Hereditas. 3. 1922
The Genotypical Response of the Plant Species to the Habitat

Göte Turesson

Institute of Genetics, Åkarp, Sweden

C. LEONTODON AUTUMNALIS.

Sets of this plant have been collected from beach habitats as well as from the inland in 1919, 1920, and 1921, and important differences are seen in the cultures between some of these sets. This is particularly true of the series transplanted in 1920 from the coast at Kullen and Arild (N. W. Scania). The plants included in these series are all smooth with coarsely dentated or nearly entire leaves. The plants making up the series transplanted from inland meadows in Scania have mostly pinnatifid leaves, usually somewhat hairy. These plants are also taller than the coast plants. The cultivated coast plants from Kullen and Arild, furthermore, have been found to flower about two weeks before the inland plants. Clones have been raised from individuals of these different series. Fig. 13 (to the left) represents an individual from a clone obtained in 1921 by dividing a plant typical of the Kullen series (transplanted 1919), while the plant to the right illustrates the individuals in a clone obtained the same year by dividing an inland meadow plant (transplanted 1919). The differences as to height and shape of leaves are plainly seen. The anatomical structure of the leaves of these clones is also different, as is seen in figs. 75 c and d. The cross-sections are made from leaves taken from the cultures in 1922. There is a slight but distinct difference in thickness. The thickness of the inland clone has been found to vary between 250μ. and 300μ., while the Kullen clone varies between 285μ. and 350μ. The loose tissues rich in intercellular spaces, which are typical of the inland leaf, contrast strongly with the compact tissues of the Kullen leaf. These differences in the anatomy of the leaf have been found to be very typical of other series from the coast (Varberg, Bohuslän) and the inland. In places where meadows and pastures run down to the shore, there is a great confusion of different forms, as series from such localities show. Such series have been transplanted from Vellinge and Hallands Väderö on the west coast, and from Västervik on the east coast. They apparently represent different combinations between the coast type and the inland type.


Fig. 13. Leontodon autumnalis. The type of the coast (to the left) and the inland type.

Plants from the clones represented in fig. 13 have been crossed. The F1-generation is multiform with regard to hairiness and to the shape of the leaf. The further discussion of the segregation will be put off till the next generations have been raised. It should be said, however, that in the F1-generation already produced types of leaves have appeared which differ from both of the parents, for instance pinnate types with extremely narrow lobes (identical with the f. coronopifolius Lge. of the floristic handbooks).

A few words should be said in regard to the treatment by systematists of some of the forms of L. autumnalis. NEUMAN (1882-83), in our country, has discussed the various forms. The above coast type from Kullen, Arild and Varberg is apparently identical with the form described under the name f. salinus Aspegr. There is another similar form, which NEUMAN describes as f. robustior. It differs from salinus mainly in being shorter in stature and in having more succulent leaves. There seems to be little doubt that this form is merely a habitat modification of the former. For it has been found that plants answering to the description of f. robustior at the time of collecting have become taller and less fleshy upon cultivation as early as the year after transplantation. It is equally true, however, that the characteristics of these cultivated plants, which now match the plants of the coast type described above, are genotypically founded and are not to be regarded as modifications of inland plants.

In summarizing the results of the cultivations the following conclusions as to the existence of different hereditary types of L. autumnalis seem to be justified.

  1. The type of the interior of Scania. The type includes rather tall plants with pinnatifid and usually hairy leaves. The leaf is composed of very loose tissues rich in air spaces.
  2. The type of the coast, as obtained from Kullen, Arild, and Varberg. This is composed of plants lower in stature than the inland type. The leaves are coarsely dentated or almost entire, and always smooth. The leaf is composed of very compact tissues, poor in air spaces, and is somewhat thicker than the leaf of the inland type. The plants of this type flower about two weeks earlier than the inland type. The f. salinus Aspegr. of systematic handbooks seems to cover this coast type.

It may safely be held that the number of types characteristic of L. autumnalis in nature is not exhausted by these two. In sandy places along the coast plants with depressed, pinnate leaves are found, which have been described as f. coronopifolius Lge. The cultivated material of this form is too young, however, to permit of any definite conclusions as to the nature and importance of the form. The same is true of cultivated material of the Alpine forms of the species. It may be said, however, that the compactness and slight succulence of the leaves, characteristic of the coast type, seem to be characteristics of the Alpine forms as well, and are even more pronounced here than in the coast type.

D. MELANDRIUM RUBRUM.

No succulent coast form of this species seems to have been recorded. Such a form grows, however, on the sea cliffs of Kullen in N. W. Scania. It has been in culture for two years and does not yet show any decrease in thickness. Fig. 76 f represents a cross-section of a leaf of the ordinary type, cultivated since 1920, so common in the beech woods of Scania. Fig. 76 e shows a cross-section from the cultivated, succulent Kullen type made in 1922. The thickness of the leaves of the cultivated type of the beech woods varies between 295μ. and 350μ., while the leaves of the cultivated plants belonging to the type from Kullen vary between 575μ. and 615μ. The measurements refer to well-developed rosette leaves in autumn.

Great variation is found within both types in regard to the shape and hairiness of the leaves, the shading of the red colour of the flower, etc., but the characteristics of the leaf thickness are peculiarly uniform in each of the two types.

It should not be thought that succulence is characteristic of the coast type of Kullen only. Alpine varieties of this species, now brought under culture, seem to be still more fleshy. The material is too young to allow definite conclusions to be drawn, but a few plants belonging to this Alpine type, transplanted from Jämtland in 1918 and cultivated by Professor NILSSON-EHLE in the same field as the abovementioned Lowland types, have been examined as to leaf succulence this autumn (1922). The leaf thickness was found to vary between 645μ. and 735μ. Fig. 76 d represents a cross-section of a rosette leaf from one of these plants.

There are, no doubt, additional examples of thin-leaved inland plants represented by succulent types both along the coast and in the mountains. Preliminary work on Rumex acetosa L. seems to point in this direction. These coast and Alpine types of one and the same species represent an interesting illustration of the fact so well known among plant geographers that one and the same species may be found both along the coast and in the Alpine regions (SCHIMPER, 1908). It has yet to be shown, however, that these species really are represented by the same hereditary type in both localities. That this is not so in the case of the above-discussed plants seems certain.

The results of the cultivation of succulent coast types discussed above leave little room for doubt as to the genotypical differences between these types and the corresponding inland types. That the presence of sodium chloride, in the amount found in the coast habitats, should in certain cases modify these coast types towards a degree of succulence not attained in the cultivated series (as we have found in the case of Solanum Dulcamara and Leontodon autumnalis) is not surprising when it is known that a great number of non-halophytic plants are also thus modified. There is — even in those series which have been found to be less succulent in the cultures than in their natural habitats — a distinct difference as to the thickness of the leaf between the cultivated coast types and the cultivated inland types from the Lowland. When the latter have been found to be thin-leaved, while the former, cultivated in the same bed under the same conditions, are found to be succulent, genotypical differences between both types as to leaf thickness must, no doubt, exist. This is made all the more plausible when the F2-segregation from the cross between the two Matricaria types, as well as the behaviour of the cultivated shade form of the Väderö Solanum, are remembered. The existence of succulent Alpine forms of some of these species also speaks in favour of the assumption that succulence by no means stands in any direct proportion to the quantity of mineral salts present in the different soils.

This does not lessen the importance of the fact of the modificatory action of salt upon the plant; it must be emphasised, however, that the results of the cultivations do not support the idea that the halophytic forms of thin-leaved inland species, found habitually along the coast, are modifications called forth in direct response to the saline soil. The classification of these forms into a group of "facultative" halophytes (WARMING, 1906) must therefore be refrained from. They are, on the contrary, to be considered as obligatory as most of the well-known species confined in nature to saline situations.

4. MATERIAL AND CULTURES OF DIFFERENT SPECIES OF HALOPHYTES.

The cultural experiments with species of halophytes have been largely confined to the genus Atriplex, and the discussion pertaining to this topic will therefore be limited to species belonging to that genus. Atriplex patulum L., although not a regular halophyte, has also been included in the discussion.

No other genus would seem so suitable for investigations of this kind as Atriplex. They are annuals, and several species of the genus have long been known to be exceptionally polymorphic. Various forms, dwarf forms, forms presumably adapted to live on sandy shores, on rocky shores, on sea-weed banks, etc. have been distinguished within several species and named by systematists. The polymorphism has also been assumed to be due, to a certain degree, to the occurrence of hybrids between some of the species. The method largely employed in my own work has been to collect series of young plants of the various forms and species from different points on the Swedish coast for my cultures. As many as possible of these transplanted individuals have then been isolated with pergamin bags at the stage of flowering, and the offspring has been compared with the mother-plants. All the species investigated seed perfectly well when isolated in that way. The isolations have been continued for several years and a considerable number of biotypes have been isolated from the different species. A remarkable multitude of different hereditary forms have thus been found to build up the species in nature. The frequent self-fertilization, occurring also in nature, tends to preserve these different forms, thus giving rise to the perplexing polymorphism found. That this polymorphism is sometimes increased by crossings between different species, and by the direct influence of environmental factors upon the plants will also be shown. It has been thought best to discuss the different species separately.

A. ATRIPLEX LITORALE.

The species occurs commonly along the Swedish coast. It has a shallow root-system and is therefore limited to the lower portion of the beach, where it often roots in the sea-weed banks thrown up. Farther up on the sandy beach only dwarfed forms are produced. The plant flourishes best in sheltered places and keeps away from the most exposed localities (cf. also TURESSON 1919 a). It reaches its best development along the Sound, especially on the beach-lines bordering the bays and coves, where masses of Fucus and Zostera have collected. The dimensions reached by the plant along the Sound are not attained on the coast farther to the north, e. g. in Halland and in Bohuslän. The forms met with in sheltered places along the Sound are often tall, erect, and very robust, while the forms farther up on the more exposed west coast are lower of stature and more spreading. The question arose whether these differences between the forms of the Sound and those of the coast line to the north are hereditary or only modificatory and wholly dime to differences in the environment. In order to answer the question collections of seed and small plants were made in the years 1916, 1917 and 1918 from various places along the Sound and the coast strip of Halland and Bohuslän. A great many of the plants raised were isolated with pergamin bags during these and the following years and a representative collection of different biotypes was obtained. It was thus found that the A. litorale populations both of the Sound and of the Halland-Bohuslän coast were made up of a large number of biotypes often differing noticeably from each other in various respects. It was also found that certain biotypes, representing certain combinations of characters, could be repeatedly isolated from the population of the Sound but not from the Halland-Bohuslän coast. The additional fact was also brought to light that certain combinations very characteristic of the northern coast strip occurred but sparsely along the Sound. Table 3 gives the characteristics of 17 biotypes, most of them in culture since 1916 -1917. The middle values of the length of the main axis and of basal branches have been based upon the measurements of the first 30 plants in the row of each biotype. The angle taken up by the main axis and the basal branches (columns. 4 and 5) has been indicated by the terms erect, ascending, procumbent and prostrate. The measurements were made in 1921, and the different biotypes are arranged as to the length of the main axis (column 2).

Field n:o Middle length of main axis (in mm) Middle length of basal branches (in mm) Position main axis Position of basal branches Leaf margin Bractlets Flowering Place and year of coll. of the mother plant
171 1005 1035 Ascending Ascending Denticulate Closed Late The Sound, 1916
172 980 965 Erect " Dentate " " "   "
175 970 990 " Erect Almost entire " " "   "
176 945 970 " " " Spreading " "   "
169 885 925 " " Entire Closed " "   "
170 860 870 " Ascending Almost entire " " "   "
177 810 840 " " " Spreading " " 1917
178 805 825   Erect Denticulate " Intermediate "   "
168 635 715   Ascending Dentate Closed Late "   "
173 580 595 Ascending " " " Early Bohuslän, 1918
174 540 600 " " " Spreading " "   "
155 510 545 " " Entire Closed Intermediate Halland, 1917
159 425 430 Procumbent Procumbent Almost entire Spreading Early "   "
160 410 460 Prostrate Prostrate " " " "   "
156 405 480 Procumbent " Dentate " Intermediate Bohuslän, 1917
158 375 415 " " " " Late The Sound, 1916
157 330 405 Prostrate " " " Intermediate, "   "

 


Fig 14. Atriplex litorale, field n:o 169. Plant to the left raised under ordinary field condition; the plant to the right raised in sand culture.

The biotype represented by field no. 169 is figured in fig. 14, field nos. 170 and 168 are shown in figs. 15 and 16, and nos. 159 and 160 in figs. 17 and 18. The biotypes represented are selected so as to give an idea of the great variability in branching habit. It should also be said that each of the 17 biotypes in the list appear to be constant throughout, the small differences seen between the individuals of the same type being certainly due to fluctuations in soil etc. They resemble in all essentials the separate mother-plants from which they originated. This fact speaks in favour of the assumption that autogamy is the rule in this species. A few mother-plants have upon isolation given offsprings showing segregation as to branching habit, time of flowering, colour of leaves, shape of leaf margin, etc., but these families have been found to be in a decided minority.

The pure cultures of biotypes of A. litorale comprised in 1921 36 different kinds. About half of the number of the pure cultures comprised only 15-25 individuals in each family and have on this account been excluded from the table. The number of isolated biotypes could easily have been increased were it not for practical difficulties with regard to the space of ground needed and the care required for a greater number of isolations. The material at hand, however, is quite sufficient to make clear the marked hereditary differences between the different biotypes. In the 17 cultures of biotypes none is found to represent the combination of characters typical of another. Between some of them only small differences are seen, but these differences repeat themselves with absolute exactness every year.

Only one character is found to be less distinct in certain years,. viz. the time of flowering. In years when development is retarded and the flowering period is detained because of cold and rainy weather the flowering times of the different biotypes almost coincide, and differences are hard to detect. Such was the case this summer, while in 1921 as well as in 1920 the differences as to this character were beautifully seen.


Fig. 15. Atriplex Morale, field n:o 170.

Fig. 16. Atriplex litorale, field no. 168.

The variations as to branching habit, leaf character, bractlets, etc.seen in a natural population of A. litorale are probably in most cases due to the occurrence in the same habitat of a mixture of distinct biotypes each representing a distinct combination of hereditary characteristics. It should not be thought, however, that environmental factors may not influence the habitus. Modificatory dwarf forms are often developed in the upper part of the beach, probably on account of an insufficient water supply. Similar plants differing much in size from individuals of the same biotype growing under or dinary field con ditions are invariably produced when allowed to develop in sand culture. Fig. 14 shows the habi tus of two indi viduals belonging to the same biotype (field no. 169) of which the one to the left has developed under the normal conditions of the experimental fields, while the dwarf to the right has been raised in sand culture. There is also another group of modifications, viz. the modificatory prostrate forms, which afford much interest. The prostrate habit of growth, which in some biotypes is a hereditary character (as in field no. 160, fig. 18) might sometimes, especially in exposed localities, become developed as a result of response to environmental factors. I have in a former publication tried to show (TURESSON 1919 a) that exposure to great light intensity produces this modificatory prostrateness. Species belonging to the genus Atriplex were also mentioned which possessed different hereditary branching types, one erect in ordinary light and one prostrate in ordinary light (A. latifolium and A. patulum), of which the former type could be induced to become more or less plagiotropic when exposed to light of great intensity. An erect type of A. litorale was also experimented with on that occasion, but this was found to be the least sensitive of the different forms (belonging to other species) tried. Since then instances of modificatory prostrateness have also been found in A. litorale. It is especially the procumbent biotypes which are apt to become modified in that way. Prostrate modifications almost always result when individuals belonging to these types are allowed to develop in the sand culture, where much light is reflected from the white sand and where the plants, because of their reduced size, do not shade one another. Fig. 17 shows the habitus of two individuals belonging to the same biotype (list no. 159) of which the upper plant has been raised in the ordinary field culture, while the lower plant illustrates the modificatory prostrate habit of growth attained in the sand culture. When modificatory and hereditary prostrates occur together on the exposed beach some difficulty is experienced in separating the different types, and the cultivating of the forms becomes necessary. Prostrate modifications of erect types have not been found. The ascending types may become influenced to a limited degree and are then hard to distinguish from the procumbent types. These latter are most readily induced to become prostrate.


Fig. 17. Atriplex litorale, field no. 159. The upper plant grown under ordinary field condition; the lower plant raised in sand culture.

Fig. 18. Atriplex litorale, field no. 160.

The distribution in nature of the different types of branching is very interesting. The erect type of branching so characteristic of the types of the Sound has not been found in any of the series of young plants and seed collections obtained from the Halland-Bohuslän coast. The tall and erect branching type, if not entirely lacking, must be but sparsely distributed in that region. Extensive series from Kristineberg and Marstrand in Bohuslän and from Varberg, Falkenberg, Halmstad, Bástad and other places along the Halland coast have upon cultivation only given plants belonging to the ascending, procumbent and prostrate types of rather low stature. A few of these types are still being kept in pure cultures (see the table). There is another characteristic of these types. They have been found to flower earlier than most of the types occurring along the Sound. The difference in the time of flowering for instance, between field no. 169 from the Sound and list no. 173 from Bohuslän (see the table) amounts to about 14 days. As has been remarked this difference in flowering-time may become less distinct in certain years, but in years with normal weather conditions the difference is always seen. The difference in flowering-time between the Sound types and the Halland-Bohuslän types is generally less than 14 days; there is usually about a week's difference between the two groups.

The tall, erect types are the most common in the Sound region, the ascending ones are somewhat less frequently distributed, while the procumbent and prostrate types are very rare. The gigantic dimensions attained by the A. litorale vegetation along sheltered sea-weed banks of the Sound, both as to height and as to extent, are thus seen to be due not only to the occurrence of more favourable localities for the plant in this region as compared with the more exposed northern coast strip but also to the occurrence of hereditarily more robust and taller types than are found in the populations to the north.

The limited number of cultivated types from the eastern coast strip seems to indicate that the A. litorale types inhabiting this coast strip differ genetically both from the plants typical of the Halland-Bohuslän coast and from those of the Sound region.

B. ATRIPLEX BABINGTONII WOODS.

This species occurs along the west coast as far to the south as Kullen. It is only found sporadically along the Sound. It has also been found quite typically at Stenshuvud on the east coast of Scania, and occurs likewise on Oland and Gottland. The plant has a deep-ground rootsystem and generally occupies the upper supralittoral belt of the beach. A number of biotypes have been isolated from this species also but no differences in the geographical distribution of the types have been found. Populations from Bohuslän, Halland and Gottland show the same mixture of types as to branching habit, leaf margin, bractlets, etc. Autogamy does not seem to be the rule in this species. In all localities where A. Babingtonii co-exists with other species of the genus, for instance with A. latifolium Wg., hybrids are found. The hybrids are in fact much more common on our west coast than A. Babingtonii itself. It is only in places, where the zones of the beach for some reason or other are devoid of other Atriplex species, that A. Babingtonii is found in its typical form.


Fig. 19. Atriplex Babingtonii. Ascending and procumbent types.

Fig. 20. Atriplex Babingtonii. A prostrate type.

The branching types usually found in A. Babingtonii are the ascending and the procumbent ones (fig. 19). The prostrate (fig. 20) type is less common. No biotype characterized by wholly erect growth has yet been isolated.

C. ATRIPLEX PATULUM L.

This species has been considered to be one of the most variable in the genus, and forms presumably adapted to live on cultivated ground, on sandy shores, and on rocky shores, etc. have been distinguished and named by systematists. I have shown in a previous paper (TURESSON, 1919 b) that the great variability assumed by some writers to be peculiar to A. patulum has been due to a confusion of this species with forms belonging to another species, which is strictly confined to the salt water shores. This species will be dealt with in the following.

TABLE 4.

Field no. Position of main axis Position of basal branches Leaves with or without scurf Margin of the terminal part of the basal leaves Bractlets
6 Erect Ascending Almost without Denticulate Small
7 " " " Entire "
8   Procumbent With Dentate "
3 " Prostrate Almost without Entire "
4 " " Without " "
5 " " Almost without Denticulate "
209 Prostrate " With Entire "
211 " " Without " "
212 " " " " Leafy

A. patulum, on the contrary, is a typical anthropophyte, which only goes down on the beach in places, where cultivated fields are close by or where refuse and rubbish has collected. Autogamy seems to be the rule in this species, and numerous biotypes, different as to branching habit, leaf character, bractlets etc., are easily raised, when different mother-plants are isolated. A great number of such isolations were made in 1916 and 1917. A few of the resulting pure cultures are still kept. Table 4 shows the combination of the most important characters realized in these different pure cultures.

Fields nos. 6 and 3 are shown in figs. 21 and 22, and field nos. 5 and 209 in fig. 24 (the main axis of this type is erect but it is sooner or later bent by the wind) and in fig. 23. When growing, as it usually does, on much trodden road-sides or in crowded fields it becomes sometimes much dwarfed. The erect type will then in most cases have an appearance corresponding with the lower plant in fig. 21, which belongs to the same biotype as the upper plant (field no. 6). It has been raised in sand culture. Plants belonging to the prostrate type will similarly take on the habitus shown by the lower plant in fig. 23, which is a modification of the upper plant (field no. 209) and is raised in sand culture. These prostrates become even more stunted and dwarfed when growing as a weed in the corn-fields of southern Sweden; they often do not exceed 10 cm. when spread out.

Fig. 21. Atriplex patulum, field no. 6. The upper under normal field conditions, the lower under sand culture conditions. Pots 12 cm. high.

 


Fig. 22. Atriplex patulum, vield no. 3.

 

Fig. 23. Atriplex patulum, field no. 209. The upper under normal field conditions, the lower under sand culture conditions. Pots 12 cm. high.

 


Fig. 24. Atriplex patulum, field no. 5.

The different biotypes occur almost indiscriminately through southern and middle Sweden in cultivated places. There is one habitat, however, which has been found to exclude all types which do not show a prostrate habit of growth. In the stubble-fields of southernmost Sweden, where A. patulum grows as a weed, only the prostrate races of the species are found, together with other annuals low in stature, such as Anagallis, Odontites, Euphorbia exigua, Stachys arvensis etc. The offspring from isolated mother-plants collected in the stubblefields might vary as to all other characters, in the shape of the leaves, in characters of leaf margin, in the size of the bractlets, etc., but they all show the prostrate habit of growth. They escape, apparently, the fatality which inevitably meets the erect races in this habitat, viz. that of being cut down and deprived of fructification.

A. patulum var. erectum Huds. of the floristic handbooks includes most of the types with an erect habit of growth, while the var. angustifolium J. E. Sm. comprises a great assemblage of prostrate types together with dwarfs and hunger forms of the preceding types.

D. ATRIPLEX SARCOPHYLLUM.

The A. patulum-like shore Atriplices with thick and fleshy leaves are here grouped under this preliminary name. I have shown in a former publication (TURESSON, 1919 b) that these forms have been erroneously grouped with A. patulum, which species they resemble in the shape of the leaves. They form together one of the most distinct species in the genus, systematically and ecologically, and occur almost along the whole Swedish coast, occupying the zone nearest to the water (the lower supralittoral belt), where no other species of the genus thrives well. They flower earlier than any other species of the genus, as early as May and June. Sets of seeds and young plants of forms belonging to this species have been collected from some thirty different localities and cultivated, and a remarkable hereditary variation has been found to occur. Furthermore, the geographical distribution of the various forms is of great interest. It is found that the eastern coast strip is inhabited by a certain group of forms distinctly different from the population inhabiting the Sound region, and both these groups of forms differ again from the forms found on the western coast strip. It is preferable to discuss these different geographical groups separately.


Fig. 25. Atriplex sarcophyllum from the east coast (Oxelösund). Pot 12 cm. high.

Fig. 26. Atriplex sarcophyllum from the east coast (Gottland). Pot 10 cm. high.

a. The type of the eastern coast strip.

The forms of this type have the general characteristics of the whole group: thick and fleshy, rhombic-lanceolate leaves; long, unbranched and, in the upper part, leafless inflorescences; early flowering etc. The eastern forms differ from the rest in the branching: they are erect, with the basal branches ascending. They have in addition somewhat thicker leaves than the forms of the Sound and the western coast (see figs. 74 e and f). The mean leaf thickness of the east coast type has been found to vary between 945μ. and 1155μ., while that of the Sound type varies between 893μ. and 1067μ. For purpose of comparison it may be mentioned that the thickness of A. patulum has been found to vary between 315μ. and 525μ. The values refer all to cultivated biotypes. The fruiting bractlets of these eastern forms are often smaller in size and more fleshy than in the other forms. When it is examined in the field, an impression of the great homogeneity of the type is obtained. When brought into culture, however, a number of different forms can be isolated. The offspring obtained from isolated plants appears to agree in all details with the respective mother-plants, indicating autogamy in a state of nature also. The following enumeration refers to biotypes now kept in pure cultures. The place and year of collection of the original mother-plants are also given together with short notes (taken in 1922) on the characteristics of the resulting pure cultures. The number of individuals in each pure culture varies this year between 15-25.

Namndö, off Stockholm, 1920 plants of rather low stature; leaves deep blue-green.
Oxelösund, Södermanland, " plants stouter and considerably more branched than the former; not quite as blue-coloured. Shewn in fig. 25.
Getå, Ostergötland, " plants resembling in all essentials the Nämndö form.
Västervik, Småland, " plants resembling in all essentials the Namndo form.
Kalinar, Småland, " plants more slender than any of the former; the margin of the fruiting bractlets denticulated.
Kastlösa, Öland, 1918 low of stature compared with most of the other forms; inflorescences short and contracted.
Slite, Gottland, 1920 low of stature; leaves and stems brightly anthocyan coloured. Illustrated in fig. 26.

The individuals of these cultivated biotypes differ markedly only in one respect from the plants in the natural habitat, viz. in size. The "wild" plants are dwarfs compared with the plants obtained in the cultures. The thickness of the leaves is often greater in the cultivated biotypes than in the plants from the natural habitat, apparently because of the better development of the former plants. The thickness of the leaf attained in the cultivated Kastlösa forms, for instance (prepared from the cultivated material, 1922) has not been found in any material from natural habitats, where the thickness of the leaves varies around 1000μ. In the Kastlösa form the leaf thickness has been found to attain 1155μ.

The forms making up this eastern type of A. sarcophyllum have been followed from the islands outside of Stockholm to Karlskrona in Blekinge (the province east of Scania). The group has been described under the specific name of A. praecox by HULPHERS (LINDMAN, 1918), and a form with contracted inflorescences apparently belonging to this group (and perhaps identical with my cultivated Kastlösa form) has been described by JOHANSON (1897) from Gottland under the name A. patulum f. globosum.


Fig. 27. Atriplex sarcophyllum from the Sound; field no. 45; ascending-erect form. Pot 12 cm. high.

b. The type of the Sound region.

When collecting material for my cultures in 1916 I was much struck by a small Atriplex form growing abundantly along the marshy shore line at Limhamn, near Malmö. It was already in flower in May and had thicker leaves than other forms of the genus growing in the same place. Several of these Limhamn plants were isolated and pure cultures were obtained, some of which are still being kept. The material has been increased by forms collected in their natural habitats from the following additional places: Vellinge, Klagshamn, Arlöv, Lomma, and Landskrona. A number of biotypes, differing in small points, have been isolated from this material. They resemble the forms of the eastern coast type in being early (they are even somewhat earlier than these), in having thick, fleshy leaves (see fig. 74 f) and with regard to the form of the leaves as well as of the inflorescences. There are, however, some important differences between the two groups of forms. Towards maturity the oldest fruiting bractlets in the inflorescences of these forms become large and leafy, and their petioles elongate, attaining a length of sometimes up to 3 cm. These characteristics are also seen in individuals collected in the natural habitat. Another point of difference from the preceding type is the colour of the leaves, which is almost always bluish-green in the praecox-type but dark green in the Sound type, both cultivated and "wild". The branching habit, which in the praecox-type is ascending-erect, is often procumbent, or even prostrate, in the Sound type. The margin of the leaves as well as of the bractlets may vary as in the case of praecox, but it is most often entire. Fig. 27 (field no. 45) shows one of the ascending-erect biotypes and fig. 28 (field no. 30) one of the prostrate ones, both isolated and pure cultured from the original Limhamn plants. It should be said that long-stalked fruiting bractlets are also developed when the plants are allowed to develop in sand culture.


Fig. 28. Atriplex sarcophyllum from the Sound; field no. 30. Prostrate form. Pot 12 cm. high.

 

1Castrations, unfortunately, cannot be made in the genus Atriplex because of technical difficulties.
Two individuals, one from a pure culture of field no. 45, the other a pure culture of the Kastlösa-praecox (see above) were isolated within the same pergamin bag in 1920 with the object of procuring hybrids between the two types1. A large number of seeds were obtained and sown, and 60 individuals from these cultures were isolated in 1921. Only one of these isolated plants has shown segregation this year, the rest are either typical no. 45 or typical Kastlösa-praecox. The segregating plant differed from the rest even in 1921, and aroused the suspicion of being a bastard-plant between the two types. It showed the branching habit of no. 45, but the fruiting bractlets resembled the praecox-type, with the exception of a few, which showed a tendency to become leafy. Fig. 29 shows this plant. The offspring of the plant shows segregation with regard to branching habit, bractlets, leaf colour etc., and a few of the individuals have much resemblance to the parent plants. A full account of this segregation, however, must be postponed until the next generation has been grown.

Fig. 29. Atriplex sarcophyllum. A F1-plant from the cross between the east coast and the Sound types.

The Sound type of A. sarcophyllum was held to be a mere modification in one of my former publications (TURESSON, 1919 b), as a result of insufficient experience in regard to the behaviour in culture of the types of the eastern and western coast strips. The hereditary nature of the characteristics peculiar to the Sound type is now beyond all doubt. The forms belonging to this type grow abundantly on the marshy coast strip along the Sound. A form belonging to the group was described as long ago as 1838 by DREJER from near Copenhagen under the specific name A. longipes. The name has disappeared from modern handbooks, or else it has been discarded with the remark that it refers to abnormal plants of other species, the abnormal characteristic meant being the long-stalked, leafy fruiting-bractlets. Probably the name A. patulum f. hololepis Fenzl., current in some of the systematic handbooks, also belongs to this type.

c. The type of the western coast strip.

Forms belonging to this type are already found in the region of Kullen, in north-western Scania, and they extend at least as far as to the middle part of Bohuslän. Cultivated material from along this coast strip shows a great hereditary variation. Individuals with leafy, long-stalked fruiting bractlets are entirely absent. The forms of this type show most resemblance to the praecox-type as to fruiting bractlets, but the branching habit is different. They are usually much branched at the base and most often prostrate both as to main axis and as to basal branches. A number of biotypes have been isolated since 1917, when the first individuals of this type were brought into culture. Table 5 characterizes the most important biotypes now in culture, and gives information as to the place and year of the collection of the original mother-plants.

TABLE 5.

Field no. Position of
main axis
Position of
basal branches
Colour of leaves Nature of
leaf margin
Place and year of coll.
of mother-plant
63 Ascending-erect Procumbent Blue-green Entire Marstrand, Boh., 1920
64 Procumbent " " " Varberg, Hall., 1919
65 " " Dark-green Denticulate "   "   "
66 Prostrate " " Entire "   "   "
67 " " Blue-green " Falkenberg, Hall., 1919
120 " Prostrate Dark-green " Hall. Väderö, Scania 1917
122 " " " Denticulate "   "   "   "
125 " " Blue-green Dentate "   "   "1918
128 " "   Entire "   "   "   "

Field nos. 63 and 64 are shown in figs. 30 and 31, and field nos. 120 and 128 in figs. 32 and 33. Some of the isolated plants gave an offspring which corresponded in every characteristic with their original mothers indicating autogamy, others showed segregation as to leaf colour, leaf margin, and branching habit. Self-fertilization seems to be the rule in most of the forms, however, as cases of vicinism are very rare in the plants sown from seed harvested from unprotected individuals in the cultures, where a number of different biotypes are grown close by.


Fig. 30. Atriplex sarcophyllum from the west coast; field n:o 63. Ascending-erect form. Pot 12 cm. high.

The biotypes tabulated above differ in a number of characters not indicated in the table. The width and length of the fruiting bractlets, for instance, vary with the different biotypes within narrow but distinct limits. The length of the fruiting bractlets of the biotypes represented by field nos. 63, 66 and 120 are given here. The fruiting bractlets are taken from one individual in the case of no. 63 and from three individuals (a, b, c) each of nos. 66, 120. The measurements were made in 1921 from the pure cultures, and 300 fruiting bractlets were taken from each individual. The distribution in the different length classes (in mm.) is as follows:

    2 4 6 8 10 12 14
no. 63       152 100 38 7 3
no. 66 a   112 159 25 4    
b   107 158 29 6    
c   104 147 35 12 2  
no. 120 a   65 195 30 6 2 2
b   72 180 29 10 5 4
c   63 170 37 15 9 6

Fig. 31. Atriplex sarcophyllum from the west coast; field no. 64. Procumbent form. Pot 12 cm. high.

Fig. 32. Atriplex sarcophyllum from the west coast; field no. 120. Prostrate form. Pot 12 cm. high.

There are also additional small points of difference between the different biotypes, although they are immaterial in this discussion. They all differ from the praecox-type in having a procumbent or a prostrate mode of growth. In rare cases the main axis may be nearly erect (cf. field 110. 63) but the basal branches show ill all cases a powerful development as compared with the main axis. The thickness of the leaves is about the same as in the Sound-type. The time of flowering is somewhat later than in the case of the longipes-  and praecox-types. It falls in the latter part of June, sometimes as late as in the first part of July.

In certain places on the western coast, where an abundance of other species of the genus, such as A. Babingtonii and A. latifolium, occur, hybrids between the west coast type and these latter species become so frequent that it is sometimes difficult to find typical individuals of the type in question. This is the more remarkable as cases of vicinism have been very few in my cultures (see above). Different forms of the type might behave differently as to the mode of fertilization. The relatively late flowering probably favours a certain degree of hybridism with other species. The western type flowers late enough to find particularly early individuals of other species flowering at the same time.


Fig. 33. Atriplex sarcophyllum from the west coast; field no. 128. Prostrate form. Pot 12 cm. high.

The northern limit of extension of the forms belonging to this western type has not yet been determined. They have been found to be common at least to the middle part of Bohuslän. Undoubted intermediates between this type and the Sound type have been found this year at Helsingborg. Their behaviour will be followed in cultures. As to the systematics of the forms belonging to the western type of A. sarcophyllum the name A. hastifolium Salisb. has been adopted to cover the group (by LINDMAN, 1918). The name A. patulum var. sarcophyllum Neum., covers some of the forms characteristic of the type.

5. MATERIAL AND CULTURES OF INLAND AND COAST FORMS OF SEDUM MAXIMUM (L.) SUTER.

The material of Sedum maximum comprises about 300 individuals transplanted from their natural habitats and now growing in the experimental field. They come from Bohuslän, from Halland, from several points in Scania, from Småland and from the islands off Stockholm. A few of the series cultivated since 1919 and 1920 (the years of transplantation) will be briefly dealt with below.


Fig. 34. Sedum maximum. The Sea cliff type of the west coast.

Fig. 35. Sedum maximum from Hallands Väderö.

My attention was first drawn to the plant in 1919 at Kullen in N. W. Scania. Small, stunted plants of the species grow here in the crevices of the exposed sea cliffs. The leaves of these individuals were found to be folded and brightly anthocyanic in colour. They were at that time thought to be modifications of the green or glaucus, flat-leaved inland-maximum called forth by the extreme habitat conditions of the sea cliff. A few of the cliff individuals were excavated and brought home in order that the reversion to the inland form might be followed more closely. This reversion has not yet taken place, however. The plants have, on the contrary, developed their original peculiarities still more with increasing age and growth. Fig. 34 (to the right) shows one of the plants, once a stunted, single stemmed dwarf, now one of the largest in culture. The leaves are permanently dark red in colour and much folded, as is plainly seen from the photograph. A second series from this locality composed of 26 individuals was brought home in 1920. The slight variation within this series, as seen in 1922, may he briefly noted: the number of flat-leaved individuals is two, the rest have folded leaves; the pure green-leaved individuals number 6, the rest have dark red leaves. The series further includes plants with ascending and with erect stems (fig. 34, to the left) in about equal number.


Fig. 36. Sedum maximum from Stenshuvud (to the left) and from S:t Olof (to the right).

It is thus seen that when a greater number of plants is taken, individuals with the characteristics of the inland form are also found to occur. It is, however, a point of much interest that these latter occur in a decided minority. The same has been found to he true of cliff series from Varberg in Halland and from Bohuslän. The Sedum maximum vegetation of the rocky shores of Hallands Väderö, which would be expected to come close to that of Kullen, is made up of individuals the majority of which are flat-leaved and glaucus. It should be remembered, how-ever, that the rich, herbiforous woods making up the interior of that island and in close touch with the rocky shore include a multitude of Sedum maximum much resembling in certain characters the ordinary inland type, which, no doubt, incessantly recruits the shores with individuals of its own type. The leaves of the plants from the woods have a typical shade-leaf structure, as has already been pointed out by LUNDEGÅRDH (1919), while the plants from the rocky shores have sun-leaf structure. The shade-leaf structure of the woodland plants is only a modification, which had entirely disappeared in 1922 in all individuals brought home from the island in 1919. Fig. 33 shows two plants, one prostrate, the other almost erect, both belonging to this woodland series.

The folded leaf maximum typical of the exposed sea cliffs of the west coast has also been found on the east coast of Scania, viz. on the cliffs at Stenshuvud. The folded individuals, however, are here mixed with flat-leaved in about equal proportions. The folding, further, is not so pronounced in the Stenshuvud plants as in the west coast plants. Fig. 36 (to the left) shows a leaf-folded individual from Stenshuvud. The colour of the leaves of these plants is pure green, not anthocyanic except towards autumn. The Stenshuvud series is thus seen to be rather different from the west coast series, even if certain characteristics of the west coast cliff type may he traced in some of the individuals.

The series cultivated and examined from further up the east coast, viz. from Oskarshamn and Västervik in Småland and from the islands off Stockholm have only been found to comprise rather long-stemmed, flat-leaved individuals, the majority of which have been found to be glaucus. Some series, for instance those from Stockholm, show all kinds of branching habit. Fig. 37 illustrates 3 plants belonging to a series of 15 brought home in 1920 from Nämndö (an island off Stockholm). The prostrate plant is the only one in the collection, the erect habit characterizes a few, while the ascending habit predominates.

Inland maximum has been collected in several places in Scania.