The purpose of our research was to find significant differences in island and mainland plant populations of the Russian White Sea. These differences could be assumed to be the result of evolution or, less probable, the result of the bottleneck effect. For this purpose 6 most polymorphic species were chosen and studied: Atriplex nudicaulis, Euphrasia frigida, Achillea millefolium s.l., Parnassia palustris, Rhodiola rosea, Carex aquatilis and C. salina aggr.. Our results indicate that some species show this difference. Unfortunately, this difference doesn’t correlate with the island age, which means that we cannot estimate the evolution pace. Our research is still in progress, and in the future we hope to obtain more significant results.
During the ice ages, 12-15 thousand years ago, the entire Northern Europe was covered with ice. 17 thousand years ago the ice was around 3 km deep and caused enormous pressure on the Earth’s crust. As a result of this pressure the whole region of the future White Sea sank to form a depression about 200-300 meters deep. Later, when the climate on the Earth became warmer, the ice started to melt. At first, this process was slow; however, about 12 thousand years ago it gained full speed and 3000 thousand years later the ice has melted completely, forming the White Sea. With the reduction of the pressure, the Earth’s crust has started to ascend at a speed of approximately 5-8 millimeters per year. As a result, numerous islands arouse. New islands were inhabited from the mainland, and as a result the population on it becomes isolated from the mainland. If so, the evolution on it then goes in a different direction. As a result the island population becomes different from the mainland populations. This can also happen due to the bottleneck effect - when the plants that got to the island represent only part of the variability of mainland populations.
The purpose of our project was to find out the degree of difference between island and mainland populations. We predicted, that such differences could then be correlated with the height and/or area of the island. If so, we could estimate the rate of evolution on the islands.
In total 49 islands in the area of Kiv Gulf were investigated. We studied seven species, known to be the most common plants with the most important taxonomic diversity: Atriplex nudicaulis, Euphrasia frigida, Achillea millefolium s.l., Parnassia palustris, Rhodiola rosea and samples of the Carex group (C. aquatilis and C. salina). On each plant we measured about 8–12 morphological characteristics, which are known to be the most variable; 59 morphological characteristics in all. For Atriplex we measured the morphological characteristics of the leaves and bracts; for Euphrasia – morphological characteristics of the shoot and flower; for Achillea - morphological characteristics of the lower leaves and capitula; for Parnassia - morphological characteristics of the flower stalk and flower; for Carex – shoot and leaf characteristics.
The plants from each population were put in a herbarium, and parts of the plants - into silica gel for future DNA analysis. From each island one or more population (topographically isolated group of plants) were studied (about 20-25 plants in each population).
In total in 2003–2005, over 2500 plants were measured in 171 populations, taken from 49 islands and mainland. The basic method of analysis was the principal components analysis (PCA). As a result of it, each plant has been transformed into a point on a two-dimensional graphic concerning the coordinates of two components. Thus, each population was projected. Extreme points of a population were incorporated with lines, depicting a population: we ended up with polygons. When analyzing our data we considered that if two polygons did not cross each other, morphological characteristics of these two populations were very different. If two polygons were somewhat crossed among themselves, we considered, that the attributes of these populations have not changed to such extent, so we could differ these populations. Clearly, our purpose was to find populations with the most different attributes.
Our results specify that some island populations significantly differ from the mainland. We can say that Achillea, Rhodiola and Carex aquatilis populations are the most diverse, and the Carex salina-recta, Euphrasia frigida and Parnassia palustris are the least diverse.
As said above, PCA analysis did show some difference between island and mainland populations. We tried to find out if this difference is somehow connected with the island height and area. To analyze our data we used cluster analysis. For all the plants of one species that grow on one island the median was calculated. Thus, each island was presented as a median value. Next we compared these values by calculating the manhattan distances between them (so called “rnorphological distance”). The more is the difference between these two parameters, the more is the morphological difference between the plants in the populations. Then we tried to find out if these differences were connected with the height and area of the islands. We found significant negative differences between the morphological distance, height and area of the island (at least for Euphrasia, Carex and Rhodiola species)
Also, on some islands more then one population of the Achillea and Rhodiola species was measured. Using the PCA analysis, we found out, that no reliable differences between such populations can be found.
Based on the statistical analysis of our data, we can conclude that there are significant and stable differences between some island and mainland populations. These differences could be the consequence of bottleneck-like processes, because we have found the significant negative correlations with the age of islands (at least for Euphrasia, Carex and Rhodiola species). Atriplex diversity is too high for unequivocal conclusions. Achillea and Parnassia demonstrate another pattern, probably related with the "wind effect" that changed morphology of plants from outlying islands.
I want to thank all those who helped with measuring and organizing the data. Last but not least, I want to say a special thanks to Polina Volkova for the technical support.
To the main page