmental conditions, Summerhayes & Elton comprise resistant taxa. (1923) gave algal lists from different types of waters Several factors influenceand determine the quality on Vestspitsbergen: from permanent ponds, tem­ and quantity of the algal floraon Spitsbergen. There porary ponds as well as from tidal ponds. are long severe winters and short cool summers. Thus Predominating groups are desmids, cyanophytes and the growth season is limited : it starts, however, before chlorophytes. The abundance of desmids is also dis­ the ice is broken; examples are given above from cussed by Prescott (1963) who mentions this group Lake Linnevatnet and differentponds. Besides the in­ together with diatoms as being most frequently oc­ solation and light periodicity, the low water tempera­ curring in the Arctic. Reasons for the development of ture is decisive as well as the water chemistry. desmids are the water chemical conditions and the On Spitsbergen we meet an arctic algal flora. large desmid chloroplasts that are effective in critical Lakes Linnevatnet and Kongressvatnet are arctic or light conditions. In the present plankton material, polar lakes belonging to the so-called "high arctic" however, the desmids play an insignificantrole both area (Holmquist 1978). Several authors seem to concerning number of taxa and total biomass. generalize in discussing arctic lakes, writing about ex­ It has been pointed out above that fo rmation of treme oligotrophy, constant thermal regime as well as spores of different kinds (cysts, akinetes, etc.) is biological simplicity. The real complexity, however, is necessary fo r survival and that the reproductive discussed, for example, by Holmquist in the paper period in this region is short. Encysted cells were just quoted. often observed in the samples. These conditions also Ecological aspects on the present material have result in a natural selection of species. In the succes­ already been discussed by Amren, especially con­ sion studies in some ponds (Figs. 9, 11, 12) it was ob­ cerning the significance of small flagellates and some vious that a few species rapidly developed in great other plankton algae as food for zooplankton (Amren quantities to soon replaced by other species often be 1964b pp. 167, 182, etc.; 1964c p. 245). belonging to another algal group. During the period Among the lakes investigated, most interest has July-middle of August Dinophyceae was followed been focused on the biggest: Lake Linnevatnet and by Bacillariophyceae and Chrysophyceae in Loc. 1. Lake Kongressvatnet. Both were covered by ice in In Loc. 20 Bacillariophyceae and Chlorophyceae June 1962, when the firstvertical samplings were per­ were followed by Chrysophyceae and in Loc. 22 the formed. In general, few taxa were recorded in June fo llowing succession was recorded: Chlorophyceae ­ and in August and all phytoplankton biomass values Bacillariophyceae - Chrysophyceae. are low. Small flagellates (Rhodomonas, Chryso­ Especially in the localities influenced by sea water chromulina), Gymnodinium, Stephanodiscus, Syned­ a single organism often developed in enormous quan­ ra, Chlamydomonas and Oocystis are some genera tities and in these localities, as mentioned above, the rather frequently observed. highest plankton biomass values were also recorded From Lake Linnevatnet B0yum & Kjensmo (Figs. 11, 12). Diatoma elongatum, Carteria spp., (1978) have published physiographical data and dis­ Chlamydomonas spp. and some small chrysophy­ cuss the heat budget of the lake. It is characterized as ceans belong to the group of algae which a "typical sulphate lake". predominates in these waters. The temporary very The meromixis of Lake Kongressvatnet is dis­ high biomasses in ponds and lakes sometimes flushed cussed by B0yum & Kjensmo (1970). They con­ by salt or brackish water seem to be of general nature sider that the meromictic stability of the lake is due to and are reported from different coastal waters inflow of mineral water from springs (high content of (Willen 1962). sulphate, calcium and magnesium) and that the lake In comparing the conditions of lakes and ponds of is an example of crenogenic meromixis. Spitsbergen some regions are of special interest. In 1968 the water "at 32.5 m depth in Kongress­ Huber-Pestalozzi (1926) compares arctic waters with vatn had a roseate colour which bears witness to the alpine lakes and ponds in Central Europe. From presence of purple bacteria" (B0yum & Kjensmo Alaska Prescott (1963, etc.) has published a series of 1970 p. 549). In the samples from June, 1962, 40 and papers about phytoplankton and arctic algal ecology, 45 m, the same phenomenon appeared due to an and from lakes of Alaska and Canada Holmquist enormous development of bacteria, mainly Chro­ ( 197 5) gives ecological information and lists of algae. matium cf. minus. Antarctic and subantarctic algae have been treated Samples from different types of ponds have been in at least two compiling works by Hirano (1965) examined. They are mostly shallow, some of them about freshwater algae and by Prescott (1979) in a dry up in summer and several are manured by birds. bibliography and checklist. Hirano compares the Besides some general information about the en- algal flora of arctic and antarctic regions and states

Acta Phytogeogr. Suec. 68 188 To rbjiirn Wi/Len

that, generally, the "algal flora of both polar regions Gj rerevoll, 0. 1975. MAB in Norway: Svalbard-the are similar". The arctic region, however, is more ecological "Galapagos" of the Arctic. - Research in complex and partly influenced by warm currents. Norway 1974 :46-54. Due to environmental differences as well as Hilliard, D.K. & Asmund, B. 1963. Studies on Chrysophyceae from some ponds and lakes in Alaska. differences in sampling and analyses of the algae it is, 11. Notes on the genera Dinobryon, Hyalobryon and naturally, difficult to get a real comparison. No es­ Epipyxis with descriptions of new species. - sential difference in the distribution of chlorophytes is Hydrobiologia 22:33 1-397. observed, the cyanophytes are less frequent in the Hirano, M. 1965. Freshwater algae in the Antarctic Arctic, while the desmids are considerably more fre­ regions. - Monographiae bioi. 15:127-1 93. quent in the Arctic. In a table, Hirano (op. cit., p. Holmquist, C. 1975. Lakes of Northern Alaska and 164) summarizes the number of arctic algae and Northwestern Canada and their invertebrate fauna. - gives the fo llowing numbers for Spitsbergen : Zoo!. Jb. Syst. 102:333-484. Cyanophyceae 31, Chlorophyceae incl. fil. Con­ 1978. Lakes of polar regions-age, complexity and trophic state. - Verh. internat. Verein. theor. angew. jugatae 35, desmid 136, Heterokontae 3, Chry­ Limnol. 20:615-617. sophyceae 1, diatoms 250 and Rhodophyceae Huber-Pestalozzi, G. 1926. Die Schwebeflora (das 1. From Bjornoya only 75 diatoms are reported. The Phytoplankton) von Seen und Kleingewiissern der table was published in 1965 and since then the alpinen und nivalen Stufe. - Zurich. 104 pp. number of observations has increased. Several Korshikow, A.A. 1953. Protococcineae of the Ukraine (in chrysophyceans from Spitsbergen have been added Russian).- Vizn. prisnov. vodor. Ukr. R.5. R5:1-439. by the present study and the list from Bjornoya has Lackey, J .B. 1942. The plankton algae and Protozoa of also been supplemented (Willen 1970). two Tennessee rivers. - Am. Midi. Nat. 27:191-202. Nauwerck, A. 1979. Zur Gattung Chrysolykos Mack. - Bot. Notiser 132 :161-183. Acknowledgements Prescott, G.W. 1963. Ecology of Alaskan freshwater algae. 11. Introduction: General considerations. - I am much indebted to Dr. H. Amren who kindly put the Trans. Am. microsc. Soc. 82:83-98. phytoplankton samples and all physico-chemical data at 1979. A contribution to the bibliography of Antarctic my disposal. Inking of the figures has been performed by and Subantarctic algae. - Bibliotheca phycologica Mrs. M. Ehrnst and Mrs. B. Krizsan. 45:1-312. R0en, U.l. 1962. Studies on freshwater Entomostraca in Greenland. 11. - Meddr. Gmnland 170:1-249. REFERENCES Skuja, H. 1956. Taxonomische und biologische Studien Amren, H. 1964a. Studies on the ecology and taxonomy of iiber das Phytoplankton schwedischer Binnengewiisser. zooplankton especially the rotifers from Spitsbergen.­ - Nov. Acta. R. Soc. Scient. upsal. Ser. IV. 16 Acta univ. upsal. Abstr. Upps. Diss. Fac. Science (3):1-404. 42:1-10. Summerhayes, V.S. & Elton, C.S. 1923. Contributions to 1964b. Ecological studies of zooplankton populations the ecology of Spitsbergen and Bear Island. - J. Ecol. in some ponds on Spitsbergen. - Zoo!. Bidr. Upps. 11:214-286. 36:161-191. Thomasson, K. 1958. Zur Planktonkunde Spitzbergens. - 1964c. Ecological and taxonomical studies on Hydrobiologia 12:226-236. zooplankton from Spitsbergen. - Zoo!. Bidr. Upps. - 1961. Zur Planktonkunde Spitzbergens. 2. 36:209-276. Hydrobiologia 18:192-1 98. Utermohl, H. 1958. Zur Vervollkommnung der quan­ Borge, 0. 1911. Die SiissvasseralgenfloraSpitzberg ens. - Skr. Vidensk.-Selsk. Christiania I. Mat.-naturv. Klasse titativen Phytoplankton-Methodik. - Mitt. internat 11: 1-38. Verein. theor. angew. Limnol. 9:1-38. B0yum, A. & Kjensmo, J. 1970. Kongressvatn. A Willen, T. 1962. The Utal Lake Chain, Central Sweden, crenogenic meromictic lake at Western Spitsbergen.­ and its phytoplankton. - Oikos, Suppl. 5:1-156. Arch. Hydrobiol. 67:542-552. 1963. Notes on Swedish phytoplankton algae. - Nova 1978. Physiography of Lake Linnevatn, Western Hedwigia 5:39-56. Spitsbergen. - Verh. internat. Verein. theor. angew. 1967. Chrysoikos, a new genus of Ochromonadales Limnol. 20:609-614. (Chrysophyta). - Phycologia 6:96-104. Foged, N. 1964. Freshwater diatoms from Spitsbergen. ­ 1970. Phytoplankton from Bjornoya, Svalbard. - Nytt Troms0 M us. Skr. 1 1:1-159. Mag. Bot. 17:17-24.

Torbjorn Willt�n, National Swedish Environment Protection Board, Water Quality Laboratory, Box 8043, S-750 08 Uppsala, Sweden

A eta Phytogeogr. Suec. 68 189 SVENSKA V A.XTG EOGRAFISKA SALLSKAPET SOCIETAS PHYTOGEOGRAPH ICA SUECANA

Adress: Vaxtbiologiska institutionen, Box 559, S-75 1 22 Uppsala, Sweden

Sallskapet har till andamal att vacka och underhalla intresse The object of the Society is to promote investigation in flora for vaxtgeografien i vidstracktaste mening, att framja utfors­ and vegetation, their history and their ecological background. kande av flora och vegetation i Sverige och andra lander och Through publication of monographs, and other activities, the att havda geobotanikens praktiska och vetenskapliga betydel­ Society tries to stimulate geobotanical research and its appli­ se . cation to practical and scientific problems. Membership is Sallskapet anordnar sammankomster och exkursioner samt open to all who have a personal interest in the ad vancement of utger tva publikationsse rier. Medlemskap kan erhallas efter phytogeography. anmalan hos sekreteraren. Foreningar, bibliotek, laroanstaJter Individual members and subscribers (societies, institutes, och andra institutioner kan inga som abonnenter. Arsavgift 50 libraries, etc.) receive the Acta Phytogeographica Suecica for kr (35 for studerande). annual dues of 50 Skr plus postage. There are additional fe es Sallskapet utger arligen Acta Phytogeographica Suecica. in years when more than one volume are issued. For member­ Medlemmar och abonnenter erhaller arets Acta mot postfor­ ship please apply to the Secretary. skott pa arsavgiftenjamte porto och expeditionskostnader. The Society also issues Viixtekologiska studier, which ap­ Vissa ar utges extraband av Acta, som erhalls mot en till­ pear irregularly and are available upon request or standing laggsavgift. order. Sallskapet utger ocksa den ickeperiodiska serien Viixteko­ Both series can be received by exchange for other scientific logiska studier. Den kan fo rvarvas efter bestallning eller ge­ publications. Please apply to the Institute Library (address as nom staende abonnemang hos Sallskapet. above). Bada serierna kan ocksa erhallas i byte mot andra publika­ tioner efter hanvandelse till Vaxtbiologiska institutionens bib­ liotek.

ACTA PHYTOGEOGRAPHICA SUECICA

I. E. Almquist, Upplands vegetation och flora. (Vegetation 10. B. Lindquist, Dalby Soderskog. En skansk lovskog i and floraof Uppland.) 1929. ISBN 91-72 10-001-X. fo rntid och nutid. (Zusamme nf. : Ein Laubwald in Scho­ 2. S. Th unmark, Der See Fiolen und seine Vegetation. 1931. nen in der Vergangenheit und Gegenwart .) 1938. 35:-. 30:-. ISBN 91-72 10-002-8. ISBN 91-72 10-0 10-9. 3. G. E. Du Rietz, Life-forms of terrestrial flowering plants. 11. N. Sllilberg, Lake Vattern . Outlines of its natural history , I. 1931. 20:-. ISBN 91-72 10-003-6. especially its vegetation. 1939. 15:-. ISBN 91-72 10-01 1-7.

4. B. Lindquist, Om den vildvaxande skogsalmens raser och 12. G. E. Du Rietz, A. G. Hannerz, G. Lohammar, R. Son­ deras utbredning i Nordvasteuropa. (Summary: The races lesson & M. Wa>rn, Zur Kenntnis der Vegetation des Sees of spontaneous Ulmus glabra Huds. and their distribution Takern . 1939. 15:-. ISBN 91-72 10-0 1 2-5. in NW. Europe.) 1932. 15:-. ISBN 91-72 10-004-4. 13. Viixtgeografiska studier tilliignade Car/ Skottsberg pa

5. H. Osvald, Vegetation of the Pacific coast bogs of North sextioarsdagen I I 12 1940. (Geobotanical studies dedica­ America. 1933. 15:-. ISBN 91-72 10-005-2. ted to C. Skottsberg.) 1940. 50:-. ISBN 91-72 10-0 13-3 . 6. G. Samuelsson, Die Verbreitung der hoheren Wasser­ 14. N. Hylander, De svenska fo rmerna av Mentha gentilis L. pflanzen in Nordeuropa. 1934. 30:-. ISBN 91-72 10-006-0. coli. (Zusammenf. : Die schwedischen Form en der Men­ 7. G. Degelius. Das ozeanische Element der Strauch- und tha gentilis L. sensu coli.) 1941. 15:-. ISBN 91-72 10- Laubflechtenflora von Skandinavien. 1935. 50:-. ISBN 014-1. 91-72 10-007-9. 15. T. E. Hasselrot, Till kannedomen om nagra nordiska um­ 8. R. Sernander, Granskar och Fiby urskog. En studie over bilicariaceers utbredning. (Zusammenf. : Zur Kenntnis der stormluckornas och marbuskarnas betydelse i den sven­ Verbreitung einiger Umbilicariaceen in Fennoscandia.) ska granskogens regeneration. (Summary: The primitive 1941. 15:-. ISBN 91-72 10-01 5-X. fo rests of Granskar and Fiby. A study of the part played 16. G. Samuelsson, Die Verbreitung der Alchemilla-Arten by storm-gaps and dwarf trees in the regeneration of the aus der Vulgaris-Gruppe in Nordeuropa. 1943 . 25:-. Swedish spruce forest.) 1936. 30:-. ISBN 91-7210-008-7. ISBN 91-72 10-016-8. 9. R. Sterner, Flora der Insel bland. Die Areale der Ge­ 17. Th. Arwidsson, Studien uber die Gefasspflanzen in den fa sspflanzen Olands nebst Bemerkungen zu ihrer Hochgebirgen der Pite Lappmark. 1943. 35:-. ISBN 91- Oekologie und Soziologie. 1938. ISBN 91-72 1 0-009-5. 72 10-0 1 7-6.

Acta Phytogeogr. Suec. 68. 190

18. N. Dahlbeck, Strandwiesen am si.idostlichen bresund. mountains in Southwestern Jamtland and adjacent parts (Summary: Salt marshes on the S. E. coa t of bresund.) of Harjedalen (Sweden) and Norway.) 1955. 30:-. ISBN 1945. 25:-. ISBN 91-72 10-0 1 8-4. 91-72 10-035-4. 19. E. von Krusenstjerna , Bladmossvegetation och blad­ 36. N. Quennerstedt, Diatomeerna i Langans sj ovegetation . mossflora i Uppsalatrakten. (Summary: Moss flora and (Su mmary: Diatoms in the lake vegetation of the Ungan moss vegetation in the neighbourhood of U ppsala.) 1945. drainage area, Jamtland, Sweden.) 1955. 30:-. ISBN 35:-. ISBN 91-72 10-0 19-2. 91-72 10-036-2. 20. N. Albertson, bsterplana hed. Ett alvaromdide pa Kin­ 37. M.-B. Florin , Plankton of fr esh and brackish waters in the nekulle. (Zusammenf. : bsterplana hed. Ein Alvargebiet Sodertalje area. 1957. 25:-. ISBN 91-721 0-037-0. auf dem Kinnekulle.) 1946. ISBN 91-7210-020-6. 38. M.-B. Florin , Insj ostudier i Mellansverige . Mikrovegeta­ 21. H. Sj ors , Myrvegetation i Bergslagen. (Summary: Mire tion och poUenregn i vikar av bstersjobackenet och insjoar vegetation in Bergslagen, Sweden.) 1948. 50:-. ISBN fran preboreal tid till nutid. (Summary: Lake studies in 91-72 10-02 1-4. Central Sweden. Microvegetation and pollen rain in inlet 22. S. Ahlner, Utbredningstyper bland nordiska barrtrad la­ of the Baltic ba in and in lakes from Preboreal time to the var. (Zusammenf. : Verbreitungstypen unter fe nnoskandi­ present day.) 1957. 15:-. ISBN 91-72 1 0-038-9. chen Nadelbaumflechten.) 1948. 40:-. ISBN 91-7210- 39. M. Fries, Vegetationsutveckling och odlingshistoria i . 022-2. Varnhemstrakten. En pollenanalytisk undersokning i 23. E. Julin , Vessers udde, Mark och vegetation i en igen­ Vastergotland. (Zusammenf. : Vegetationsentwicklung vaxande lovang vid Bjarka-Saby. (Zusammenf. : Vessers und Siedlungsgeschichte im Gebiet von Varnhem. Eine udde. Boden und Vegetation in einer verwachsenden pollenanalytische Untersuchung aus Vastergotland Laubwiese bei Bjarka-Saby in bstergotland, Siidschwe­ (Siidschweden).) 1958. 20:-. ISBN 91-72 1 0-039-7. den.) 1948. 30:-. ISBN 91-7210-023-0. 40. Bengt Pellersson, Dynamik och konstans i Gotlands flora 24. M. Fries , Den nordiska utbredningen av Lactuca alpina, och vegetation. (Resume: D'•namik und Konstanz in der Aconitum septentrionale, Ranunculus platanifolius och Flora und Vegetation von Gotland, Schweden.) 1958. Polygonatum verticillatum. (Zusammenf. : Die nordische 65:-. ISBN 91-72 1 0-040-0. Verbreitung von Lactuca alpina ...) 1949. 20:-. ISBN 41. E. Uggla, Skogsbrandfalt i Muddus nationalpark . (Sum­ 91-7210-24-9. mary : Forest fire areas in Muddus National Park, North­

25. 0. Gjrerevoll , Snji'Sieievegetasjonen i Oviksfjellene. (Sum­ ern Sweden.) 1958. 25:-. ISBN 91-72 10-04 1-9. mary: The snow-bed vegetation of Mts Oviksfjallen, Jamt­ 42 K. Th omasson, Nahuel Huapi. Plankton of some lakes in land, Sweden.) 1949. 25:-. ISBN 91-72 10-025-7. an Argentine National Park, with notes on terrestrial vege­ 26. H. Osvald, Note on the vegetation of British and Irish tation. 1959. 25:-. ISBN 91-72 10-042-7. mos es. 1949. 15:-. ISBN 91-72 10-026-5. 43 . V. Gillner, Vegetations- und Standortsuntersuchungen in 27 . S. Selander, Floristic phytogeography of South-Western den Strandwiesen der schwedischen Westki.iste . 1960. Lule Lappmark (Swedish Lapland). I. 1950. 30:-. ISBN 40:-. ISBN 91-72 1 0-043-5. 91-72 10-027-3. 44. E. Sj ogren, Epiphytische Moosvegetation in Laubwal­ 28. S. Selander, Floristic phytogeography of South-Western dern der Insel bland, Schweden. (Summary : Epiphytic Lule Lappmark (Swedish Lapland). II. Karlvaxttloran i moss communities in deciduous woods on the i land of ydva tra Lule Lappmark. (Summary: Va cular flora.) bland, Sweden.) 1961. 30:-. ISBN 91-72 1 0-044-3 (ISBN 1950. 25:-. ISBN 91-72 10-028- 1. 91-71 1 0-444-9) . 29. M. Fries, Pollenanalytiska vittnesbord om enkvartar ve­ 45 . G. Wistrand, Studier i Pite Lappmarks karlvaxtflora, med getationsutveckling, sarskilt skogshistoria, i nordvastra arskild hansyn till skogslandet och de isolerade fjallen. Gotaland. (Zusammenf. : Pollenanalytische Zeugnisse der (Zu ammenf. : Studien i.iber die Gefasspflanzenflora der spatquartaren Vegetationsentwicklung, hauptsachlich der Pite Lappmark mit besonderer Beri.icksichtigung des Waldgesc hichte, im nordwe tlichen Gotaland, Si.id­ Waldlandes und der isolierten niederen Fj elde.) 1962. schweden.) 1951. 30:-. ISBN 91-72 1 0-029-X. 45:-. ISBN 91-72 10-045- 1 (ISBN 91-72 1 0-445-7). 30. M. Wtern, Rocky-shore algae in the bregrund Archipela­ 46. R. lvarsson, Lovvegetation i Mollosunds socken. (Zu­ go . 1952. 45:-. ISBN 91-72 10-030-3. sammenf. : Die Laubvegetation im Kirchspiel Mollosund,

31. 0. Rune, Plant life on serpentines and related rocks in the Bohuslan, Schweden.) 1962. 35:-. ISBN 91-72 10-046-X North of Sweden. 1953. 25:-. ISBN 91-72 10-03 1-1. (ISBN 91-72 10-446-5). 32. P. Kaaret, Wasservegetation der Seen Orlangen und Tre­ 47. K. Thomasson, Araucanian Lakes. Plankton studies in horningen. 1953. 15:-. ISBN 91-7210-032-X. North Patagonia, with notes on terrestrial vegetation. 33. T. E. Hasselrot, Nordliga lavar i Syd- och Mellansverige. 1963. 45:-. ISBN 91-72 10-047-8. (Nordliche Flechten in Si.id- und Mittelschweden.) 1953. 48. E. Sj ogren, Epilitische und epigaische Moosvegetation in 30:-. ISBN 91-72 10-033-8. Laubwaldern der Insel bland, Schweden. (Summary : 34. H. Sjo rs , SHitterangar i Grangarde fi nnmark . (Summary: Epilithic and epigeic moss vegetation in deciduous woods Meadows in Grangarde Finnmark, SW. Dalarna, Swe­ on the island of bland, Sweden.) 1964. 50:-. ISBN den.) 1954. 25:-. ISBN 91-72 10-034-6. 91-72 10-048-6 (ISBN 91-72 10-448- 1).

35. S. Kilander, Karlvaxternas ovre granser pa fjall i sydvast­ 49. 0. Hedberg , Features of afroalpine plant ecology. (Re­ ra Jamtland samt angransande delar av Harjedalen och sume fran�ais.) 1964. 50:-. ISBN 91-72 1 0-049-4 (ISBN Norge. (Summary: Upper limits of vascular plants on 91-72 10-449-X).

Acta Phytogeogr. Suec. 68 191

50. The Plant Cover of Sweden . A study dedicated to G. vasculaires dans la fo ret dense humide d'Afri que occiden­ Einar Du Rietz on his 70th birthday by his pupils. 1965 . tale.) 1974. ISBN 91-72 10-059- 1 (ISBN 91-72 10-459-7). 100:-. ISBN 91-72 1 0-050-8. 60. H. Olsson. Studies on South Swedish sand vegetation, 51. T. Flensburg, Desmids and other benthic algae of Lake 1974 . 70:-. ISBN 91-72 1 0-060-5 (ISBN 91-7210-460-0). Kavsjon and Store Mosse, SW Sweden. 1967 . 50:-. ISBN 61. H. Hytteborn, Deciduous woodland at Andersby, Eastern 91-72 10-05 1-6 (ISBN 91-72 10-451-1). Sweden. Above-ground tree and shrub production. 1975.

52. £. Skye, Lichens and air pollution. A study of cryptoga­ 55:-. ISBN 91-72 10-061-3 (ISBN 91-72 10-46 1-9). mic epiphytes and environment in the Stockholm region. 62. H. Persson, Deciduous woodland at Andersby, Eastern 1968. 60:-. ISBN 91-72 1 0-052-4 (ISBN 91-72 1 0-452-X). Sweden: Field-layer and below-ground production . 1975 . 53. Jim Lundqvist, Plant cover and environment of steep 50:-. ISBN 91-72 10-062- 1 (ISBN 91-72 10-462-7). hillsides in Pite Lappmark. (Resume: La couverture vege­ 63 . S. Brakenhielm, Vegetation dynamics of afforested fa rm­ tale et !'habitat des flancs escarpes des collines de Pite land in a district of South-eastern Sweden. 1977. 65:-. Lappmark.) 1968 . 60:-. ISBN 91-72 10-053-2 (ISBN 91- ISBN 91-72 10-063-X (ISBN 91-72 1 0-463-5). 72 I 0-453-8). 64. M. Ammar, Vegetation and local environment on shore 54 . Conservation of Vegetation in Africa South of the Saha­ ridges at Vickleby, bland, Sweden. An analysis. 1978. ra . Proc . of ymp. at 6th plen. meeting of AETFAT. Ed. 65:-. ISBN 91-72 10-064-8 (ISBN 91-72 10-464-3). by Inga and Olov Hedberg. 1968 . 80:-. ISBN 91- 65. L. Kul/man, Change and stability in the altitude of the 72 1 0-054-0 (ISBN 91-72 10-454-6). birch tree-limit in the southern Swedi h Scandes 1915- 55. L.-K. Konigsson, The Holocene history of the Great Al­ 1975 . 1979. 65:-. ISBN 91-72 10-065-6 (ISBN 91-72 10- var of bland. 1968. 70:-. ISBN 91-72 10-055-9 (ISBN 465- 1).

91-72 10-455-4). 66. £. W aldemarson Jensen, Successions in re lation hip to 56. H. P. Hallberg, Vegetation auf den Schalenablagerungen lagoon development in the Laitaure delta, North Sweden. in Bohuslan , Schweden. (Summary: Vegetation on shell 1979. 65:-. ISBN 91-72 1 0-066-4 (ISBN 91-7210-466-X). deposits in Bohuslan, Sweden.) 1971. 60:-. ISBN 91- 67. S. Tu hkanen, Climatic parameters and indices in plant 72 1 0-056-7 (ISBN 91-72 10-456-2). geography. 1980. 65 :-. ISBN 91-72 1 0-067-2 (ISBN 91- 57. S. Fransson, Myrvegetation i ydvastra Varmland. 72 I 0-467 -8). (Summary : Mire vegetation in south-western Varmland, 68. Studies in plant ecology dedicated to Hugo Sj ors . Ed. Sweden.) 1972. 55:-. ISBN 91-72 10-057-5 (ISBN 91- Erik Sjogren. 1980. 95:-. ISBN 91-72 1 0-068-0 (ISBN 91-

72 1 0-457-0). 72 1 0-468-6) 0 58 . G. Well/in , Lovskogsvegetation i Sj uharadsbygden. (Summary: Deciduous woodlands in Sj uharadsbygden, Limited numbers of cloth-bound copies of Acta 44 , 45 , 46, 48, Vastergotland, south-western Sweden.) 1973 . 55:-. ISBN 49, 51, 52. 53, 56, 57, 61, 63 , 66, 67 , 68 are available through 91-72 10-058-3 (ISBN 91-72 1 0-458-9). the Society at an additional cost of 15:- per copy. ISBN nos.

59. D. Johansson, Ecology of vascular epiphytes in West in brackets refer to cloth-bound copies. Nos. I, 9, 20, 59 are African rain forest . (Re ume: Ecologie des epiphytes out of print.

Acta Phytogeogr. Suec. 68 192

V AXTEKOLOGISKA STUDIER

I. S. Brakenhielm & T. lngelog, Vegetationen i Kungs­ 7. L. Rodenborg, Bodennutzung, Pflanzenwelt und d ihre hamn-Morga naturreservat med forslag till skotselplan. Veranderungen in einem alten Weidegebiet auf MMittei­ (Summary: Vegetation and proposed management in the Oland , Schweden. 1976. 20:-. ISBN 91-72 10-807-X .. Kungshamn-Morga Nature Reserve south of Uppsala.) 8. H. Sj ors & Ch . Nilsson, Vattenutbyggnadens effekt

72 1 0-802-9 0 9. J. Lundqvist & G. Wistrand, Strandtlora inom ovnre och 3. H. Sj ors och medarb., Skyddsvarda myrar i Kopparbergs mellersta Skelleftealvens vattensystem. Med en samnman­ Jan. (Summary: Mires considered for protection in Kop­ fattning betraffande botani ka skyddsvarden. (Sumnmary: parberg County (Prov. Dalarna, Central Sweden).) 1973 . Riverside vascular flora in the upper and middle ccatch­ 20:-. ISBN 91-72 10-803-7. ment area of the River Skelleftealven, northern Swereden.) 4. L. K arlsson, Autecology of cliff and cree plants in Sarek 1976. 25:-. ISBN 91-72 10-809-6. National Park, northern Sweden. 1973 . 20:-. ISBN 10. A. Miiller-Haeckel, Migrationsperiodik einzelliger /Algen

91-72 I 0-804-5. in Flie sgewassern. 1976. 10:-. ISBN 91-72 10-8 10-X . 5. B. Klasvik, Computerized analysis of stream algae . 1974. 11. Sjo din , Index to distribution maps of bryopbhytes A. 20:-. ISBN 91-72 10-805-3. 1887-1975. I. Musci. 1980. 60:- (hard-bound). I ISBN

6. Y. Dahlstrom-Ekbohm, Svensk miljovards- och omgiv­ 91-72 1 0-8 11-8 0 ningshygienlitteratur 1952-1971. Bibliografi och analys. 12. Sjo din , Index to di tribution maps of bryopbhytes A. 1975 . 20:-. ISBN 91-72 10-806- 1. 1887-1975. 11. Hepaticae. 1980. 40:- (hard-bound). I ISBN 91-7210-8 1 2-6.

Distributors: Svenska Vaxtgeografi ska Sallskapet, Box 559, S-75 1 22 Uppsala, Sweden

Almqvist & Wiksell International , Box 62, S- 101 20 Stockholm, Sweden.

Acta Phytogeogr. Suec. 68.