HISTORY AND AGE OF SOME ARCTIC PLANT SPECIES 21 1 



partly glaciated. I have shown (Hadac, 1941, 1960) that practically all its 

 rare flora, all its relics, grow in places never glaciated during the Last Ice Age. 



Spitsbergen, Siberia, Alaska, eastern North America and Greenland have 

 vast areas of never glaciated territories where, by adaptation and isolation 

 during the Ice Age or Ages, new species could be formed, as, for example, 

 Coptidiiim spitsbergense, Pediciilaris dasyantha, PuccineUia vilfoidea, etc. 



Meanwhile, during the Mindel and Riss Ice Ages, many Arctic elements 

 penetrated southwards. I must say that I was very sceptical towards reports 

 of Arctic plant species from far distant mountains in the south until I myself 

 collected Oxyria digyna in the mountains of the Iraqi Kurdistan. 



In discussions on the Arctic flora and its migrations several authors have 

 emphasized that the migration followed step by step (i.e. not, or very seldom, 

 as single plants but rather in whole plant communities) which is possible only 

 over land connections. Dahl (1958) has shown (by statistical analysis of the 

 Scandinavian Amphi-Atlantic element and its adaptation to, or incapability 

 of, long-distance dispersal) that the occurrence of "western" elements in the 

 Scandinavian flora is not due to long-distance dispersal. This is a negative 

 proof. But we can test this problem also in another way. 



Let us suppose, for example, that the flora of Spitsbergen was supplied by 

 long-distance dispersal from the south, east, and west, and that its plant 

 communities were formed by free combination of all these geographical 

 elements according to their requirements. In such a case the different geo- 

 graphical elements of the flora should be represented in plant communities of 

 Spitsbergen according to the law of probabihty, and not all grouped together. 



The most common plant association of the Spitsbergen tundra in the Inner 

 Fjord zone is Tomentohypnetiim involuti (cf. Hadac, 1946, p. 155). The 

 climax community seems to be Cassiopetum tetragonae. The most luxuriant 

 community of this region is Trisetetum spicati. Let us now compare the 

 percentage occurrence of different geographical groups (which will be dis- 

 cussed later on) in the associations mentioned with the average of the whole 

 flora, i.e. with the "probable" occurrence: "probable" under the supposition 

 that their species came to Spitsbergen independently by long-distance dis- 

 persal (Table 1). 



More instructive may be the actual and "expected" numbers of species in 

 geographical groups (Table 2). 



We can see that the actual numbers of plant species in individual groups 

 very seldom agree with the "supposed" numbers and that the differences in 

 several cases (designated with*) exceed the standard deviation ; the differences 

 are thus statistically significant. 



It is also evident that in most communities plants of the same origin 

 remain together; they came to Spitsbergen not as single species but as 

 members of the same plant community. This can be proved as well for Iceland 

 and Greenland, as elsewhere. 



