6408 Dryas L.
- L., Sp. Pl.: 501 (1753).
Notes: Many of Yurtsev's notes refer to previous drafts and versions. Several points emphasized by Yurtsev below have been taken care of in the current version. Elven has partly edited the notes with some corrections and comments marked by brackets.
As a positive feature I note the adoption of the Dryas punctata aggregate and of D. incisa, and all four subsections of sect. Dryas. In my turn I agree to reduce (until it is further investigated) D. chamissonis to a comment to D. integrifolia. Some doubts concerning this taxon were noted by Yurtsev (1984c), though the treatment of it as the hybrid D. integrifolia x octopetala is obviously incorrect. Below the controversial aspects of the checklist of Dryas are briefly discussed.
1. That two different hypotheses of the evolutionary history of the genus (mine and Kozhevnikov's 1979) are based on essentially the same facts is a normal situation. This gives the possibility of testing in various ways. My version of the history of the genus is a development of the classical works by Kuznetsov (1922) and Juzepczuk (1919, 1929, 1941). The former suggested a North American origin of the genus with Cowania as a present-day relative of the presumed predecessor. The latter drafted a phylogenetic system of Dryas with two sections: Nothodryas (two species) and Dryas (Eudryas) with three series, each of them with a probable ancestor in eastern Siberia and morphologically connected to D. grandis, the eastern Siberian species of sect. Nothodryas. This indicates eastern Siberia as the probable 'centre of origin' of Dryas. Porsild (1947) raised Yuzepchuk's series to subsections and Yurtsev (1994) added one more subsection: Incisae. The closest relatives to the presumed less advanced D. grandis in the respective subsections and series are: D. oxyodonta (subsect. Dryas, ser. Oxyodontae), D. ajanensis subsp. ochotensis (ser. Ajanenses), D. viscosa, and to some extent D. punctata subsp. punctata (subsect. Punctatae), D. sumneviczii (subsect. Incisae), and D. crenulata (subsect. Tenellae).
The supposition by Elven and Murray or rather Hultén, as Elven and Murray have no opinion on this matter that D. grandis might acquire some characters of sect. Dryas through hybridization is improbable. Dryas grandis (like D. drummondii) is an extremely conservative type, almost nonvariable (though able to produce hybrids with representatives of sect. Dryas).
2. Of fundamental importance is the fact of appearance of D. octopetala s. lat. at ca. 2.5 m years B.P. at the coasts of Iceland, the western and northernmost islands of the Canadian Arctic Archipelago, and northern Greenland, in the primary landscapes of a hypoarctic foresttundra which followed soon after the first opening of Bering Strait and migration of Pacific mollusks and foraminifers of the Atlantic. At the same time a plant close to D. ajanensis s. lat. could get from ancient alpine ("goltsy") areas to the primary deforested landscape of the coast of Arctic Ocean in strongly oceanic climate, prior to the formation of a permanent ice cover over the Arctic ocean. It permits to put some framework to the history of the genus.
3. Canescent forms in many species of Dryas are by no means modifications. In a number of species or subspecies they are absent (e.g., in D. oxyodonta and D. punctata subsp. alaskensis) and in some very rare (D. grandis, in flood plains together with non-canescent plants). In others they may be the only form as, e.g., in D. ajanensis subsp. ochotensis. Normally, the canescent forms occupy the windswept, dry positions in a hillocky or mountainous landscape and are replaced in lower parts of the slopes and in valleys by plants where the leaves are green and glabrate above. At intermediate positions, both green and canescent plants are met with side by side. But such a situation occurs only in certain areas, e.g., in Asian D. integrifolia only in the easternmost Chukchi Peninsula and in Wrangel Island. In other areas the green form occupies both dry snowfree ridges and snowbeds. Thus, canescent and green-leaved forms are normal ecotypes and deserve rank as varieties.
4. All species of Dryas are allogamic with entomophilous flowers open to a wide circle of pollinators, and with weak (or no?) reproductive barriers. All are diploids (2n = 18). Hybrids arise easily among nearly all sympatric entities of sect. Dryas (more rarely between contrasting ecotypes, see above). Hybrids are also known between D. grandis (sect. Nothodryas) and some representatives of sect. Dryas, though only rarely. The hybrid D. grandis x punctata is often sterile (Yakutia). However, in one locality in central Chukotka, at the northern boundary of D. grandis, we have found three local populations of it with plants intermediate in almost all characters and sometimes more abundant than at least one of the parents.
Despite almost free hybridization between species of Dryas, even introgressive in some areas, hybrid species or subspecies are nearly unknown, with the possible exception of D. octopetala subsp. subincisa (D. subincisa) suspected to be a hybrid between D. octopetala s. str. and D. incisa. Some diversity of species and races is kept even in sympatric situations. In subsections and series allopatric speciation (or raceforming) occurs. What factors support upkeep of the distinction between the sympatric species? This is probably due to geographic and ecological replacement. There are some almost or completely 'monopolistic' species with regional populations occupying huge areas. Such are D. punctata in eastern Siberia, where in many places it populates nearly all site types from snowfree summits down to snowbeds and mires, and D. integrifolia s. str. in most parts of Canada and Greenland. Dryas octopetala subsp. subincisa and D. punctata share dominance; in Taimyr the latter drastically predominates (including plakors), whereas in western Siberia, Polar Ural, Novaya Zemlya, and northeastern European Russia the dominance passes to D. octopetala subsp. subincisa, except for some sites where populations of D. punctata occur in their typical form (the same is probably the case in the Khibiny Mountains in the Murman area and probably also in northern Scandinavia). In these areas numerous hybrid plants or even populations occur described as D. x vagans. In Svalbard and northeastern Greenland, where D. punctata probably arrived much later than D. octopetala subsp. subincisa, almost full embracement of it by the former latter! may be suspected (the presence of lenseshaped and other glands in some populations), but there occur also sets of individuals in those parts looking as more or less typical D. punctata. Such forms are completely absent from, e.g., D. octopetala s. str. and from other Dryas species in regions where representatives of subsect. Punctatae are absent today and are not suspected to have been present in earlier times.
In contrast, in Central Beringia there are areas where up to five or six species and races of sect. Dryas share the diversity of site types within a mountainous or hillocky landscape, almost all with two contrasting ecotypes (green and canescent). The basis for replacement is provided for by contrasting snowfree (windswept) and snowrich (protected from winds) sites and similarly contrasting basic and acidic rocks. One may observe the dominance over significant areas of such species as D. incisa (e.g., on ultramafic rocks in the Anadyr drainage and the Koryak Mountains), D. crenulata s. str. (on limestones in subarctic Yakutia), D. ajanensis subsp. beringensis in Alaska and the easternmost Chukchi Peninsula (especially on acidic rocks though it also occurs on basic ones with D. integrifolia s. lat.), and D. punctata subsp. alaskensis in the southeastern Chukchi Peninsula and in Alaska and the Yukon Territory. Thus, in the case of sympatry of several species of the genus, it is ecological specialization and disruptive selection that support the upkeep of species diversity despite interfertility.
The last general remark. Elven and Murray following Hultén obviously underestimate the taxonomical value of such characters as a shape of leaf blade, the size and form of teeth, the condition of the upper and structure of lower surface of blade, the direction and pubescence of median and lateral veins, feature of stipules, scapes, the size of flowers, shape and pubescence of sepals and petals, etc. This results in a very schematical concept of taxa, based on few characters (the types of scales and glands, the presence or absence of teeth on margins of leaves, etc.).
Below follow my comments to the species and subspecies:
Dryas octopetala and D. ajanensis are considered by Elven and Murray as two subspecies of one species in a previous version - D. octopetala sensu Elven - despite each of them is divided by a complex of characters into two (D. octopetala) and even five races (D. ajanensis). This is due to obvious underestimation of the ordinary morphological characters, separating rather extensive geographic populations. Among these are shape of leaves; brightly green vs. dull surfaces of leaves; leaf thickness, dentation, and vein direction; shape and pubescence of sepals, etc. Dryas octopetala s. str. (subsp. octopetala) is easily recognized at a glance (photographs of specimens from LINN seen). In the Arctic it is replaced by subsp. subincisa which is also common in northernmost Scandinavia and the Khibiny Mountains not supported by molecular data. The type race is more similar to the races of D. ajanensis, whereas the arctic subsp. subincisa more resembles D. incisa (see below). This could be explained either by hybridization between D. octopetala s. str. and D. incisa (in this case one should adopt the rank of a species for D. subincisa), or by convergence in the true arctic environment.
Among the five races of D. ajanensis, four are restricted to alpine areas of Russian, Japanese, and Manchurian Far East and thus have fragmented ranges. Only the northernmost race - subsp. beringensis - has an extensive zonal range from the easternmost Chukchi Peninsula east to the eastern Brooks Range. It is one of the most distinctive subspecies of D. ajanensis, similar in shape of leaf blade to the eastern form of D. punctata but lacking glands on leaves. The subarctic alpine canescent subsp. ochotensis is the most similar - among all five races - both to D. oxyodonta and to D. octopetala.
About D. punctata subsp. punctata. Sporadic occurrence of rather typical specimens of the taxon in Svalbard and eastern Greenland does not permit us to exclude these areas from the range of this species. At least one should put a question mark. Neither it is possible to exclude Alaska and the Yukon Territory from the range of this taxon. Taking into account that in the case of subspecies not each individual can be identified to subspecies, it is risky to include subsp. hookeriana into the Panarctic Flora (put question mark and try to check specimens).
About D. incisa and D. crenulata. Despite their resemblance in leaves often being dentate throughout (though the uppermost part of a leaf often remains entire), the complex of subtle characters listed in Yurtsev (1984c) and Yurtsev in Nordal et al. (1999) does not permit to merge the taxa. These characters also testify that D. crenulata belongs to subsect. Tenellae. It is interesting that throughout the mountain alpine? range of D. crenulata, a form appears with subentire leaves but otherwise similar to the normal D. crenulata (Chersky Range, Transbaikalia, East Sayan). The sympatry of both taxa in southern Siberia is largely due to their basiphily and the pattern of distribution of basic rocks.
About D. chamissonis and D. integrifolia. There is a long set of evidence (including the features of the related genera Cowania and Fallugia and of the species of subgenus Nothodryas) that support the conclusion that the ancestral form of the genus had regularly dentate (serrate) leaves and that D. integrifolia therefore is the most advanced derived type within the genus. From this viewpoint, D. integrifolia might have arisen from a progenitor similar to the modern D. chamissonis - probably in unglaciated refugia in the westernmost Canadian Arctic Archipelago and then 'returned' to central and western Beringia. A weak degree of reproductive isolation might support a merger of these two taxa, though in West Chukotka D. chamissonis is the only representative of the D. integrifolia aggregate. Besides, in the areas of sympatry of both D. integrifolia and D. chamissonis, the latter drastically predominates (or sometimes is present alone) on drier windswept sites on subneutral soils, whereas D. integrifolia s. str. is more typical of carbonate mires and snowbeds, which could be attributed to the effect of disruptive selection. Another interpretation of D. chamissonis, as a hybrid D. integrifolia x incisa (or x octopetala), may be appropriate but only for a minor part of 'chamissonislike' plants with additional other characters of the second parent (especially if the latter is present as well). These comments should to be put in to mark an urgent problem for the nondistant future.
Thus, I consider D. integrifolia as the most advanced, specialized type in the evolution of Dryas, with well expressed feature of a pulvinate dwarf shrub (in mesic sites). As to D. integrifolia subsp. sylvatica, it could hardly be a progenitor of D. integrifolia s. str., rather a secondary microthermic ecological race.
-
Source: Indolences at Wikispecies
