CC BY
0CYTO-ECOLOGIC PROBLEMS IN NORWEGIAN FLORA GROUPS
DISTRIBUTION AND SIGNIFICANCE OF POLYPLOIDY1
By GUNVOR KNABEN
botanical laboratory, university of oslo, norway
(Received April 6th, 1961)
I. INTRODUCTION
IN 1876 the late Axel Blytt, Professor at the University of Oslo, published a paper which created great interest. Its title was: "Essay on the immigration of the Norwegian flora during alternating rainy and dry periods." Norway, with its long coast and mountains occupying the interior, has a highly variable topography. The mountain ridges follow a direction parallel to the coast, and owing to the different degrees of influence of the sea, the climate varies very considerably from one part of the country to the other. The fjord landscapes in Western Norway have a pronounced oceanic climate with cool summers, mild winters, and a high rainfall, evenly distributed throughout the year. The valley landscapes in the "rain shadow" in Eastern Norway have warm and dry summers and cold winters with snow lying for several months. The climate in the mountains of Southern Norway with peaks up to 2,468 m. resembles that of the areas north of the Arctic Circle, in the counties of Nordland, Troms and Finmark.
The composition of the flora in the different climatic areas differs greatly. Blytt separated four groups of lowland plants, climatically conditioned, which he named after the post-glacial periods in which he thought they had immigrated to Norway: the Boreal, the Atlantic, the Sub-Boreal and the Sub-Atlantic groups. The Boreal and Sub-Boreal groups are today found only in Eastern Norway, the first as far west
1 The present work was delivered as a lecture on a topic, chosen by the author, at her disputation for the degree of Doctor of Philosophy at the University of Oslo, February 1960.
29 — Hereditas V
as in the inner fjord districts of Western Norway and Trondelag, the latter only in the south-eastern part of Eastern Norway. The Atlantic and the Sub-Atlantic groups are found only along the coasts in Western Norway with somewhat varying boundaries to the east and north. Today, we know much more about the migratory history of Norwegian flora after the Ice Age, than did Blytt, thanks to the work of our pollen analysts. Although now out of date Blytt's theories still contain a considerable amount of truth and are therefore referred to when phyto-geographical problems are discussed.
On reading the discussions of the varying significance of polyploidy in European and Arctic floras, the present author found it of interest to compare the frequency of polyploids in the different flora groups in Norway against the background of the migratory history of the flora after the retreat of the ice sheet, which covered most of the country in the last Ice Age, i.e. about 30,000 years ago.
II. OCCURRENCE OF POLYPLOIDY
The modern trend of phyto-geography, which can also be called cyto-ecology, is to investigate the variations and distributions of the higher plants in nature on the basis of the chromosomes, their numbers and structures. It can be dated back to the year 1917 when 0ivind Winge formulated his hypothesis of hybridization followed by polyploidy as a means of species formation in nature. Polyploid species are genetically related species with chromosome numbers forming multiple series with a common basic number. Winge's hypothesis has been verified by analyses of polyploid complexes of natural species and of crop plants, and further, experimentally, by synthesis of allopolyploids. Today we know from thousands of counts of chromosome numbers that polyploidy is widespread among the higher plants, though very irregularly distributed. It has had no significance in the evolution of new species in the Gymnospermae, but occurs frequently in a variety of groups within all categories of the Pteridophyta and in the Angiosper-mae. In the Pteridophyta it seems that polyploidy has reached a much higher level than in any other group of plants. This is proved by the high chromosome numbers of the most different polyploid complexes.
It seems probable that, with increasing age, genera with high chromosome numbers arise. High chromosome numbers have, therefore, been regarded as a sign of antiquity in a genus. This is particularly true of polyploids, which, having outlived their low-numbered ancestors, are
regarded as antique and as representing the last links of highly developed polyploid species. Many categories of the Pteridophyta can be mentioned as examples of such old polyploids. Psilotum (diploid with 2n = c. 100, tetraploid with 2n = c. 200) and Tmesipteris with 2n = over 400 are regarded as high-polyploid survivors of the primitive Psilo-tales. In Equisetum all species have 2n = 216, and Ophioglossum vul-gatum has 2n = 500 odd (Manton, 1950).
Genera of modern ferns, however, in spite of their high chromosome numbers, contain, like the Angiospermae of today, species with several related chromosome numbers. Especially in the Polypodiaceae, as studied by Manton in Great Britain, there exist genera rich in species with polyploid chromosome numbers dispersed throughout the flora in a variety of environments. They give the general impression, says Irene Manton, of having resulted from a recent wave of polyploidy which has affected the flora as a whole and led to a partial replacement of the low-numbered species by their high-numbered descendants, a process which, she thinks, is perhaps still continuing.
The chromosome numbers are generally much lower in the Angiospermae than in the ferns. They are most often lower than 50. They are seldom as high as 100, depending on the degree of polyploidy in the genus. This agrees with the younger age of the Angiospermae, the youngest of the three large groups of higher plants. The polyploid Angiosperm genera contain, typically, one or more polyploid species in contrasting environments. Polyploidy, however, is very irregularly distributed among genera and families, and the phenomenon seems to have been of varying significance in different geographical areas. Of this, Stebbins (1950) has given a survey.
On an average, the highest percentages of polyploids are found in perennial herbs, a smaller percentage in annuals, and the lowest percentage in woody plants. The following genera are almost completely devoid of polyploidy: Carpinus, Corylus, Castanea, Fagus, Quercus (28 diploid species, 1 tetraploid, the latter in Japan), Ulmus, Ficus, Morus, Ilex, which all belong to temperate regions. Arctic and alpine areas have few woody plants, and the following genera that occur represent polyploidy: Betula, AInus, Sorbus, Salix. Moreover, most of the dwarf scrubs dominating in these areas belong to polyploid genera.
There are also many variable genera of perennial herbs completely devoid of polyploidy. Epilobium can be mentioned as a good example of a Norwegian genus with many species, which, without any change in chromosome numbers, have adopted themselves to contrasting en-
vironments. Its 11 species, ail with 2n=36 chromosomes, have occupied the whole country, from sea level to the highest mountains; from south to north, a distance covering more than 13 degrees of latitude, that is, from temperate regions to the Arctic; and, further, from the moist oceanic climate in the west, to the warm dry valleys in the east, in all kinds of environments. Umbelliferae can be quoted as an example of a family with a small degree of polyploidy in all geographical areas.
III. SIGNIFICANCE OF POLYPLOIDY IN EUROPEAN
FLORAS
The cause of the apparently varying significance of polyploidy in different geographical areas has been vigorously discussed in our time. The varying occurrence in European floras was debated first for the obvious reason that it was these floras which are best known as to chromosome numbers. Tischler, supported by LOve, has tried to explain that there is an increase in the number of polyploid species from south to north in Europe, and that arctic areas and mountain floras have a relatively high content of polyploids, compared with that of lowland floras.
To demonstrate some of this variation a few of the percentages of polyploids found in various floras from Tischler and Love's comparisons have been listed in Table 1. They show that the proportion of polyploid species in Europe increases northwards in the following se-
TABLE 1. Contents of diploids and polyploids in European floras
( Angiospermae). After Tischler, 1955.
Diploids Polyploids D + P N
Cyclades 365 = 63 % 214 = 37 % 56 1186 53.5
Hungary 754 = 51.4 714 = 48.6 136 2039 78.7
Rumania 1117 -= 53.2 982 = 46.8 159 3365 67.1
Central Europe 1038 = 49.1 1076 = 50.9 210 2909 79.9
Schleswig-Holstein 438 = 45.5 525 = 54.5 79 1081 96.4
Great Britain 671 = 46.7 765 = 53.3 113 1778 87.1
Sweden 568 -= 43.1 753 = 56.9 106 1526 93.5
D+P = Number of species with intra-specific polyploidy N = Number of species in the flora
fc= Percentage of species known as to chromosome numbers
TABLE 2. Contents of diploids and polyploids in arctic floras ( Angiospermae).
Dicotylédones Monocotyledones D + P N o<
Dipl. Polypi. Dipl. Polypi.
Iceland Tischler, 1955 149 196 22 135 502 100
L. & L., 1956 = 43.2 % -56.8 % = 14 % -86%
Greenland 89 117 14 70 46 437 69
J., S. & W., 1958 -43.24 % = 56.8 % — 16.7 % = 83.3 %
Pearyland 13 23 3 25 64 100
Holmen, 1951 -56.5 % = 43.5 % = \1% = 88%
West Spitzbergen 28 52 1 41 135 82.1
Flovik, 1940 = 35 % = 65% = ÏÂ% = 97.6%
D + P = Number of species with intra-specific polyploidy
N = Number of species in the flora
% = Percentage of species known as to chromosome numbers from within the area
quence: the Cyclades, Hungary, Rumania, Central Europe, SchleswigHolstein, England, Sweden from 33.7 % in the south to 56.9 % in the north. In the Arctic, the percentages are high in all areas. In Iceland, and Greenland as a whole, there are 69.3 % polyploids in each country (Löve and Löve, 1956; Jorgensen, Sorensen and Westergaard, 1956). The numbers are higher in other, more restricted areas. In Pearyland in northernmost Greenland there are 81.5 % polyploids (Holmen, 1951), in West Spitzbergen 73.6 % (Flovik, 1940).
In Table 2 is listed the distribution of diploids and polyploids in the Monocotyledones and Dicotyledones of these latter areas. As to the arctic floras, it must be stressed that percentages mentioned above are only those calculated from chromosome numbers determined on material from within the areas. Besides the Greenland flora there is only one arctic, extra-European flora which has been investigated as to chromosome numbers, viz., the one from Kolguyev Island east of Cape Kanin, where there are 64 % polyploids according to Sokolovskaja and Strelkova (1941).
Exception has been taken to the percentages of polyploidy having been calculated for the European floras even if the chromosome numbers have not been determined on material from within the areas. It is only Tischler's investigations of the chromosome numbers of the flora of Schleswig-Holstein which have been based solely on material from
within the area. Tischler considered all species within the areas, with the exception of the microspecies of Rubus. Jorgensen, Sorensen and Westergaard omitted the Cyperaceae. Even if the aneuploid numbers of the genera originate through modified forms of polyploidy, it is difficult to separate the diploid and polyploid numbers. Species which show intra-specific polyploidy are omitted in all cases as we do not know the geographical distribution of their diploid and polyploid numbers. There are surprisingly many species showing intra-specific polyploidy. In the flora of Central Europe these amount to 8.0 % of the species.
The principal theories formulated to explain the varying significance of polyploidy in the different floras are as follows. Tischler, supported by A. and D. LOve, claims, in accordance with Hagerup's original hypothesis that polyploid species are more tolerant of extreme ecological conditions than their diploid relations. They attribute the higher percentages of polyploids in northern areas and in the mountains to the great hardiness of the polyploids. They maintain that the polyploids are more resistant to severe cold, for instance, than the diploids. A. and D. LOve explain the high proportion of polyploids in arctic areas in the same manner. They consider that the polyploids must have been more resistant than the diploids during the glaciations, and that for this reason more of the polyploids survived the Ice Age than did their low-numbered relations.
These theories have been criticized by many authors. It has been stressed that a complete change to new and severe conditions cannot possibly result from a mere increase in chromosome sets. It seems as though we cannot give a general explanation of the phenomenon, and that each flora has to be analysed as to its content of species. It has, for example, been experimentally shown that frost resistance of related species is not correlated to differences in chromosome numbers (Bow-den, 1940; Clausen, Keck and Hiesey, 1945).
Gustafsson (1948) has reported that certain ecologic plant groups have a greater number of polyploid species than others, this being connected with certain life forms disposing of polyploidy. As mentioned above, annuals of a complex are most often diploid, whereas, on an average, high percentages of polyploids are found in perennial herbs, and especially among those having a strong vegetative propagation (cf. also MUntzing, 1936).
The Mediterranean floras are characterized by many diploid annuals and by woody genera in which the faculty of polyploidy is absent. The
flora in Sicily has about 30 % polyploids, the same low percentage as in the Cyclades (Tischler, 1935). In the Arctic and in the mountain areas in Northern Europe, there are few annuals. There is a single annual in the Norwegian mountains, Koenigia islándico, which is tetra-ploid. These areas are dominated by grasses, or by grassy plants of the families Juncaceae and Cyperaceae. A large percentage of the grasses in the Arctic are perennials, with strong vegetative propagation by runners, vivipary or apomictic seed formation, a life form which almost always is connected with polyploidy. "Grasses" as a collective group can comprise more than V3 of all plant species in arctic areas. On Scandinavian mountains a grass region exists at greater heights than the region with dwarf scrubs, Betula nana and mountain willows. More than 90 Jo of these grasses are polyploids, which involves a rise in the percentage of polyploids in the area.
Stebbins claims that the high percentage of polyploidy in arctic and alpine areas is, in the main, dependent on the dominance of grasses in the vegetation cover of these areas. A fact supporting this view can be mentioned here. According to Tischler's "Chromosome Atlas" the grasses only amount to 14 % of the inventory of lowland species in Central Europe, and of these many are diploids. But the problem is not as simple as it might appear. If we analyse the floras along other lines, analysing their inventory of polyploid genera on, for instance, the basis of historical facts it can be proved that other factors must be taken into account. Or again, the question of how to explain that 56.8 % of the Greenland Dicotyledones are polyploid, is still unanswered. Originally, the intention of Jorgensen, Sorensen and Westergaard in their cytological studies of the Greenland flora was to contribute to the general discussion of arctic polyploidy, but this they refrained from doing, chiefly because so few chromosome numbers are known from arctic America and Siberia.
IV. POLYPLOIDY IN NORWEGIAN FLORA GROUPS
To procure a preliminary orientation in the distribution of polyploid species in Norway the present author has ascertained the percentages of polyploids in the different climatically conditioned flora groups in this country (Tables 3 and 4). But she did not think it of interest to consider the genera which have hitherto been known to include almost only polyploids. Therefore, the Pteridophyta and the apomicts belonging to the family Rosaceae and to the genera Taraxacum and Hieracium
TABLE 3. Grouping of the Norwegian flora.
Groups Distribution
Larger groups:
1. Lowland Lowland plants occurring from sea level up into the coniferous forest belt.
2. Mountains Mountain plants occurring in the sub-alpine and alpine regions. Ubiquitous or disjunct distribution.
3. Whole country Over the whole country, from sea level and up among the mountains.
Narrower groups:
4. Bi- or unicentric species Mountain plants with disjunct distribution.
Lowland plants:
5. Nordland Occurring in Southern Norway with northern limit in Nordland. 66° N. Lat.
6. Trondelag Occurring in Southern Norway with northern limit in Trondelag. 64° N. Lat.
7. S. of Dovre Mts. Occurring in Southern Norway, but seldom north of the Dovre mountains. 62° N. Lat.
8. Boreal group (Blytt) South-Eastern Norway and inner fjord districts in Western Norway and Trondelag.
9. Sub-Boreal group (Blytt) Only in South-Eastern Norway.
10. Southern species Occurring in southernmost Norway, south of the Oslo Fjord. 60° N. Lat.
Coastal plants:
11. F^egbi's group Coastal plants in Norway.
12. Oceanic spp. Oceanic species also farther south in Europe.
13. Bromus Benekenii group Oceanic accentuated, but somewhat restricted in distribution.
14. Shore plants Halophilous or shore plants.
have been omitted. Further, all apomictic grasses and the Cyperaceae have also been omitted.
However, we must fully realize the hazards involved in finding the percentages of polyploids also in other Norwegian plant groups, since almost all chromosome numbers have been determined on material from outside Norway. The latest cytological investigations from Central Europe and England have, namely, revealed that a large number of the most common species show intra-specific polyploidy. All species show-
TABLE 4. Content of diploids and polyploids in Norwegian
flora groups.
Diploid Polyploid hr. numb, inknown v'umb. of species 'Ufoa/li!
Groups £ » ■a ^ .a s* a 3 & ¿. .s So s 3 c. % D + P Nnmber c species Number of species
2i * Y. " u D P D + P
Larger groups:
1. Lowland 245 77.3 72 22.7 41 17 Weeds
2. Mountains 96 65.8 50 34.3 7 9 omitted
3. Whole country 67 53.2 59 46.8 9 —
Narrower groups:
4. Bi- or unicentric spp. 36 50.0 36 50.0 3 7 0 0 0
Lowland:
5. Nordland 26 63.4 15 36.6 17 2 8 2 3
6. Trondelag 29 80.6 7 19.4 11 3 4 4 3
7. South of Dovre Mts. 23 76.7 7 23.3 3 2 3 1 1
8. Boreal group 39 90.7 4 9.3 6 4 2 3 0
9. Sub-Boreal group 42 93.3 3 6.7 8 3 2 2 0
10. Southern spp. 18 65.0 10 35.0 4 3 0 0 1
Coastal plants:
11. F/EGRi's group 12. Oceanic spp. 13. B. Benekertii group 69 23 15 75.0 57.5 78.0 23 17 4 25.0 42.5 22.0 13 2 1 4 3 1 Weeds omitted
14. Shore plants 26 57.8 19 42.2 1 2
D = Diploids, P = Polyploids, D + P = species with intra-specific polyploidy.
ing intra-specific polyploidy outside Norway, but which have not been investigated as to chromosome conditions with us, have been omitted in the calculations as we do not know which number, diploid or polyploid, occurs here. However, some uncertainty is still attached to the percentages found, since we do not definitely know whether the remaining species comprise one chromosome race only.
As a matter of course, the percentages found by the present author are not directly comparable with the ones determined from other geographical regions, as these are based on another selection of the species concerned.
Discussing the occurrence of polyploidy in the different climatic areas in Norway on the basis of the percentages listed in Table 4, even if some uncertainty is attached to the numbers, the present author has
come to conclusions that bring to light new points of view, which may be of interest.
Tables 3 and 4 contain two different groupings of our flora. Three larger and 11 narrower groups have been considered. The three larger groups are set up against each other according to the vertical distribution in Norway of the species included, irrespective of their total area. One of these, group 1, includes all lowland species growing from sea level up into the coniferous forest belt, some of them from south to north, but none ascending into the mountain birch belt. Another group (2) includes species occurring in the mountain birch woods or in the tree bare, alpine region, as well as in the arctic north. There is 34 % polyploids in this wide group of arctic-alpine species, that is, more polyploids than in the group (1), comprising the lowland plants, which has 22.71 % polyploids. The percentage of polyploids is again higher (46.8 %) in the group (2) of species growing over the whole country at all heights, from south to north.
Below, we will chiefly consider the narrower groups of species with more restricted boundaries.
1. The arctic-alpine flora groups
The percentages of polyploids differ in the two groups of arctic-alpine species, considered. The larger group 2, mentioned above, including all our arctic-alpine species, both ubiquists occurring throughout the whole mountain range, from south northwards to Finmark, as well as the species with disjunct distribution, has a lower number of polyploids than the narrower group 4, 34 % as against 50 % in the latter. The narrow group contains only our rare species, viz. the ones which are said to have bicentric or unicentric distribution because their stations fall within two separate areas, one, smaller, in the mountains in Southern Norway, the other, larger in extension, north of the Arctic Circle (Fig. 1).
As is well known it is especially the latter group of species, with disjunct distribution, which has been discussed in connection with the theory of glacial survival in ice-free areas in Norway, either along the More coast outside the southern inland "island" of rare species, or at different places along the coast north of the Arctic Circle. If the theorj' of glacial survival is correct, this group of rare arctic-alpine species represents the most ancient flora element in Scandinavia.
The polyploids amount to 50 % in the group (4) of rare mountain
species, which is a high percentage considering how the percentages were computed, and the question is in what manner this can be explained.
We must make comparisons with other alpine floras. In the course of the last 10 years many polyploid complexes belonging to the alpine
flora of Central and Southern Europe have been investigated from a cyto-ecologic point of view, and many interesting facts MVS been revealed. During the Quaternary period the ice sheet covered the mountains of Central and Southern Europe. However, it has been proved that in the peripheric Alps, which lay within the border of the ice sheet but were less glaciated than the high, central Alps, there grow today many species with restricted ecology and distribution; and the interesting point is that they are diploid. The origin of these species has been discussed. They have been considered as ancient types, with roots going directly back into the Tertiary flora of the Alps. It is thought that this Tertiary alpine flora must have developed from the local flora when conditions changed at the end of the Tertiary period, adapting themselves to the climate of greater heights, without any change in chromosome numbers. It is also supposed that the lowland types, on the plains, became extinct during the glaciations, whereas the Tertiary alpine types survived on nunataks, and have moved very little since the ice melted (cf. Favarger, 1958).
The occurrence of polyploids in the flora of the Alps as a whole has been discussed in relation to arctic polyploidy. The percentage of polyploids amounts to 56.2 in the Alps, a number which is much lower than the ones found in arctic floras (Table 2). The former number has been calculated in the same way as the latter, all genera being taken into consideration. Favarger maintains that the reason why the percentage of polyploids is lower in the Alps than in the Arctic is chiefly because the ancient element of Tertiary plants includes so many diploids which are missing in the north. He considers that the high number of these diploids lowers the percentage of polyploids in the Alps. Both Favarger and MerxmOller emphasize that the central high Alps, which were more glaciated than the peripheric Alps, comprise a higher proportion of polyploids than the lower, peripheric Alps. They discuss in that connection Love's theory of arctic polyploidy. As mentioned above, this theory suggests that the polyploids could stand the hard conditions during the Quaternary glaciations better than the diploids, and that, therefore, more polyploids than diploids survived the Ice Age.
The investigations — now in progress in Switzerland and Austria (Favarger and his collaborators, Ehrendorfer) — of the detailed geographical distribution of the varying chromosome races of species showing intra-specific polyploidy in the Alps and the surrounding lowland will certainly clear up many problems concerning the recent
history of the flora of the Alps. The number of species with varying chromosome numbers have appeared to be surprisingly high, also in species in which the morphological differentiation is not pronounced, even if the ecological variation is clear, the chromosome races belonging to different milieus or climatic areas.
The large content of polyploids in the High Alps is not first and foremost attributed to the strong resistance of polyploids to severe conditions, but rather to the longer distances the plants have traversed before reaching the High Alps. With MerxmUller we can say that a "Sippengliederung" has taken place contemporaneously with an "Arealbildung". It seems as though the diploid hibernated species have been the progenitors of new polyploids, which occupied the areas laid bare after the melting of the ice.
Favarger points out that many of the immigrants to the Alps which originally came from the Himalayas and other Asiatic mountain ranges belong to the polyploid element in the Alps. As example he mentions species of the genera Androsacae, Primula and Gentiana. During the wanderings of plants over smaller or greater distances the ecologic and climatic conditions change, and miscellaneous populations arise. All these new conditions offer chances for new polyploids to appear and to establish themselves in the floras.
Sokolovskaja and Strelkova's analyses of the alpine flora in the Near East coincide with these points of view. Especially is this true of the flora of the Caucasian mountains. Here they found a percentage of polyploids amounting to 50 only, calculated according to the old methods, all genera being taken into consideration. The authors explain this relatively low occurrence of polyploids in accordance with Favarger. The last elevation of the mountains in Caucasia, to twice the height, took place during the Quaternary period, and they maintain that a large part of the Caucasian alpine flora of today originates directly from the local Tertiary flora at lower heights. This local flora moved with the elevation of the mountains, and many species adjusted themselves to life at greater heights. The Russian authors think that these species represent the large number of diploid species belonging to the alpine flora of today, whereas most of the species which wandered long distances to reach Caucasia are polyploid. It may be mentioned in this connection that Skottsberg (1939), similarly connects the high percentage of polyploids in the Hawaiian Islands with the long migratory distances. Many species in these islands originate from antarctic areas.
Generally, we must think that isolated, polyploid species with high chromosome numbers, which grow far removed from the area where several of their ancestors with lower chromosome numbers still grow, must be of advanced age. They must certainly be much older than polyploid species with chromosome numbers which are topping a complex, but which still grow in the near neighbourhood of, or together with, their ancestors. This particularly applies to polyploids growing in areas which have been subject to heavy climatic oscillations, such as Europe during Quaternary and post-glacial times.
Returning to the Arctic, we shall study a polyploid complex represented here, with a long series of high chromosome numbers, such as the Papaver radicatum complex. It has a series of numbers ranging from 14 to 84, with basic number 7. It is striking that only the highest chromosome numbers 56, 70 and 84 belong to the European sector of the Arctic, whereas also the lower numbers are found in America, preferably in the western part, and in Asia. There is a marked differentiation as to phenotype and chromosome structure at each polyploid chromosome level, which shows that the polyploids must be of great age in the European area (Knaben, 1959). We must suppose that they migrated there in pre-Quaternary or inter-glacial times, and that the populations became split during the glaciations.
This part of the Arctic owns other genera or subgenera with isolated polyploids, which are now far removed from their low-numbered relations, or without relations in other areas: Arenaria ciliata subsp.pseudo-frigida, A. humifusa and A. norvegica, Braya linearis and B. purpura-scens, Ledum palustre, Primula scandinavica and P. stricta, Nigritella nigra, Platanthera oligantha, Saxifraga hieraciifolia, etc. Similar to the high polyploid Papavers, these polyploids may have migrated to this part of the Arctic in pre-glacial or inter-glacial times.
The old arctic-alpine elements of the Scandinavian flora, which most possibly are glacial survivors, have, like the old diploid species in the Alps, spread very little in post-glacial times. They seem to grow nowadays not very far from where they hibernated, either in ice-free coastal strips or on nunataks farther inland. It also seems as though, like the diploid species in the Alps, they must have wandered in groups, seeing that today they are found isolated on the same mountains. Strikingly, all, or most of them, are little variable. This is the case with regard to both the diploid and polyploid members. All these conditions suggest that both categories of plants, diploids and polyploids, like the diploids in the Alps, are of great age within the Scandinavian area, indicating
that polyploid species of an old flora element have not the same faculty of colonizing and spreading as a newly arisen, "raw" polyploid type.
2. The lowland flora groups
Considerable importance has been attached to the great faculty of newly arisen polyploids to colonize. Analyses show that they must have far greater ability to occupy new areas than their diploid relations. Their new, unstabilized gene combinations have much selection value on virginal soils. They evidently have less ability to penetrate into a ripe vegetation cover.
In late glacial and post-glacial times there were in Europe several categories of new soils ready for colonization. Firstly, there were the areas laid bare after the melting of the ice, both in the lowlands and in the mountains. Later, there were the shores along the coasts and along the large inland lakes, which rose above sea level, subsequent to the general elevation of the countries in post-glacial times.
It is supposed that at the beginning of the Quaternary period Europe, north of the Alps, had a flora of about the same kind as today. This flora hibernated on ice-free refuges to the north-west, south and east of the ice cover, and migrated back again, when the climate improved, shortly after the retreat of the ice.
We shall consider briefly how the different climatically conditioned flora groups of Europe vary in distribution according to the variation in climate today. The variation today, in the climate in Europe changing gradually northwards, concerns firstly the light conditions. The change in day and night lengths during the year differs very much from south to north. The temperatures also vary with the latitude, but the climate in Europe, both with regard to temperature and moisture, depends more upon the situation of the areas in relation to the large mountain ranges and their direction, and then, on the distance of the areas from the Atlantic Ocean. In Western and Central Europe the conditions change along a west—east gradient.
On Troll's map (Fig. 2) three zones are marked in Western Europe which are influenced by the Atlantic Ocean with decreasing effect eastwards: the eu-Atlantic zone along the Atlantic coast, the sub-Atlantic farther inland, and the eury-Atlantic zone penetrating in between the sub-arctic coniferous tree region to the north-east and the arid region with steppe vegetation to the south-east. The oceanic flora elements of these three zones are in Norway concentrated in the coastal region of
Fig. 2. Reprint of Troll's map of floristic regions in Europe. 1: Eu-Atlantic, 2: Sub-Atlantic, 3: Eury-Atlantic, 4: Boreal and Arctic, 5: Semi-arid regions.
Western Norway. They are characterized by a northern limit going farther to the north in Western Norway than in the warmer and drier districts in Eastern Norway. The latter districts for their part have — as character plants — an element of thermophilous and xerophilous species growing preferably on edaphically good localities. Many of them show connections with the arid steppe vegetation of south-eastern Europe.
We shall not consider in detail the distribution of the climatically conditioned plant groups elsewhere in Europe, but only mention that the sub-oceanic plants in the Mediterranean climb up on the mountain sides; these therefore have been noted as sub-oceanic montane species.
It seems obvious when investigating occurrence and significance of polyploidy in Europe to base this on our knowledge of the distribution of the climatically conditioned plant groups, the recent wanderings of which have been discussed for a century, i.e. since Axel Blytt (1876)
formulated his theories of the immigration of the Norwegian plant groups in alternating dry and rainy periods in post-glacial times.
I have computed the percentages of polyploid species by different groupings of our lowland species.1 We see that the group of lowland species has a very low proportion of polyploid species, either their northern limit falls in Nordland at about the Arctic Circle, in Trondelag, or we consider only the species in South-Eastern Norway, south of the Dovre range (Groups 5—9, Tables 3 and 4). The two latter groups of species (8 and 9) correspond to Blytt's boreal and sub-boreal plants.
The dominance of diploid species in the southern and south-eastern flora elements disagrees with the theory that it is the autochtone floras which include most diploids. Analysing the contents of species it appears that these groups comprise many genera with little or no faculty of polyploidy. On the whole, 35 of the south-eastern species belong to different genera with solely diploid numbers. Of these, 14 genera are monotypic and not variable. It is remarkable that all the diploid species of the south-eastern group belong to genera with the northern limit in Southern Norway, and that this is the case with many of the polyploid members of the groups, too. It is not only the species, but the genera to which they belong that have their northernmost findings in Europe in Southern Norway. The south-eastern species represent a characteristic, alien element in our flora, behaving as if they were relics from a warmer period. Now they occur in the warmest and driest parts of the country. The many diploids must have roots back in the European Tertiary flora. They adjusted themselves during the changing conditions in the late Tertiary and Quaternary periods at the outskirts of the ice sheet, and after the retreat of the ice, in a favourable climatic period, wandered northwards again, probably together with other steppe plants. Here they found environments of about the same type as those to which they had become accustomed, and that is where they grow today.
It was mentioned above that recent investigations of the glacial flora in Central Europe have revealed that the Quaternary climatic oscillations were followed by a vivid formation of polyploids within many genera after the melting of the ice. An obvious question here is whether the Norwegian flora which came from the south is characterized by polyploid types of Quaternary or perhaps post-glacial origin. The
1 The grouping of the species is based on the maps of distribution in HultEn's Flora Atlas and on the lists of species worked out for the mapping of the Norwegian flora by F>egri, Gj^revoll, Lid and Nordhagen.
30 — Heredltas 47
answer must be that we certainly have such types. Some of them have arisen farther south in Europe, and reached us in company with ecologically similar, older species, whereas others have arisen when the boreal — sub-boreal and arctic — sub-arctic flora groups met each other in Southern Scandinavia during migratory times.
We will study some examples. Most Galium species show intra-specific polyploidy in Europe. Two species belonging to our southern lowland group, Galium mollugo and G. verum, came from South-Eastern Europe. Fagerlind (1937) dealt with these two species. In Norway they grow in somewhat changing environments on dry slopes and on crags. In South-Eastern Europe both show intra-specific polyploidy.
Fagerlind investigated the geographical distribution of the chromosome numbers. He found that the diploid types grow only in SouthEastern Europe, in some places together with the tetraploids, whereas he only found tetraploids to the north of the Balkans and the Carpathians. Populations from about 70 localities were investigated, from the Balkans and Transcaucasia, northwards to southernmost England and the Stockholm area.
Fig. i. The distribution of Saxífraga adscendens (•) and S. tridactylites (O) in Fenno-Scandia. From Knaben (1954).
It must be supposed that probably the polyploid types arose during or after the Quaternary period, either from inter-variatal hybrids, or through a purely quantitative increase in chromosome numbers. Judged from the distribution of the chromosome numbers it seems that the diploids have not been able to spread northwards, whereas the tetra-
ploids have colonized the land laid bare after the melting of the ice. In Norway we will certainly find only tetraploids of both species, which, in accordance with modern theories, we can consider as relatively young categories of plants. The two Galium species have a little wider distribution in Southern Norway than the species of Blytt's boreal and sub-boreal groups. In South-Eastern Europe, diploid and tetraploid mollugo are ecologically very alike, growing among each other in the localities, whereas diploid and tetraploid verum are found in ecologically different niches. The two species must have reached Norway in company with the south-eastern, temperate flora element, which migrated into Scandinavia shortly after the retreat of the ice when the climate suddenly improved. Sub-fossil deposits in southernmost Sweden show that a sub-arctic flora with pines overtook the arctic Dryas flora of late glacial times, and that the temperate flora from the south-east followed the sub-arctic flora before the spruce invaded the country on a broad front.
In Scandinavia we have a polyploid species in an area where the south-eastern flora must have met the sub-arctic, and perhaps also some arctic plants which had followed the retreating ice. This species, Saxífraga osloensis, is regarded as an amphidiploid, arisen after chromosome doubling in the hybrid 5. adscendensXtridactylites (Knaben, 1954). S. osloensis (Fig. 3) grows across Southern Sweden, preferably near the great lakes. In Norway it is found only in a restricted area at the head of the Oslofjord, and almost only on land laid open for colonization after the elevation of the country at the end of the post-glacial warm period (Blytt's Sub-Boreal period). It grows below the marine limit in Oslo and its surroundings. The supposed ancestors: sub-arctic adscendens and temperate tridactylites (Fig. 4) still grow within the area of S. osloensis. S. adscendens is rare in this lowland area, where it occurs at its southern limit in Scandinavia, whereas tridactylites grows abundantly and in company with osloensis in the localities. S. tridactylites belongs to our boreal group of plants with connections to the south-east in Europe.
3. The oceanic species
Before considering other possible sources of our eastern polyploid lowland species, we shall consider briefly the occurrence of polyploids in the coastal flora. F^egri's group of coastal plants in Table 4 is not very appropriate for our purpose. It comprises the species with coastal
occurrences which F^egri (1960) treats in the first part of Norway's Flora Atlas, just published. All Norwegian coastal plants are included in this group even if the species have a rather eastern distribution farther south in Europe. The present author has separated two narrower groups from this larger group. One includes coastal plants along our western coast with oceanic distribution also farther south in Europe, the other includes the species which have a similar distribution in Norway to that of Bromus Benekenii (cf. below p. 472).
There are 42.5 % polyploids in the group of pronounced oceanic species of Western Norway. It is a high percentage, the mode of calculation taken into account. In accordance with the points of view maintained above, the question is whether this high percentage can be explained in accordance with modern theories. The moist coasts along the Atlantic Ocean, with cool summers and mild winters and heavy rainfall, formed new land with new kinds of localities, which became ready for colonization during the post-glacial rainy periods. This new land differs in our time very much from the soils and climatic conditions to which the late Tertiary flora had adapted itself during the Quaternary period.
We shall notice that in the oceanic element there are as many as 12 of the species belonging to diploid or almost purely diploid genera, lacking the faculty of polyploidy, which have spread to the moist coasts, adapting themselves without any change in chromosome numbers. It is further noticeable that all the oceanic polyploids belong to variable genera rich in differing taxa, with series of related chromosome numbers, all of which occur today within the oceanic-influenced areas of Scandinavia or farther south in Western Europe.
These conditions are suggestive of the occurrence within the oceanic zones of relatively young polyploid species that arose in post-glacial times. If we wish to study a polyploid complex in the Atlantic zones of Europe, the Luzula campestris-multiflora complex is a good example. It comprises the following series of chromosome numbers: 12, 24, 36, and 48 distributed in Western Europe today.
The species, L. congesta, with 2n = 48, which is the highest polyploid of the complex, is eu-Atlantic, restricted to the outermost coastal strip from Western Norway to Spain. It belongs to the strongly humified zone developed in post-glacial times. It may be a polyploid of recent origin. However, as shown by Hedda Nordenski<}ld, all the Luzulas are in reality so variable in Western Europe today, with series of related types varying as to chromosome numbers, that we ought to discuss
whether "a recent wave of polyploidy" has not taken place within the genus, to use Irene Manton's mode of expression.
A pair of species, Cardamine hirsuta and C. flexuosa, may be mentioned in this connection. The first is diploid, the latter tetraploid. Both species are oceanic-accentuated as to distribution, but only the tetraploid is regarded as distinct oceanic. The diploid, hirsuta, is included among the species with distribution in Norway similar to Bromus Bene-kenii (F^egri, 1960). The Bromus Benekenii group has a lower percentage of polyploids than the more pronounced oceanic species: 16.2 % as against 42.5 % in the latter. The Bromus Benekenii group of species does not, like the distinct oceanic, reach the outermost coastal strip, but is confined edaphically to better localities farther inland. It is strange that 15 of the 18 species of the Benekenii group are diploids. The group includes 20 species, but two are not known as to chromosome numbers. From a cyto-ecologic point of view we could perhaps also include one of the three polyploids among the group of diploids, viz. Hedera helix. Certainly, it is a tetraploid species in the genus; in Australia there is a species H. australiana with 2n = 24 chromosomes as against 2n = 48 in H. helix. But it has the lowest chromosome number of the European species of Hedera, and it has the widest distribution across the European continent. On its borders to the west and south there are octoploids, H. hibernica in Ireland and H. canariensis in the Mediterranean and the Canaries, both with 2n = 96. To the east a species with 2n = 192 is known, viz. H. colchica in Caucasia and Persia (Jacobsen, 1954). Accordingly, provided that H. helix has 2n = 48 throughout Europe it may be a tetraploid of advanced age in Europe, and probably the progenitor of relatively younger polyploids on the borders of its area.
The question is whether the restricted distribution of the Benekenii group in Western Norway is connected with old, diploid or higher, chromosome numbers. The species are no longer sufficiently variable and cannot adjust themselves to new conditions. It is often seen that species on the border of their area are restricted to the edaphically better localities. The Benekenii species were probably crowded out and pressed back to their localities of today when the soils in the outermost coastal region were heavily leached out in the cool climate with high rainfall which has favoured the formation of acid raw humus.
If it be correct that the distribution of the Benekenii species can today be explained by supposing that they were not able to adjust themselves to the new conditions, the distribution of the two species Cardamine
hirsuta and flexuosa can be understood. C. hirsuta, itself a diploid, can then be said to have a tetraploid descendant, C. flexuosa, as a substitute in the pronounced oceanic element. It is evident that the tetraploid in this case has been able to penetrate into new and strange localities, whereas the diploid has had more restricted possibilities. In Southern Europe C. flexuosa is oceanic-montane, whereas hirsuta is more restricted.
V. CONCLUSIONS
Summing up the results hitherto attained of the analyses of the Norwegian flora groups, it seems clear that in order to provide a logical explanation of the variable distribution of the polyploid species, the floras have to be analysed separately as to content of polyploid complexes in relation to various factors. Especially the Quaternary history of the plants, the stations where they hibernated, their migratory ways and the long distances they may have travelled, in addition to their adaptability to climatic oscillations have been factors of great importance.
We must be aware of the fact that our oceanic flora element may include recently arisen polyploids, just as our eastern flora or the whole lowland flora, probably comprises such polyploids. This being so, it seems that many phyto-geographical problems concerning the Norwegian flora, which have been discussed for a long time but have hitherto been difficult to elucidate, can now be given a logical explanation.
Finally, in support of these points of view, some of the results of the above-mentioned investigations during the last decade of the chromosome conditions of the glacial flora of Central Europe in relation to distribution will be briefly cited, especially those of interest for our lowland flora.
Surprisingly enough, it would appear that some of our lowland species originate from the ancient flora of the Alps, viz., from the local age-old element of diploid species. These diploid species have, through intra- or inter-specific polyploidy, been the source of new types, which wandered down into the lowland and occupied the areas north of the Alps. As Favarger has stressed, for alpine species the life of lower heights, on the plains, represented new conditions. It seems now to be proved that many polyploid races of a series of alpine species have spread down into the lowland, unlike the diploids themselves, which still grow where they hibernated in the peripheric Alps.
It has long been known that the Scandinavian arctic-alpine flora is related to the flora in northern arctic areas, and that it has less connection with the flora of the Alps. Against this background it is surprising to see that the lowland flora has been recruited from the flora in the Alps. SOllner and Favarger, in 1950 and 1954, were the first to assert that diploid alpine species have given birth to polyploid lowland races. It was races of Cerastium arvense, Lotus species and of the Anthoxan-thum odoratum-alpinum complex that they discussed.
As to Cerastium arvense, we no longer consider, as Hegi did, that its subsp. strictum, with restricted distribution, is an alpine ecotype, separated from subsp. commune that grows abundantly in the lowland north of the Alps. On the contrary, we maintain that the polyploid, widely dispersed subsp. commune originates from the alpine type. In 1959 Favarger showed that, within Aster alpinus and Chrysanthemum leucanthemum polyploid races had similarly arisen through intra-specific polyploidy, and that the diploid ancestors were found in the relic alpine flora in the refuge area of the Alps. The latter, of interest to our flora, has diploid, tetraploid, hexaploid and octoploid races in the European glacial area. However, it must be noted that besides the alpine diploid race studied by Duckert and Favarger, occurring today on the western brim of the Alps, there is also a diploid growing at the present time on the plains in Central Europe northwards to Denmark and Southern Sweden. This diploid race may have hibernated in the refuge area in the north-western part of the Continent or in West Jutland (Bocher and Larsen, 1957).
Bocher and Larsen show that the specific name Leucanthemum must probably be retained for the diploid race, as the specimen belonging to the Linnean Herbarium of the Linnean Society in London accords as to phenotype with a diploid specimen. The Leucanthemum complex has not yet been unravelled. In Norway we probably have one or more of the polyploid races.
Ehrendorfer's investigations of the Galium pumilum complex are of interest for the present analyses. In this complex it is also the diploid which has restricted, and in this case, disjunct, distribution in the peripheric alpine nunatak-area, whereas the tetraploid, hexaploid and octoploid races have a much wider distribution in the lower parts of Central Europe. The distribution of the polyploids solely in the once glaciated areas of Europe suggests also in this case that the race formation has taken place during or after the Quaternary climatic oscillations.
In Norway there are two races of Galium pumilum, one in the north,
on the island of Vega in Nordland (subsp. Normanii O. Dahl), the second in a small area in South-Eastern Norway. Their chromosome numbers are unknown, but probably they are polyploid. The same reasoning can be applied to Galium palustre and G. uliginosum, which both show intra-specific polyploidy. In England, Hancock (1942) has found both diploid and polyploid races.
The lowland localities in Norway must be regarded as terminals for the wanderings of the species occurring here after the Ice Age. It would be of great interest to ascertain whether the genera contain relatively more polyploids and less diploids in our area than farther south and east, as a result of a connection between increase in polyploidy and long distances traversed, and between polyploidy and the faculty of colonizing. For the present we can only conclude by stating that in the lowlands, both within the eastern xerophile flora and within the oceanic group, there exists, besides an ancient diploid element, a polyploid element which may contain species of recent origin.
Similarly, the arctic-alpine flora contains a multitude of still unsolved problems. The Empetrum nigrum problem must be reconsidered owing to Favarger and his collaborators having found the tetraploid hermaphroditum in the Alps. Previously, it was only known in northern areas. The diploid E. nigrum grows all over the lower parts of Europe, where hermaphroditum is missing or rare. The latter is an arctic-alpine species, probably belonging to an ancient flora element. Nor is the Antoxanthum problem easily explained. In contrast to the Empetra, the diploid Anthoxanthum races are arctic-alpine, with the tetraploids in the European lowland. The history of these two polyploid complexes is, perhaps, not easily unravelled.
In addition, we have all the problems concerning common species, with varying ecology, or with no clear connections with other areas. Some of them have been subjected to the searchlight of our phyto-geographers, such as the mountain birch met with in a belt above the coniferous forests of spruce and pine. On Scandinavian mountains and in the European Arctic it is the mountain birch that dominates the forests in a belt at the upper forest limit. The mountain birch forest has been characterized as an Atlantic phenomenon from post-glacial wet periods. It constitutes a vegetation type without relations far back in time or with other areas, the origin of which has been difficult to explain.
The mountain birch is not only ecologically distinctive. It is morphologically fairly well separated from the birch at lower heights. Scandi-
navian botanists have considered it as specifically distinct. It has been named Betula tortuosa Ledeb., to distinguish it from B. pubescens coll. at lower heights. Both B. tortuosa and B. pubescens are polyploid, with the same number of 2n = 56 chromosomes. B. verrucosa Ehrh. in regio silvatica and B. nana L. in regio alpina are both diploid, with 2n = 28.
In our country, up to the present time, problems of plant distribution of this nature have been approached according to the descriptive methods. Even if this older method is far from being exhausted, the time is perhaps now ripe for attacking the problems according to modern methods, basing the work on cyto genetic characters. Hitherto, it has been first and foremost the aristocrats in our flora, the rare mountain species and the ecologically clearly limited groups of species, which have been analysed as to their migratory history and origin, and also from a chorological point of view. An intensive study with chromosome counts of the more common, trivial plants will now, it seems, certainly assist in throwing light upon problems concerned with the origin of our flora as a whole.
The present account cannot be terminated without citing Stern's (1946) work on Paeonia, a widespread genus in the northern hemisphere. Stern's conclusions with regard to the distribution and significance of polyploidy in this genus accords with the present author's views. Stern thinks that the distribution of the species today must be due to changes which took place in the Ice Age. In Europe, the oldest species, that is the diploid species, are only found in Portugal, in the islands of the Mediterranean, in the Crimea and Volga regions, and in a small area in the Caucasus. The distribution of these species suggests that they are pre-glacial relics from which tetraploid species arose. These latter, new species, have proved themselves better adaptable to post-glacial conditions than the old diploid species. They have spread to areas which were covered by the ice sheet and laid bare after the melting of the ice. Here their diploid near relatives do not grow. An example is P. daurica, which occurs in the Crimea and the surrounding neighbourhood, whereas allied tetraploid species, P. mascuht and P. ba-natica, stretch westwards across Europe, through Hungary and Germany to France.
Manchuria and China were not covered by the ice sheet. In these countries nine paeonie species are found; they are all diploid and are considered to be pre-glacial relics. Likewise, in California in North America, diploid paeonie species are found in areas which were not glaciated. They also are considered as ancient relics which have de-
scended from the same source as the species in Western China. It is curious, Stern says, that in Asia and America all the species are diploid, with only one known tetraploid, whereas in Europe there are more than twice as many tetraploid species as diploid species.
In the genus Pyrola there is only one known tetraploid species, viz. P. media. It is supposed to have arisen as an amphidiploid from the cross between the two diploid species P. minor and P. rotundifolia (Hagerup, 1941; Knaben, 1943). It grows today, similar to the tetraploid paeonie species, in once glaciated areas, viz., in Europe, North Asia and Asia Minor, but not in Central Asia, Manchuria and Japan. The family Pyrolaceae is thought to have had its origin in the latter areas, which appear to be rich in Pyrola species. P. minor and P. rotundifolia must have emanated from there. The distribution of P. media has hitherto been difficult to explain. If the species is considered to be of glacial or post-glacial origin, arisen during periods of upheavals caused by changes during the Ice Age, then its distribution can the more easily be understood.
Literature cited
Blytt, A. 1876. Essay on the immigration of the Norwegian flora during alternating
rainy and dry periods. — Chria. pp. 1—89. Bowden, M. 1940. Diploidy, polyploidy, and winter hardiness relationships in the
flowering plants. — Am. J. Bot. 27: 357—371. b0cher, T. W. and LArsen, K. 1957. Cytotaxonomical studies in the Chrysanthemum
Leucanthemum Complex. — Watsonia i: 11—16. Clausen, J., Keck, 1). D. and Hiesey, W. M. 1945. Experimental studies on the nature of species. II. Plant evolution through amphiploidy and autoploidy, with examples from the Madiinae. — Carnegie Inst, of Wash. Puhl. 504: 1—174. Contandriopoulos, J. and Favarger, C. 1959. Existence de races chromosomiques chez Chrysanthemum alpinum L.; leur repartition dans les Alpes. — Rev. générale de Bot. 66: 1—17. Darlington, C. D. and Wylie, A. P. 1950. Chromosome Atlas of Flowering Plants. — London. 519 pp.
Duckert, M.-M. et Favarger, C. 1956. Sur l'existence dans le Juron d'une forme diploide de Chrysanthemum Leucanthemum L. — Bull. Soc. bot. suisse 66: 134—146.
Ehrendorfer, F. 1949. Zur Phylogenie der Gattung Galium. I. Polyploidy und geographisch-ökologische Einheiten der Gruppe des Galium pumilum Murray (Sect. Leptogalium Lange sensu Rouy) im österreichischen Alpenraum. — österr. Bot. Z. 96: 109—137.
Fagerlind, F. 1937. Embryologische, zytologische und bestäubungseksperimentelle Studien in der Familie Rubiaceae nebst Bemerkungen über einige Polyploiditäts-probleme. — Acta Horti Bergiani 11: 195—470.
Favarger, C. 1954. Sur le pourcentage des polyploides dans la flore de l'étage nival des Alpes Suisses. — VIIIe Congrès international de Bot., pp. 51—56.
— 1959. Quelques problèmes de géobotanique alpine. — Texte présenté au IXe Congrès de bot. Montreal.
— 1959. Distribution en Suisse des races chromosomique Chrysanthemum Leu-canthemum L. — Bull. Soc. bot. suisse 69: 26—46.
Favarger, C., Richard, J.-L. et Duckert, M.-M. 1959. La Camarine noire. Empetrum
nigrum et E. hermaphroditum en Suisse. — Bull. Soc. bot. suisse 69:249—259. Favarger, C. et Söllner, R. 1949. Nombres chromosomiques et structure du noyau
de quelques Cerastium des Alpes. — Bull. Soc. bot. suisse 59: 87—90. Flovik, K. 1940. Chromosome numbers and polyploidy within the flora of Spitzbergen. — Hereditas 26: 430—440. F/Egri, K. 1960 in F/EGri, K., Gj/EREvoll, O., Lid, J. and Nordhagen, R. Maps of distribution of Norwegian vascular plants. I. Coast plants. — Univ. Bergen Skr. 26: 1—131.
Gustafsson, A. 1948. Polyploidy, life-form and vegetative reproduction. — Hereditas
34:1—22.
HAGERUP, O. 1931. Über Polyploidie in Beziehung zu Klima, Ökologie und Phylo-genie. — Hereditas 16: 19—40.
— 1941. Zytoökologische Bicornes-Studien. — Planta 32:6—14.
Hancock, B. L. 1942. Cytological and ecological notes on some species of Galium
L. Em. Scop. — The new Phyt. 41: 70—78. Holmen, K. 1952. Cytological studies in the flora of Pearyland, North Greenland. —
Medd. Grönland 128. pp. 1—40. HultÉn, E. 1950. Atlas över växternas utbredning i Norden. — Stockholm. 512 pp. Jacobsen, P. 1954. Chromosome numbers in the genus Hedera L. — Hereditas 40: 252—254.
j0rgensen, C. A., Sorensen, Th. and Westergaard, M. 1958. The flowering plants of Greenland. A taxonomical and cytological study. — Biol. Skr. Dan. Vid. Selsk. 9, No. 4, pp. 1—172.
Knaben, G. 1943. Studier over norske Pyrola-arter. Forelöbig meddelelse. — Bergens Mus. Arbok. Naturv. rekke 6: 1—18.
— 1954. Saxifraga osloënsis n. sp. A tetraploid species of the Tridactylites section.
— Nytt mag. bot. 3: 117—138.
— 1959. On the evolution of the Radicatum-group of the Scapiflora Papavers as studied in 70 and 56 chromosome species. Part A. Cytotaxonomical aspects. — Opera Bot. (Suppl. Bot. Not.) 2:3, 75 pp. — Part B. Experimental studies. — Ibid. 3:3, 96 pp.
Löve, À. 1949. The geobotanical significance of polyploidy. I. Polyploidy and latitude.
— Port. Acta Biol. Ser. A. R. B. Goldschmidt Jubilee Volume, pp. 273—352.
— 1953. Subarctic polyploidy. — Hereditas 39: 113—124.
Löve, A. and Löve, D. 1943. The significance in the distribution of diploids and polyploids. — Hereditas 29: 145—163.
— 1948. Chromosome numbers of northern plant species. — Univ. Inst. Appl. Sc., Dep. Agr. Rep. Ser. B, 3, 131 pp.
— 1956. Cytotaxonomical conspectus of the Icelandic flora. — Acta Horti. Got. 20: 65—291.
— 1957. Arctic Polyploidy. — Proc. Gen. Soc. Canada, pp. 23—27.
Manton, I. 1950. Problems of cytology and evolution in the Pteridophyta. — Cambridge. 316 pp.
Merxmüller, H. 1952—1954. Untersuchungen zur Sippengliederung und Arealbildung in den Alpen. — Jahrbuch Ver. Schutze d. Alpenfl. und Tiere 1952 Bd. 17 (Teil I); 1953, Bd. 18 (Teil II); 1954, Bd. 19 (Teil III), pp. 1—105. Müntzing, A. 1936. The evolutionary significance of autopolyploidy. — Hereditas 21:
Nordenskjöld, H. 1951. Cyto-taxonomical studies in the genus Luzula. — Hereditas 37:325—372.
— 1954. A cyto-taxonomic investigation into the campestris-multijlora complex of the genus Luzula. — VIIIe Congrès international de Bot., Paris, Section 9, Cytologie.
Skottsberg, C. 1939. Remarks on the Hawaiian flora. — Proc. Linn. Soc., London, 151: 181—186.
Sokolovskaja, A. P. and Strelkova, O. S. 1940. Karyological investigations of the alpine flora of the main Caucasus range and the problem of geographical distribution of polyploids. — Comt. rend. (Doklady) Acad. Sei. U.R.S.S. 29: 415— 418.
— 1941. Polyploidy and karyological races under conditions in the Arctic. — Ibid. 32: 144—147.
Stebbins, G. L. 1950. Variation and evolution in plants. — New York. 643 pp.
Stern, F. C. 1946. Plant distribution in the northern hemisphere. — Geogr. Journ. 705:24—40.
Söllner, R. 1950. Polyploidie intraspécifique chez Cerastium arvense L. et nombre chromosomiques de quelques autres Cerastium. — Experientia 6:335—340.
Tischler, G. 1935. Die Bedeutung der Polyploidie für die Verbreitung der Angiospermen. — Engler Bot. Jahrbuch 67: 1—36.
— 1950. Die Chromosomenzahlen der Gefässpflanzen Mitteleuropas. — Haag. 263 pp.
— 1955. Der Grad der Polyploidie bei den Angiospermen in verschiedenen Grossarealen. — Cytologia 20: 101—118.
Troll, K. 1925. Ozeanische Züge im Pflanzenkleid Mitteleuropas. — Freie Wege vergl. Erdkunde. — Festgabe E. v. Drygalsky zum 60. Geburtstag. Pp. 307—335.
WlNGE, 0. 1917. The chromosomes, their numbers and general importance. — Comt. Rend. Trav. Lab. Carlsberg, 13: 131—275.
263—378.
Contents
I. Introduction ..............................
II. Occurrence of polyploidy ..................
III. Significance of polyploidy in European floras
IV. Polyploidy in Norwegian flora groups ......
452 454
457 460 465 470
1. The arctic-alpine flora groups
2. The lowland flora groups
3. The oceanic species
V. Conclusions .........
Literature cited ..........
473 477
