Plants of Central Asia - Plant Collection from China and Mongolia, Vol. 5: Verbenaceae-Scrophulariaceae PDF

INTRODUCTION¹

This is the fifth volume of the series Plants of Central Asia and covers the two very large families of order Tubiflorae-Labiatae and Scrophulariaceae-and two smaller Central Asian families Solanaceae and Verbenaceae.

Family Labiatae is represented in the studied territory by 41 genera and 157 species, of which only 1 genus and 22 species are endemic. The Central Asian regions of the former USSR (desert and desert-steppe Kazakhstan, Northern and Central Tien Shan and Eastern Pamir) would add to this list 7 more genera and about 60 species. Family Scrophulariaceae is represented in the studied territory by 18 genera and 152 species, of which 23 species are endemic. Two more genera and about 60 species would add to this list from the Central Asian part of the former USSR.

These two large families, however, are of little florogenetic interest. In Central Asia, they represent plants that have migrated relatively recently, mainly from the Mediterranean and, to a lesser extent, from Eastern Asia Nepeta, Stachys, Elsholtzia, Phyllophyton, Pedicularis, Lagotis, Oreosolen). Most species of these families are confined to the mountain borders of our region in the south-east, north-west and specially the west. They are predominantly high-mountain-alpine and hill-steppe-as well as steppe species. Only a few of the species of Labiatae and Scrophulariaceae have contributed to the desert flora of Central Asia, for example Lagochilus ilicifolius, Schizonepeta annua, Eremostachys molucelloides, Dodartia orientalis. The endemic species of these families are less indigenous and are essentially neoendemics. The recently established lone endemic Tien Shan monotypic genus Metastachys, very poorly differentiated from genus Stachys, is no exception. The more typical subendemic East Siberian-Central Asian genus Cymbaria with 2 steppe species is clearly of East Asian origin.

In Central Asia we found the so called terminal segments and links of taxonomic chains of Labiatae and Scrophulariaceae extending from outside. There are no indigenous local taxonomic groups-sections, cycles or series. This is also true even for such a large genus as Pedicularis.

Distribution of the high-montane monotypic genus Lancea should be regarded as a very prominent, but not the only proof of direct floristic links in the past. The major part of the distribution range of Lancea tibetica falls in the Himalayas, Tibet and adjoining montane regions of West China (see Map 3) while a small isolated part has been found in the main Hangay mountain range in Northern Mongolia with a disjunction of a thousand kilometres, now falling in the desert expanse of Central Mongolia. Similar instances of distribution are known in other families too, for example among Megadenia from Cruciferae and Mannagettaea from Orobanchaceae detected by Popov (1954) in the Eastern Sayans and Rhododendron from Ericaceae. Such disjunct distribution ranges represent more a heritage of the cold Pleistocene Age when the alpine species of East Asia may have penetrated uninterruptedly along the high Mongolian plateaus to the north. Popov's view (1954) that Mannagettaea and Megadenia represented Paleogene subtropical relics, in our opinion, is not adequately substantiated. Neither their distribution ranges, confined to the high mountains in the eastern fringes of Tibet (Qinghai and Weitzan), nor the floristic affinities of host plants (Salix, Caragana, Ribes) on which all the known species of Mannagettaea parasitise favour such a conclusion. The relationships of these genera, if not extended theoretically, fall within the boreal forest and Mediterranean floras.

Family Solanaceae is far more interesting. This very ancient pantropic family is represented in Central Asia by only 9 genera and 21 species of these, 1 genus and 4 species are endemic and 3 more species subendemic. Endemic genus Przewalskia (see Map 2), surprisingly, is characteristic of the most younger Tibetan province whose flora may have formed only after the Glacial period (see Introduction, Vol. 1). This genus is undoubtedly close to genus Scopolia but is well distinguished by its unique characteristics and presence of major distinctive features. It was first established by C.I. Maximowicz as a monotypic genus but it is now clear that it comprises 2 distinct species. These are rather small rosetted plants with narrowly tubular axillary flowers and typical swollen fruit in which the accrescent inflated calyx forms an outer coarse membrane for the small pod, dehiscent with an operculum as in Scopolia. Such is the fruit structure in subendemic eastern Tibetan species Scopolia tangutica but the calyx accrescent in fruit, unlike in Przewalskia, is open, cup-shaped, its fringe closed at the top with folds only at the end, as the fruit ripens. The seeds of Scopolia tangutica and both species of Przewalskia are so similar that they could hardly be differentiated from external characteristics. Scopolia tangutica has a large branched stem but the early phase of growth of this species in spring resembles a rosette-a shaft with developed leaves emerges from the soil all at once. Among the eastern Tibetan species, Przewalskia tangutica has only a short underground stem and an ovate calyx nearly closed on top but a surface stem is sometimes seen in southern Tibetan P. shebbearei but the calyx in fruits is open at the top, with large teeth. The thought arises spontaneously that Przewalskia may have developed as a branch of the Tertiary genus Scopolia in the course of evolution adapted to the severe conditions of high mountains through neoteny. Since this genus is undoubtedly young, such a high tempo of evolution can only be explained in this manner. Moreover, in the case of P. shebbearei, this process is not complete-the surface stem is not yet fully reduced and the accrescent in calyx in fruit is not completely closed above, resembling its form in Scopolia tangutica (see Plates IV and V). The latter is a forest and high-montane species of Qinghai and the southeastern border of Tibet; on the contrary, compared to its closely related Chinese forest species S. sinensis Hemsl. it represents somewhat the first step towards such a process of evolution.

Another interesting endemic of the same type was the new species of genus Mandragora described here from eastern Tibet-Weitzan. According to established concepts, this ancient Mediterranean genus was known in East Asia from the single high-montane species M. caulescens distributed in Kam, Eastern Himalayas and in eastern Southern Tibet. Unlike the Mediterranean large-leaved rosetted species, M. caulescens has a surface stem, small leaves and broadly campanulate nutant flowers. Our new, proper Tibetan species M. tibetica undoubtedly represents the closest relative of Kam-Himalayan species (it has in particular the same form of leaves and scales with the same characteristic ciliate pubescence as in M. caulescens) but grows in even more severe dry and cold high uplands of eastern Tibet and is a small densely rosetted plant. Thus, even in this case, evolution evidently proceeded along the line of inherited juvenile stage of the vegetative phase.

The large (with more than 100 species) extratropical genus Lycium in the Old World is confined mainly to southern Africa and the Mediterranean and is represented in our territory only by 6 species, of which 1 is endemic and 2 subendemic. The endemic Junggar-Tien Shan L. flexicaule and subendemic MongolianL. potaninii, according to A. I. Pojarkova (1950), belong to a series of eastern Asian species while the subendemic Central Asian L. truncatum belongs to a series of eastern Mediterranen species. Thus we find here rare instances of 2 branches of an old genus arriving from opposite directions taking root in the Central Asian flora, although it must be conceded that all of these species are very closely related. Unlike the above erythrocarpous species common in Central Asia, melanocarpous L. ruthenicum is an extensively distributed Mediterranean desert species.

In our area, there are only 2 species from the large pantropic family Verbenaceae; they provide very striking proof of the direct penetration into Central Asia of the typical eastern Asian warm-temperate forest genus Caryopteris (some ten species distributed in China, Korea and Japan). While C. tangutica enters only from the east into the hilly forest Qinghai belt preserving the typical mesophytic features of the genus (tall shrub with broad dentate leaves), its closest relative C. mongholica enters the composition of desert-steppe and desert flora of Central Asia to a iimited extent over a vast distribution range and is a subendemic. It acquired a totally xerophytic habit-low subshrub with densely pubescent entire, narrowly lanceolate leaves. Its affinity with C. tangutica is revealed not only by its flowers, but also by var. serrata with interrupted dentate leaves found from time to time along the southern fringe of its distribution range (see Map 1). This species evidently penetrated into Central Asia from the eastern boundary where it encountered no orographic barriers, as early as in the Tertiary period and became a total native long ago.

In this volume, the maps of distribution ranges of species of families Labiatae and Scrophulariaceae were drawn by their respective authors and of families Verbenaceae and Solanaceae by O.I. Starikova, senior laboratory assistant. She also translated all Chinese texts from labels of herbarium specimens and from floristic literature. Further, she rendered much assistance to the authors in drawing up distribution ranges and references to the geographic distribution of species for which the authors express their sincere gratitude. Plates of plant drawings were prepared by artists G.M. Aduevska (plate IV) and T.N. Shishlova.

¹By V.I. Grubov

Author: V I Grubov