of the dissolved substances to pass through them. This is, of course, true of all types of plant, for otherwise they would be unable to obtain the small quantities of other mineral salts essential to their growth. Many halophytes, however, absorb very large amounts of common salt; the ash of some of them, like the glassworts, was formerly used on a large scale to provide soda for glass-making, and certain plants, as for example thrift, actually excrete surplus salt from the glands on their leaves. That the development of succulent leaves is closely connected with the absorption of salt is borne out by the behaviour of many non-halophytes when they grow in places exposed to sea-water. Many inland plants found on open beaches or on cliffs within reach of sea-spray possess much more fleshy leaves than they have in their normal habitats; bird’s-foot trefoil (Lotus corniculatus), kidney vetch (Anthyllis vulneraria) and the greater knapweed (Centaurea scabiosa) are species which often show this effect. Some years ago I analysed the ash of certain inland plants which had been exposed to sea-spray in this way and found 13.5 per cent of salt in that of the kidney vetch. Evidence such as this points strongly to the conclusion that some chemical action connected with salt is the primary cause of succulence in halophytes, and there seems little reason to associate it with the problem of conserving water. The similarity between succulent xerophytes and halophytes is remarkable and we must leave it at that.
FIG. 9.—Comparison of root-systems of annual glasswort and rice-grass:-a. Small plant of annual glasswort; b. Two joints from stem of annual glasss wort enlarged to show leaf-scales; c. Base of a rice-grass plant showing vertical anchoring roots and horizontal feeding roots and stolons (Fig. c from Tansley after Oliver, 1926).
ROOT-SYSTEMS
The roots of coastal plants are very characteristic and present many points of interest. The majority of true halophytes (i.e. plants which normally grow where the soil-water is saline) possess very deep roots, generally markedly woody. Most of these plants are perennials and the chief value of their long roots in a salt-marsh is to enable them to secure a firm anchorage in relatively unstable mud. It also allows them to derive their main water supplies from regions where the concentration of salt is less variable than it is in the surface layers. Annual glasswort (Salicornia stricta) (Pl. XIII) possesses only quite short roots, and as a result is liable to become dislodged from the unstable mud if there is a strong tidal flow (Fig. 9(a)). This is in marked contrast to rice-grass (Spartina townsendii) (Pl. XIV), which occupies much the same position as a pioneer colonist in many of the south coast salt-marshes. This plant develops a most extensive root-system and becomes so firmly anchored in the mud that it can easily resist the strong currents produced by the ebb and flow of the tides (Fig. 9(c)). Its remarkable powers of spreading over soft mud and stabilising the surface are described more fully on page 71. Plants growing on sea-cliffs also develop very long roots, which serve the dual purpose of anchoring them firmly against uprooting by the violent winds encountered in these exposed places and of enabling them to tap deep-seated supplies of water. Well-established plants of thrift or samphire (Crithmum maritimum) (Pl. 10) frequently possess roots several feet long, which penetrate deeply into the crevices between the rocks.
Extensive root-systems are also a characteristic feature of xerophytes in all parts of the world. In sand-dunes and shingle this enables the plants to utilise the moisture which is always present some way below the surface (see here). In addition, the elaborate root-systems developed by many dune plants perform the important function of binding blown sand. All pioneer colonists on sand-dunes, and to a lesser extent those on mobile shingle, have to contend with the possibility of being periodically swamped by loose sand or shingle. Most of them have, in varying degrees, the ability to form fresh shoots easily when they are submerged in this way, and to grow up through this covering. Marram-grass easily outstrips all other plants in the vigour with which it can do this. When once established in loose sand, it soon produces a mass of underground runners from which new shoots continually spring. Where these new shoots occur, fresh adventitious roots are produced under them; as the stems and leaves become buried by sand, further shoots and leaves are produced at a higher level on the stem (Fig. 10). The thick tufts of leaves and young shoots are very stiff and offer a considerable obstruction to the wind, causing it to drop some of its load of sand round them. Provided the blown sand does not accumulate too rapidly over the plants, marram-grass can continue to grow upwards through many feet of sand. In this way dunes rising to considerable heights can be produced, their size depending on the supply of available sand and the strength of the prevailing winds. The whole interior of the dune remains closely penetrated by a mass of rhizomes (underground stems) and fine roots which bind it together. As these rhizomes and roots are constantly being renewed at higher levels, the lower ones gradually die off, but their dead remains persist in the lower regions of a dune for a long time and continue to exercise a stabilising effect on it (Pl. XXI).
FIG. 10—Leaf production at different levels shown by marram-grass (from) Firistch & salisbury, 1946).
To a lesser extent the sea couch-grass (Agropyron junceiforme) can produce a similar result. This plant has much the same habit as ordinary couch-grass or twitch (Agropyron repens), which is an all too familiar agricultural weed. Like marram-grass, it produces a mass of rhizomes from which new shoots spring up at frequent intervals, but its runners tend to spread more rapidly in a horizontal direction than vertically when covered by loose sand. As a result, the dunes it produces are comparatively low compared with those formed by marram-grass. None the less, its powers of binding sand are considerable, and single plants have been shown to cause dunes as much as 20 feet across during a few years’ growth. The sea lyme-grass (Elymus arenarius) has a similar root-system and occasionally forms low dunes of the same type on certain parts of the coast.
Many other pioneer plants on shifting sand can accumulate small amounts of sand round them to form miniature dunes, provided they are sufficiently virile to shoot up again when they become buried. For example, the sea-sandwort (Honckenya (Arenaria) peploides) is a low-growing plant not more than a few inches high, but it possesses surprisingly extensive creeping roots and readily produces fresh shoots when it is covered. Even those common pioneers of sandy beaches, the sea-rocket (Cakile maritima) (Pl. VII) and the prickly saltwort (Pl. I), which are only annuals, can collect tiny dunes round their long trailing branched stems. Their dead remains usually persist for a considerable time in the winter and continue to hold the sand, although the principal agent here is the stem rather than the roots.
Marram-grass has relatively little stabilising effect on the surface sand, and it is the later colonists which establish themselves between the clumps that are responsible for its eventual consolidation. Notable amongst these are the sand-sedge (Carex arenaria) and the sand-fescue (Festuca rubra var: arenaria), both of which produce horizontally creeping roots just below the surface. Tufts of foliage arise from these runners at frequent intervals, and in the case of the former often appear spaced out along a nearly straight line for many feet, showing clearly the course of its immense roots (Pl. X).
Plants growing on shingle similarly have to endure periodical swamping by stones. Very unstable shingle is usually devoid of vegetation, but during storms those less mobile parts of the beach, which carry quite a large amount of vegetation, may also
