SUCCESSION

288. Concept. Succession is the phenomenon in which a series of invasions occurs in the same spot. It is important, however, to distinguish clearly between succession and invasion, for, while the one is the direct result of the other, not all invasion produces succession. The number of invaders must be large enough, or their effect must be sufficiently modifying or controlling to bring about the gradual decrease or disappearance of the original occupants, or a succession will not be established. Partial or temporary invasion can never initiate a succession unless the reaction of the invaders upon the habitat is very great. Complete and permanent invasion, on the other hand, regularly produces successions, except in the rare cases where a stable formation entirely replaces a less stable one without the intervention of other stages. Succession depends in the first degree upon invasion in such quantity and of such character that the reaction of the invaders upon the habitat will prepare the way for further invasion. The characteristic presence of stages in a succession, which normally correspond to formations, is due to the peculiar operation of invasion with reaction. In the case of a denuded habitat, for example, migration from adjacent formations is constantly taking place, but only a small number of migrants, especially adapted to somewhat extreme conditions, are able to become established in it. These reach a maximum development in size or number, and in so doing react upon the habitat in such a way that more and more of the dormant disseminules present, as well as those constantly coming into it, find the conditions favorable for germination and growth. The latter, as they in turn attain their maximum, cause the gradual disappearance of the species of the first stage, and at the same time prepare the way for the individuals of the succeeding formation. It is at present impossible to determine to what degree this substitution is due to the struggle for existence between the individuals of each species and between the somewhat similar species of each stage, and to what degree it arises out of the physical reaction.

It is evident that geological succession is but a larger expression of the same phenomenon, dealing with infinitely greater periods of time, and produced by physical changes of such intensity as to give each geological stage its peculiar stamp. If, however, the geological record were sufficiently complete, we should find unquestionably that these great successions merely represent the stable termini of many series of smaller changes, such as are found everywhere in recent or existing vegetation.

289. Kinds of succession. The fundamental causes of succession are invasion and reaction, but the initial causes of a particular succession are to be sought in the physical or biological disturbances of a habitat or formation. With reference to the initial cause, we may distinguish normal succession, which begins with nudation, and ends in stabilization, and anomalous succession, in which the facies of an ultimate stage of a normal succession are replaced by other species, or in which the direction of movement is radically changed. The former is of universal occurrence and recurrence; the latter operates upon relatively few ultimate formations. In the origin of normal successions, nudation may be brought about by the production of new soils or habitats, or by the destruction of the formation which already occupies a habitat. In a few cases, the way in which the habitat arises or becomes denuded is not decisive as to the vegetation that is developed upon it, but as a rule the cause of nudation plays as important a part in the development of a succession as does the reaction exerted by the invaders. The importance of this fact has been insisted upon under invasion. New soils present extreme conditions for ecesis, possess few or no dormant disseminules, and in consequence their successions take place slowly and exhibit many stages. Denuded soils as a rule offer optimum conditions for ecesis as a result of the action of the previous succession, dormant seeds and propagules are abundant, and the revegetation of such habitats takes place rapidly and shows few stages. The former may be termed primary succession, the latter secondary succession.

PRIMARY SUCCESSIONS

290. These arise on newly formed soils, or upon surfaces exposed for the first time, which have in consequence never borne vegetation before. In general they are characteristic of mountain regions, where weathering is the rule, and of lowlands and shores, where sedimentation or elevation constantly occur. The principal physical phenomena which bring about the formation of new soils are: (1) elevation, (2) volcanic action, (3) weathering, with or without transport.

291. Succession through elevation. Elevation was of very frequent occurrence during the earlier, more plastic conditions of the earth, and the successions arising as a result of it must have been important features of the vegetation of geological periods. To-day, elevation is of much less importance in changing physiography, and its operation is confined to volcanic islands, coral reefs, and islets, and to rare movements or displacements in seacoasts, lake beds, shore lines, etc. There has been no investigation of the development of vegetation on islands that are rising, or have recently been elevated, probably because of the slow growth of coral reefs and the rare appearance of volcanic islands. On coral reefs, the first vegetation is invariably marine, but as the reef rises higher above the surf line and the tide, the vegetation passes into a xerophytic terrestrial type adapted to an impervious rock soil, and ultimately becomes mesophytic. In volcanic islands, unless they are mere rocks over which the waves rush, the succession must always begin with a xerophytic rock formation. The best known example of a rising coast line is found in Norway and Sweden, where the southeastern coast is rising at the rate of five or six feet a century. There can be little question that such changes of level will produce marked changes in vegetation, but the modification will be so gradual as to be scarcely perceptible in a single generation. It is probable that the forests of the Atlantic coastal plains are the ultimate stages of successions initiated at the time of the final elevation of the sea bottom along the coast line.

Fig. 60. A lichen formation (Lecanora-Physcia-petrium), the first stage of the typical primary succession (Lecanora-Picea-sphyrium) of the Colorado mountains.

292. Succession through volcanic action. The deposition of volcanic ashes and flows of lava are relatively infrequent at present, occurring only in the immediate vicinity of active volcanoes, chiefly in or near the tropics. Successions of this sort are in consequence not only rare, but they are also relatively inaccessible to investigators. They have been studied in a few cases, for example, those of Krakatoa by Treub, but this study has been confined to the general features of revegetation. Ash fields and lava beds are widely different in compactness, but they agree in having a low water- and nutrition-content. The pioneer plants in both will be intense xerophytes, but the soil differences will determine that these shall be sand-binders in the former, and rock-weathering plants in the latter.

293. Weathering. Practically all primary successions start on soils produced by weathering. This is also true of coral or volcanic islets and of lava beds, for no terrestrial vegetation can secure a foothold upon them until the surface of the rock has been to some extent decomposed or disintegrated. Weathering, as is well known, consists of two processes, disintegration and decomposition, which usually operate successively, though they are sometimes concomitant. Disintegration usually precedes, especially in rock masses, and unless it is soon followed by decomposition, results in dysgeogenous soils. Decomposition often goes hand in hand with disintegration, or it takes place so rapidly and perfectly that it alone seems to be present. In either case, the resulting soil is eugeogenous. The relation of decomposition to disintegration determines the size and compactness of the soil particles, and upon the latter depend the porosity, capillarity, and hygroscopicity of the soil. These control in large degree the character of the first vegetation to appear on the soil.

Another point of fundamental value in determining revegetation is the disposition of the weathered rock. If it remains in situ, it will evidently differ in respect to compactness, homogeneity, nutrition-content, water-content, disseminules, etc., from weathered material which has been transported. An essential difference also arises from the fact that a rock may be weathered a long distance from the place where the decomposed particles are finally deposited, and in the midst of a vegetation very different from that found in the region of deposit. The disposition of the weathered material affords in consequence a satisfactory basis for the arrangement of primary successions. The following classification is proposed, based upon the soil groups established by Merrill.[[39]]

294. Succession in residuary soils. Residuary soils are always sedentary, i. e., they are formed in situ. They show certain differences dependent upon the rock from which they originate, which may be mixed crystalline shale, sandstone, or limestone, but the thoroughness of decomposition causes these differences to be comparatively small. Residuary soils are typically eugeogenous; their successions in consequence usually begin with mesophytes, and consist of a few stages. The soluble salt-content is comparatively low, since all soluble matters are readily leached out. Successions in these soils are especially characteristic of shale, sandstone, and limestone ledges or banks. Cumulose deposits, like residuary ones, are sedentary in character, but as they are produced by the accumulation of organic matter, they will be considered under reactions of vegetation upon habitat.

Fig. 61. Talus arising from the disintegration of a granitic cliff; the rocks are covered with crustose lichens.

295. Succession in colluvial soils. Colluvial deposits owe their aggregation solely or chiefly to the action of gravity. They are the immediate result of the disintegration of cliffs, ledges, and mountain sides, decomposition appearing later as a secondary factor. The masses and particles arising from disintegration are extremely variable in size, but they agree as a rule in their angular shape. The typical example of the colluvial deposit is the talus, which may originate from any kind of rock, and contains pieces of all sizes. Gravel slides differ from ordinary talus in being composed of more uniform particles, which are worn round by slipping down the slope in response to gravity and surface wash. Boulder fields are to be regarded as talus produced by weathering under the influence of joints, resulting in huge boulders which become more and more rounded under the action of water and gravity. This statement applies to those fields which are in connection with some cliff that is weathering in this fashion; otherwise, boulder fields are of aqueous or glacial origin. The character of the successions in talus will depend upon the kind of rock in the latter. If the rock is igneous or metamorphic, decomposition will be slow, and the soil will be dysgeogenous. Successions on such talus consist of many stages, and the formations are for a long time open and xerophytic. In talus formed from sedimentary rocks, especially shales, limestones, and calcareous sandstones, decomposition is much more rapid, and the successions are simpler and more mesophytic.

296. Succession in alluvial soils. Alluvial soils are fluvial when laid down by streams and rivers, and litoral when washed up by the waves or tides. They are formed when any obstacle retards the movement of the water, decreasing its carrying power, and causing the deposit of part or all of its load. They consist of more or less rounded, finely comminuted particles, mingled with organic matter and detritus. Alluvial deposits are especially frequent at the mouth of streams and rivers, on their terraces and flood plains, and along silting banks as compared with the erosion banks of meanders. The filling of ponds by the erosion due to surface drainage, and of lakes by the deposition of the loads of streams that enter them, results in the formation of new alluvium. A similar phenomenon occurs along coasts, where bays and inlets are slowly converted into marshes in consequence of being shallowed by the material washed in by the waves and tides. Such paludal deposits are invariably salt water or brackish. Contrasted with these, which are uniformly black in consequence of the large amount of organic matter present, are the sandbars and beaches, which, though due to the same agents, are light grey or white in color, because of the constant leaching by the waves. Two kinds of alluvial deposits may accordingly be distinguished: (1) those black with organic matter, and little disturbed by water, and (2) those of a light color, which are constantly swept by the waves. The successions corresponding to these are radically different. In the first, the pioneer vegetation is hydrophytic, consisting largely of amphibious plants. The pioneer stages retard the movement of the water more and more, and correspondingly hasten the deposition of its load. The marsh bed slowly rises in consequence, and finally the marsh begins to dry out, passing first into a wet meadow, and then into a meadow of the normal type. A notable exception to this sequence occurs when the swamp contains organic matter or salts in excess, in which case the vegetation consists indefinitely of swamp xerophytes, or halophytes. The first vegetation on fresh water sandbars is xerophytic, or, properly, dissophytic, unless they remain water-swept, and the ultimate stages of their successions are mesophytic woodlands composed of water-loving genera, Populus, Salix, etc. It seems certain, however, that these will finally give way to longer-lived hardwoods. Maritime sandbars and beaches are always saline, and their successions run their short course of development entirely within the group of halophytes, unless the retreat of the sea or freshwater floods change the character of the soil. The chemical action of underground waters also produces new soils, which might be classed as alluvial. These soils are essentially rock deposits, travertine, silicious sinter, etc., made by iron and lime springs and by geysers, and they must be changed by decomposition into soils proper to be comparable with alluvial soils.

Fig. 62. Talus arising from the decomposition of granite; the gravel is covered with a formation of foliose lichens (Parmelia-chalicium), the second stage of the primary talus succession; the herbs are pioneers of the next stage.

297. Succession in aeolian soils. The only wind-borne soils of geological importance at the present time are those which form dunes, both inland and coastal. Aeolian deposits consist largely of rounded sand particles, which are of almost uniform size in any particular dune, but vary greatly in dunes of different ages. The reaction of the pioneers on dunes plays an important part in building the latter, but the immense dunes of inland deserts, which are entirely destitute of vegetation, seem to indicate that its value has been overestimated. The first stages in dune successions are dissophytic, i. e., the plants grow in a soil of medium or high water-content, but in an atmosphere that is extremely xerophytic. The ultimate stages vary widely in accordance with the region in which they occur; they may be xerophytic heaths or mesophytic meadows and forests. Because of their striking character and economic significance, dunes have received much attention, with the result that their successions are the most thoroughly known of all. Prairie and steppe formations are probably to be regarded as the ultimate stages of successions established on wind-borne loess, and it is possible that the same is true of sand-hill vegetation in the prairie province.

298. Succession in glacial soils. The formation of glacial deposits is at present confined to alpine and arctic regions. Recent successions in such soils are localized in these regions, and are in consequence relatively unimportant. There can be little question, however, that the thorough investigation of succession in and near the moraines of existing glaciers will throw much light upon the successions of the glacial period. Moraines, drumlins, eskars, and alluvial cones represent the various kinds of glacial deposits. They agree in being heterogeneous in composition, and are covered to-day with ultimate stages of vegetation, except in the immediate vicinity of glaciers.

SECONDARY SUCCESSIONS

299. Generally speaking, all successions on denuded soils are secondary. When vegetation is completely removed by excessive erosion, it is an open question whether the resulting habitat is to be regarded as new or denuded. Erosion is rarely so extreme and so rapid, however, as to produce such a condition, even when it results from cultivation or deforestation. It is, moreover, especially characteristic of newly formed soils, and in studying succession in eroded habitats, it is fundamentally important to determine whether erosion has produced denudation, or has operated upon a new soil. The great majority of secondary successions owe their origin to floods, animals, or the activities of man, and they agree in occurring upon decomposed soils of medium water-content, which contain considerable organic matter, and a large number of dormant migrants. These successions consist of relatively few stages, and are rarely of extreme character.

300. Succession in eroded soils. Eroded soils show considerable differences, as they arise in consequence of erosion by water or by wind, though the initial stages of revegetation derive their character more from the aggregation of the soil than from the nature of the erosive agent. Eroded soils are as a rule xerophytic. In the case of erosion by water, dysgeogenous soils are readily worn away in consequence of their lack of cohesion, as in sand draws, etc., while eugeogenous soils are easily eroded only on slopes, as in the case of ravines, hillsides, etc. In the former, the extreme porosity and slight capillarity of the sand and gravel result in a low water-content. In the finer soils, the water-content is also low, on account of the excessive run-off, due to compactness of the particles and to the slope. The erosive action of winds upon soils bearing vegetation is not very general; it is found to some extent in more or less established dunes, and exists in a marked degree in buttes, mushroom rocks, and blowouts. The first two are regularly xerophytic, the last as a rule, dissophytic. The early stages of successions in eroded soils are composed of xerophytes. In loose soils, these are forms capable of binding the soil particles together, thus preventing wash, and increasing the accumulation of fine particles, especially of organic matter. In compact soils, the effect is much the same; the pioneers not only decrease erosion, but at the same time also increase the water-content by retarding the movement of the run-off.

301. Succession in flooded soils. The universal response of vegetation to floods is found in the amphibious plant, which is a plastic form capable of adjustment to very different water-contents. Floods are confined largely to river basins and coasts. In hilly and mountainous regions, where the slope is great, any considerable accumulation of flood waters is now impossible, although of frequent occurrence when land forms were more plastic.

In all streams that have become graded, the fall is insufficient to carry off the surplus water in the spring when snows are melting rapidly, or at times of unusual precipitation. These waters accumulate, and, overflowing the banks, spread out over the lowlands, resulting in the formation of a well-defined flood plain. This is a periodical occurrence with mature streams, and it occurs more or less regularly with all that are not torrent-like in character. The effect of the overflow is to destroy or to place at a disadvantage those plants of the flood plain that are not hydrophytes. At the same time, a thin layer of fresh silt is deposited upon the valley floor of sand or alluvium. Flooding is most frequent and of longest duration near the banks of the stream. It extends more or less uniformly over the flood plain, and disappears gradually or abruptly as the latter rises into the bench above. Floods destroy vegetation and make a place for secondary successions by drowning out mesophytic species, by washing away the aquatic forms of ponds and pools, and by the erosion of banks and sandbars. They affect the amphibious vegetation of swamp and shore to a certain extent, but, unless the period of flooding is long, they tend to emphasize such formations rather than to destroy them. The still-water formations of many cutoff and oxbow lakes owe their origin to a river which cuts across a meander in time of flood. This result is more often attained by the alternate silting and erosion of a meandering river by which it cuts across a bend in its channel. The usual successions in flooded lands are short as a rule; amphibious algae, liverworts, and mosses soon give way to ruderal plants, and these in turn to the original mesophytes of meadows, or dissophytes of sandbars. In the case of ponds and pools, the process of washing-out or silting up merely removes or destroys the vegetation, without effectively modifying the habitat, and the secondary successions that follow are extremely short.

302. Succession by subsidence. Subsidence is a factor of the most profound importance in changing vegetation. It operates over vast areas through immense periods of time. For these reasons, the changes are so slow as to be almost imperceptible, and the resulting successions can be studied only in the geological record. Extensive subsidence is confined to-day to coastal plains, as in Greenland, the south Atlantic coast, and the region of the Mississippi delta, where its effects are merged with the paludation of tidal rivers, and the wave and tide erosion of the sea shore. Such successions are unique, inasmuch as the denuding force operates very slowly instead of quickly, and the first pioneers of the new vegetation appear before the original formation has been destroyed. In all cases, the succession is from mesophytic or halophytic formations to paludose, and, finally, marine vegetation. In small areas of subsidence, such as shore slips along lakes and streams, sink holes, and sunken bogs, the succession is usually both short and simple, mesophytes giving place to amphibious and ultimately to aquatic forms.

303. Successions in landslips. Landslips occur only in montane and hilly regions, and here they are merely of local importance. In many respects, they are not unlike talus; they show essential differences, however, in that they are not sorted by gravity, and in that they destroy vegetation almost instantly. The succession arises as a rule, not upon the original soil, but upon that of the landslip, and, as pointed out elsewhere, might well be regarded as primary.

304. Succession in drained, or dried soils. In geological times, the subsidence of barriers must often have produced drainage and drying-out, just as elevation frequently resulted in flooding and lake formation. At the present time, the drying-out of lakes and ponds is the result of artificial drainage, or of climatic changes. The former will be considered under successions brought about by the agency of man. Climatic changes when general operate so slowly that the stages of such successions are perceptible only when recorded in strata. More locally, climate swings back and forth through a period of years, with the result that in dry years the swamps and ponds of wetter seasons are dried out, and the vegetation destroyed or changed. If the process be gradual, the succession passes from hydrophytic through amphibious to mesophytic, and, in dry regions, xerophytic conditions. When the process of drying-out occurs rapidly, as in a single summer, the original formation is destroyed, and the new vegetation consists largely of ruderal plants. A peculiar effect of climate occurs in regions with poor drainage, where the result of intense evaporation is to produce alkaline basins and salt lakes, in which the succession becomes more and more open, and is finally represented by a few stabilized halophytes, or disappears completely.

Fig. 63. A typical gravel slide (talus) of the Rocky mountains, before invasion.

305. Succession by animal agency. Successions of this class are altogether of secondary importance, the instances in which animals produce denudation being relatively few. Such are the heaps of dirt thrown up by prairie dogs and other burrowing animals, upon which ruderal plants are first established, to be finally crowded out by the species of the original formation. Buffalo wallows furnish examples of similar successions in which the initial stages are subruderal, while overstocking and overgrazing frequently produce the same result with ruderal plants.

306. Succession by human agency. The activities of man in changing the surface of the earth are so diverse that it is impossible to fit the resulting successions in a natural system. While man does not exactly make new soils, he exposes soils in various operations: mining, irrigation, railroad building, etc. He destroys vegetation by fires, lumbering, cultivation, and drainage, and if he can not control climate, he at least modifies its natural effects by irrigation and the conservation of moisture. The operations of man extend from seacoasts and swampy lowlands through mesophytic forests and prairies to the driest uplands and inlands. Since the adjacent formations determine in large degree the course and constitution of a succession, it will be seen that the effects of any particular activity upon vegetation will differ greatly in different regions. For convenience, all classes of successions arising from the presence and activity of man will be considered in this place, though, as indicated above, some might well be regarded as producing primary successions, while others produce anomalous ones.

307. Succession in burned areas. It will suffice merely to point out that “burns” may arise naturally through lightning, volcanic cinders, lava flows, etc., but the chances are so slight that these causes may be ignored. The causes of fires are legion, and as they have little or no effect upon results, they need not be considered. From their nature, fires are of little significance in open vegetation, deserts, polar barrens, alpine fields, etc., since the area of the burn can never be large. In closed formations, the extent of fires is limited only by the area of the vegetation, and the effect of wind, rain, and other forces. Forest fires usually occur during the resting period, except in the case of coniferous forests. In grassland, the living parts are underground during autumn and winter, when prairie fires commonly occur. As a consequence, the repeated annual burning of meadow or prairie does not result in denudation and subsequent succession. On the contrary, it acts in part as a stabilizing agent, inasmuch as it injures the typical vegetation forms of grassland much less than it does the woody invaders. All formations with perennial parts above ground, viz., thicket, open woodland, and forest, are seriously injured by fire. A severe general fire destroys the vegetation completely; a local fire destroys the formation in restricted areas; while a slight or superficial burn removes the undergrowth and hastens the disappearance of the weaker trees. In the latter case, while the primary layer of the forest remains the same, succession takes place in the herbaceous and shrubby layers. These successions are peculiar in that they are composed almost wholly of the proper species of the forest, and that they are very short, showing only a few poorly defined stages. A local fire initiates a succession in which the pioneers are derived largely from the original formation, particularly when the latter encloses the burned area more or less completely. The constitution of the intermediate and ultimate stages will depend in a larger degree still upon the size and position of the burn. When a particular formation is destroyed wholly or in large part, the first stages of the new vegetation are made up by invaders from the adjacent formations. In the most perfect types of succession, this dissimilarity between the new and the old vegetation continues to the last stage, in which the reappearance of the facies precedes that of the subordinate layers. In many forest successions, however, the general physical similarity of the ultimate stages permits the early reappearance of the herbaceous and shrubby species, and the final stages affect the facies alone. Successions in burned areas operate usually within the water-content groups. The reconstruction of a mesophytic forest takes place by means of mesophytes; of the rarer xerophytic and hydrophytic forests, through xerophytes and hydrophytes respectively. This is due to the fact that the alteration of the soil is slight, except where the burning of the vegetation permits the entrance of erosion, as on mountain slopes.

Fig. 64. Gravel slide formation (Pseudocymopterus-Mentzelia-chalicium), stage III of the talus succession.

308. Succession in lumbered areas. Commercial lumbering, especially where practiced for wood-pulp as well as for timber, results in complete or nearly complete destruction of the vegetation by removal and the change from diffuse light to sunlight, or by the action of erosion upon the exposed surface. In the first place, short mesophytic successions will result; in the second, the successions will be long and complex, passing through decreasingly xerophytic conditions to a stable mesophytic forest. Where a forest is cut over for certain species alone, the undisturbed trees soon take full possession, though the causes effective in the beginning will ultimately restore the original facies in many instances. Such successions are anomalous, and will be treated under that head.

309. Succession by cultivation. The clearing of forests and the “breaking” of grassland for cultivation destroy the original vegetation; the temporary or permanent abandonment of cultivated fields then permits the entrance of ruderal species, which are the pioneers of new successions. This phenomenon takes place annually in fields after harvest, resulting in the secondary formations of Warming, in which practically the same species reappear year after year. In fields that lie fallow for several years, or are permanently abandoned, the first ruderal plants are displaced by newcomers, or certain of them become dominant at the expense of others. In a few years, these are crowded out by invaders from the adjacent formations, and the field is ultimately reclaimed by the original vegetation, unless this has entirely disappeared from the region. The number of stages depends chiefly upon whether the final formation is to be grassland or woodland. Other activities of man, such as the construction of buildings, roads, railways, canals, etc., remove the native vegetation, and make room for the rapid development of ruderal formations. In and about cities, where the original formations have entirely disappeared, the chance for succession is remote, and the initial ruderal stages become more or less stabilized. Elsewhere the usual successions are established, and the ruderal formation finally gives way to the dominant type. In mountain and desert regions, where ruderal plants are rare or lacking, their place is taken by subruderal forms, species of the native vegetation capable of rapid movement in them. These, like ruderal plants, are gradually replaced by other native species of less mobility, but of greater persistence, resulting in a short succession operating often within a single formation. From the nature of cultivated plants, succession after cultivation generally operates within the mesophytic series.

310. Succession by drainage. Successions of this kind show much the same stages as are found in those due to flooding. They proceed from aquatic or swamp formations to mesophytic termini, either grassland or woodland. When drainage takes place rapidly and completely, the pioneer stages are usually xerophytic; cases of this sort, however, are infrequent.

311. Succession by irrigation. Irrigation produces short successions of peculiar stamp along the courses of irrigating canals and ditches, and in the vicinity of reservoirs. These are recent, as a rule, and are usually found in the midst of cultivated lands, so that their complete history is still a matter of conjecture. The original xerophytes are forced out not only by the disturbance of the soil, but also by its increased water-content. A few of them often thrive under the new conditions, and, together with the usual ruderal plants and a large number of lowland mesophytes and amphibious forms derived from the banks of the parent stream, constitute a heterogeneous association. This is doubtless to be regarded as an initial stage of a succession, but it is an open question whether the succession will early be stabilized as a new formation, or whether the original vegetation will sooner or later be reestablished under somewhat mesophytic conditions. From the number of mesophytes and from the behavior of valleys, it seems certain that the banks of such canals will ultimately be occupied by a formation more mesophytic than hydrophytic, into which some of the surrounding xerophytes of plastic nature have been adopted.

312. Anomalous successions are those in which the physical change in the habitat is relatively slight, resulting in a displacement of the ultimate stage, or the disturbance of the usual sequence, merely, instead of the destruction and reconstruction of a formation, or the gradual development of a new series of stages on new soil. In nature, the ultimate grass or forest stage of a normal succession is often replaced by a similar formation, especially if the facies be few or single. It is evident that certain trees naturally replace others in the last stages of a forest succession, without making the latter anomalous. The last occurs only when a normal stage is replaced by one belonging properly to an entirely different succession, as when a coniferous forest replaces a deciduous one in a hardwood region. The presence and development of such successions can be determined only after the normal types are known. The interpolation of a foreign stage in a natural succession, or a change of direction, by which a succession that is mesotropic again becomes hydrophytic, is easily explained when it is the result of artificial agents, as is often the case. In nature, anomalous successions are commonly the result of a slow backward and forward swing of climatic conditions.

313. Perfect and imperfect successions. A normal succession will regularly be perfect; it passes in the usual sequence from initial to ultimate conditions without interruption or omission. Imperfect succession results when one or more of the ordinary stages is omitted anywhere in the course, and a later stage appears before its turn. It will occur at any time when a new or denuded habitat becomes so surrounded by other vegetation that the formations which usually furnish the next invaders are unable to do so, or when the abundance and mobility of certain species enable them to take possession before their proper turn, and to the exclusion of the regular stage. Incomplete successions are of great significance, inasmuch as they indicate that the stages of a succession are often due more to biological than to physical causes, the proximity and mobility of the adjacent species being more determinative than the physical factors. Subalpine gravel slides regularly pass through the rosette, mat, turf, thicket, woodland, and forest stages; occasionally, however, they pass immediately from the rosette, or mat condition, to an aspen thicket which represents the next to the last stage. Such successions are by no means infrequent in hilly and montane regions; in regions physiographically more mature or stable, perfect successions are almost invariably the rule.

Fig. 65. Half gravel slide formation (Elymus-Muhlenbergia-chalicium), stage IV of the talus succession.

314. Stabilization. It may be stated as a general principle that vegetation moves constantly and gradually toward stabilization. Each successive stage modifies the physical factors, and dominates the habitat more and more, in such a way that the latter seems to respond to the formation rather than this to the habitat. The more advanced the succession, i. e., the degree of stabilization, the greater the climatic or physiographic change necessary to disturb it, with the result that such disturbances are much more frequent in the earlier stages than in the later development. Constant, gradual movement toward a stable formation is characteristic of continuous succession. Contrasted with this is intermittent succession, in which the succession swings for a time in one direction, from xerophytic to mesophytic for example, and then moves in the opposite direction, often passing through the same stages. This phenomenon usually is characteristic only of the less stable stages, and is generally produced by a climatic swing, in which a series of hot or dry years is followed by one of cold or wet years, or the reverse. The same effect upon a vast scale is produced by alternate elevation and subsidence, but these operate through such great periods of time that one can not trace, but can only conjecture their effects. A normal continuous succession frequently changes its direction of movement, or its type, in transition regions or in areas where the outposts of a new flora are rapidly advancing, as in wide mesophytic valleys that run down into or traverse plains. Here the change is often sudden, and grass and desert formations are replaced by thickets and forests, resulting in abrupt succession. Species guilds are typical examples of this. More rarely, a stage foreign to the succession will be interpolated, replacing a normal stage, or slipping in between two such, though finally disappearing before the next regular formation. This may be distinguished as interpolated succession.

The apparent terminus of all stabilization is the forest, on account of the thoroughness with which it controls the habitat. A close examination of vegetation, however, will show that its stable terms are dependent in the first degree upon the character of the region in which the formation is indigenous. It is obviously impossible that successions in desert lands, in polar barrens, or upon alpine stretches should terminate in forest stages. In these, grassland must be the ultimate condition, except in those extreme habitats, alpine and polar, where mosses and lichens represent the highest type of existing vegetation. Forests are ultimate for all successions in habitats belonging to a region generally wooded, while grassland represents the terminus of prairie and plains successions as well as of many arctic-alpine ones.

CAUSES AND REACTIONS

315. The initial cause of a succession must be sought in a physical change in the habitat; its continuance depends upon the reaction which each stage of vegetation exerts upon the physical factors which constitute the habitat. A single exception to this is found in anomalous successions, where the change of formation often hinges upon the appearance of remote or foreign disseminules. The causes which initiate successions have already been considered; they may be summarized as follows: (1) weathering, (2) erosion, (3) elevation, (4) subsidence, (5) climatic changes, (6) artificial changes. The effect of succeeding stages of vegetation upon a new or denuded habitat usually finds expression in a change of the habitat with respect to a particular factor, and in a definite direction. Often, there is a primary reaction, and one or more secondary ones, which are corollaries of it. Rarely, there are two or more coordinate reactions. The general ways in which vegetation reacts upon the habitat are the following: (1) by preventing weathering, (2) by binding aeolian soils, (3) by reducing run-off and preventing erosion, (4) by filling with silt and plant remains, (5) by enriching the soil, (6) by exhausting the soil, (7) by accumulating humus, (8) by modifying atmospheric factors. The direction of the movement of a succession is the immediate result of its reaction. From the fundamental nature of vegetation, it must be expressed in terms of water-content. The reaction is often so great that the habitat undergoes a profound change in the course of the succession, changing from hydrophytic to mesophytic or xerophytic, or the reverse. This is characteristic of newly formed or exposed soils. Such successions are xerotropic, mesotropic, or hydrotropic, according to the ultimate condition of the habitat. When the reaction is less marked, the type of habitat does not change materially, and the successions are xerostatic, mesostatic, or hydrostatic, depending upon the water-content. Such conditions obtain for the most part only in denuded habitats.

316. Succession by preventing weathering. Reactions of this nature occur especially in alpine and boreal regions, in the earlier stages of lichen-moss successions. They are typical of igneous and metamorphic rocks in which disintegration regularly precedes decomposition. The influence of the vegetation is best seen in the lichen stages, where the crustose forms make a compact layer, which diminishes the effect of the atmospheric factors producing disintegration. In alpine regions especially, this protection is so perfect that the crustose lichens may almost be regarded as the last stage of a succession. There are no recorded observations which bear upon this point, but it seems certain that the pioneer rock lichens, Lecanora, Lecidea, Biatora, Buellia, and Acarospora, cover alpine rocks for decades, if not for centuries. Ultimately, however, the slow decomposition of the rock surface beneath the thallus has its effect. Tiny furrows and pockets are formed, in which water accumulates to carry on its ceaseless work, and the compact crustose covering is finally ruptured, permitting the entrance of foliose forms. The latter, like the mosses, doubtless protect rock surfaces, especially those of the softer rocks, in a slight degree against the influence of weathering, but this is more than offset by their activity in hastening decomposition, and thus preparing a field for invasion. Rocks and boulders (petria, petrodia, phellia) furnish the best examples of this reaction; cliffs (cremnia) usually have a lichen covering on their faces, while the forces which produce disintegration operate from above or below.

Fig. 66. Thicket formation (Quercus-Holodiscus-driodium), stage V of the talus succession.

317. Succession by binding aeolian soils. Dunes (thinia) are classic examples of the reaction of pioneer vegetation upon habitats of wind-borne sand. The initial formations in such places consist exclusively of sand-binders, plants with masses of fibrous roots, and usually also with strong rootstalks, long, erect leaves, and a vigorous apical growth. They are almost exclusively perennial grasses and sedges, possessing the unique property of pushing up rapidly through a covering of sand. They react by fixing the sand with their roots, thus preventing its blowing about, and also by catching the shifting particles among their culms and leaves, forming a tiny area of stabilization, in which the next generation can establish a foothold. The gradual accumulation of vegetable detritus serves also to enrich the soil, and makes possible the advent of species requiring better nourishment. Blowouts (anemia) are almost exact duplicates of dunes in so far as the steps of revegetation are concerned; while one is a hollow, and the other a hill, in both the reaction operates upon a wind-swept slope. Sand-hills (amathia) and deserts (eremia) show similar though less marked reactions, except where they exhibit typical inland dunes. Sand-binders, while usually classed as xerophytic or halophytic, are in reality dissophytes. Their roots grow more or less superficially in moist sand, and are morphologically mesophytic while their leaves bear the stamp of xerophytes. The direction of movement in successions of this kind is normally from xerophytes to mesophytes, i. e., it is mesotropic. In sand-hills and deserts, the succession operates wholly within the xerophytic (dissophytic) series. Along seacoasts, the mesophytic terminus is regularly forest, except where forests are remote, when it is grassland.

318. Succession by reducing run-off and erosion. All bare or denuded habitats that have an appreciable slope are subject to erosion by surface water. The rapidity and degree of erosion depend upon the amount of rainfall, the inclination of the slope, and the structure of the surface soil. Regions of excessive rainfall, even where the slope is slight, show great, though somewhat uniform erosion; hill and mountain are deeply eroded even when the rainfall is small. Slopes consisting of compact eugeogenous soils, notwithstanding the marked adhesion of the particles, are much eroded where the rainfall is great, on account of the excessive run-off. Porous dysgeogenous soils, on the contrary, absorb most of the rainfall; the run-off is small and erosion slight, except where the slope is great, a rare condition on account of the imperfect cohesion of the particles. In compact soils, the plants of the initial formations not merely break the impact of the raindrops, but, what is much more important, they delay the downward movement of the water, and produce numberless tiny streams. The delayed water is largely absorbed by the soil, and the reduction of the run-off prevents the formation of rills of sufficient size to cause erosion. As in dunes, such plants are usually perennial grasses, though composites are frequent; the root system is, however, more deeply seated, and a main or tap root is often present. On sand and gravel slopes, the loose texture of the soil results generally in the production of sand-binders with fibrous roots. Unlike dunes, such slopes exhibit a large number of mats and rosettes with tap-roots, which are effective in preventing the slipping or washing of the sand, and run little danger of being covered, as is the case with duneformers. In both instances, each pioneer plant serves as a center of comparative stabilization for the establishment of its own offspring, and of such invaders as find their way in. From the nature of these, slopes almost invariably pass through grassland stages before finding their termini in thickets or forests. Bad lands (tiria) furnish the most striking examples of eroded habitats. The rainfall in the bad lands of Nebraska and South Dakota is small (300 mm.); yet the steepness of the slope and the compactness of the soil render erosion so extreme that it is all but impossible for plants to obtain a foothold. Their reaction is practically negligible, and the vegetation passes the pioneer stages only in the relatively stable valleys. Mountain slopes (ancia), and ridges and hills (lophia) are readily eroded in new or denuded areas. This is especially true of hill and mountain regions which have been stripped of their forest or thicket cover by fires, lumbering, cultivation, or grazing. Where the erosion is slight, the resulting succession may show initial xerophytic stages, or it may be completely mesostatic. Excessively eroded habitats are xerostatic, as in the case of bad lands, or, more frequently, they are mesotropic, passing first through a long series of xerophytic formations. Sandbars (cheradia, syrtidia) should be considered here, though they are eroded by currents and waves, and not by run-off. They are fixed and built up by sand-binding grasses and sedges, usually of a hydrophytic nature, and pass ultimately into mesophytic forest.

319. Succession by filling with silt and plant remains. All aquatic habitats into which silt, wash, or other detritus is borne by streams, currents, floods, waves, or tides are slowly shallowed by the action of the water plants present. These not only check the movement of the water, thus greatly decreasing its carrying power, and causing the deposition of a part or all of its load, but they also retain and fix the particles deposited. In accordance with the rule, each plant becomes the center of a stabilizing area, which rises faster than the rest of the floor, producing the well-known hummocks of lagoons and swamps. All aquatics produce this reaction. It is more pronounced in submerged and amphibious forms than in floating ones, and it takes place more rapidly with greatly branched or dissected plants than with others. In pools (tiphia) and lakes (limnia), debouching streams and surface waters deposit their loads in consequence of the check exerted by the still water and the marginal vegetation, and delta-like marshes are quickly built up by filling. Springs (crenia) likewise form marshes where they gush forth in sands, the removal of which is impeded by vegetation. The flood plains and deltas of rivers show a similar reaction. The heavily laden flood waters are checked by the vegetation of meadows and marshes, and deposit most of their load. The banks of streams (ochthia) and of ditches (taphria) are often built up in the same fashion by the action of the marginal vegetation upon the current. The presence of marginal vegetation often determines the checking or deflecting of the current in such a way as to initiate meanders, while natural levees owe their origin to it, in part at least. Along low seacoasts, waves and tides hasten the deposit of river-borne detritus, causing the water to spread over the lowlands and form swamps. They often throw back also the sediment that has been deposited in the sea, the marsh vegetation acting as a filter in both cases. Successions of the kind indicated above are regularly mesotropic. Where the soil is sandy, and the filling-up process sufficiently great, or where salts or humus occur in excess, xerophytic formations result. In certain cases, these successions appear to be permanently hydrostatic, changing merely from floating or submerged to amphibious conditions, but this is probably due to the slowness of the reaction. As a rule, the accumulation of plant remains is relatively slight, and plays an unimportant part in the reaction. In peat bogs and other extensive swamps, the amount of organic matter is excessive, and plays an important role in the building up of the swamp bed.

Fig. 67. Pine forest formation (Pinus-xerohylium), stage VI of the talus succession.

320. Succession by enriching the soil. This reaction occurs to some degree in the great majority of all successions. The relatively insignificant lichens and mosses produce this result upon the most barren rocks, while the higher forms of later stages, grasses, herbs, shrubs, and trees, exhibit it in marked progression. The reaction consists chiefly in the incorporation of the decomposed remains of each generation and each stage in the soil. A very important part is played by the mechanical and chemical action of the roots in breaking up the soil particles, and in changing them into soluble substances. Mycorrhizae, bacterial nodules, and especially soil bacteria play a large part in increasing the nutrition-content of the soil, but the extent to which they are effective in succession is completely unknown. The changes in the color, texture, and food value of the soil in passing from the initial to ultimate stages of a normal succession are well known, and have led many to think them the efficient reactions of such successions. It seems almost certain, however, that this is merely a concomitant, and that, even in anomalous successions where facies replace each other without obvious reasons, the reactions are concerned more with water-content, light, and humidity than with the food-content of the soil.

321. Succession by exhausting the soil. This is a reaction not at all understood as yet in nature. A number of phenomena, such as the “fairy rings” of mushrooms and other fungi, the peripheral growth and central decay of lichens, Lecanora, Placodium, Parmelia, and of matforming grasses, such as Muhlenbergia, and the circular advance of the rootstalk plants, indicate that certain plants at least withdraw much of the available supply of some essential soil element, and are forced to move away from the exhausted area. It is probable that the constant shifting of the individuals of a formation year after year, a phenomenon to be discussed under alternation, has some connection with this. It will be impossible to establish such a relation, however, until the facts are exactly determined by the method of quadrat statistics. So far as native formations are concerned, there can not be the slightest question that prairies and forests have existed over the same area for centuries without impoverishing the soil in the least degree, a conclusion which is even more certain for the open vegetation of deserts and plains. With culture formations, the case is quite different. The exhaustion of the soil by continuous or intensive cultivation is a matter of common experience in all lands settled for a long period. Calcium, phosphorus, and nitrogen compounds especially are used up by crops, and must be supplied artificially. The reason for this difference in reaction between native and culture formations seems evident. In harvesting, not merely the grain, but the stems and leaves, and in gardening often the root also, are removed, so that the plant makes little or no return to the soil. In nature, annual plants return to the ground every year all the solid matter of roots, stems, leaves, and fruits, with the exception of the relatively small number of seeds that germinate. Perennial herbs return everything but the persistent underground parts. Shrubs and trees replace annually an immense amount of material used in leaves and fruits, and sooner or later, by the gradual decay of the individuals or by the destruction of the whole formation, they restore all that they have taken from the soil. This balance is further maintained to an important degree by the activity of the roots, which take from the deep-seated layers of the soil the crude materials necessary for the formation of leaves and fruits. Upon the fall and decay of these, their materials are incorporated with the upper layers of the formation floor, from which they may be absorbed by the undergrowth, or find their way again into the layers permeated by the tree roots. From the universal occurrence of weeds in cultivated regions, the pioneers in impoverished or exhausted fields are uniformly ruderal plants. As is well known, the seed production and ecesis of these forms are such that they take possession quickly and completely, while their demands upon the soil are of such a nature that the most sterile field can rapidly be covered by a vigorous growth of weeds. As indicated elsewhere, ruderal formations ultimately yield to the native vegetation, though in regions so completely given over to culture that native formations are lacking or remote, it is probable that successions reach their final stage within the group of ruderal plants.

Fig. 68. Spruce forest formation (Picea-Pseudotsuga-hylium), stage VII, the ultimate stage of the talus succession.

322. Succession by the accumulation of humus. This is the characteristic reaction of peat bogs and cypress swamps (oxodia), in which the accumulation of vegetable matter is enormous. The plant remains decompose slowly and incompletely under the water, giving rise to the various humic acids. These possess remarkable antiseptic qualities, and have an injurious effect upon protoplasm. They affect the absorption of water by the root-hairs, though this is also influenced by poor aeration. The same acids are found in practically all inland marshes and swamps, but the quantity of decomposing vegetation in many is not great enough to produce an efficient reaction. Formations of this type usually start as freshwater swamps. The succession is apparently hydrostatic, but no thorough study of its stages has as yet been made.

323. Succession by modifying atmospheric factors. All layered formations, forests, thickets, many meadows and wastes, etc., show reactions of this nature, and are in fact largely or exclusively determined by them. The reaction is a complex one, though it is clear that light is the most efficient of the modified factors, and that humidity, temperature, and wind, while strongly affected, play subordinate parts. In normal successions, the effect of shade, i. e., diffuse light, enters with the appearance of bushes or shrubs, and becomes more and more pronounced in the ultimate forest stages. The reaction is exerted chiefly by the facies, but the effect of this is to cause increasing diffuseness in each successively lower layer, in direct ratio with the increased branching and leaf expansion of the plants in the layer just above. In the ultimate stage of many forests, especially where the facies are reduced to one, the reaction of the primary layer is so intense as to preclude all undergrowth. Anomalous successions often owe their origin to the fact that certain trees react in such a way as to cause conditions in which they produce seedlings with increasing difficulty, and thus offer a field favorable to the ecesis of those species capable of enduring the dense shade. Successions of this kind are almost invariably mesostatic, as it is altogether exceptional that layered formations are either xerophytic or hydrophytic.

LAWS OF SUCCESSION

324. The investigation of succession has so far been neither sufficiently thorough nor systematic to permit the postulation of definite laws. Enough has been done, however, to warrant the formulation of a number of rules, which apply to the successions studied, and afford a convenient method for the critical investigation of all successions upon the basis of initial causes, and reactions. Warming has already brought together a few such rules, and an attempt is here made to reduce the phenomena of succession, including its causes and effects, to a tentative system. At present it is difficult to make a thoroughly satisfactory classification of such rules, and they are here arranged in general conformity with the procedure in succession.

I. Causation. The initial cause of a succession is the formation or appearance of a new habitat, or the efficient change of an existing one. II. Reaction. Each stage reacts upon the habitat in such a way as to produce physical conditions more or less unfavorable to its permanence, but advantageous to the invaders of the next stage. III. Proximity and mobility.

(1) The pioneers of a succession are those species nearest at hand that are the most mobile. (2) The number of migrants from any formation into a habitat varies inversely as the square of the distance. (3) The pioneer species are regularly derived from different formations, as the latter nearly always contain permobile species capable of effective ecesis. (4) The plants of the initial stages are normally algae and fungi, with minute spores, composites, and grasses, which possess permobile fruits, or ruderal plants, on account of their great seed production.

(1) All the migrants into a new, denuded, or greatly modified habitat are sorted by ecesis into three groups: (1) those that are unable to germinate or grow, and soon die; (2) those that grow normally under the conditions present; (3) those that pass through one or more of the earlier stages in a dormant state to appear at a later stage of the succession. (2) Wherever ruderal vegetation is present, it contributes a large number of the pioneer species of each succession, on account of the thorough ecesis. In other regions this part is played by subruderal native species. (3) Annuals and biennials are characteristic of the early stages of secondary successions, on account of their great seed production and ready ecesis. (4) In layered formations, heliophytes appear before sciophytes; they ultimately yield to the latter, except where they are able to maintain a position in the primary layer. (5) Excessive seed production and slight mobility lead to the imperfect ecesis of individuals in dense stands, and in consequence usually produce great instability. (6) Each pioneer produces about itself a tiny area of ecesis and stabilization for its own offspring, for the disseminules of its fellows, or of invaders. (7) Species propagating by offshoots, or producing relatively immobile disseminules in small number, usually show effective ecesis, as the offspring appear within the area of the reaction of the parent forms.

(1) Stabilization is the universal tendency of vegetation. (2) The ultimate stage of a succession is determined by the dominant vegetation of the region. Lichen formations are often ultimate in polar and niveal zones; grassland is the final vegetation for plains and alpine stretches, and for much prairie, while forest is the last stage for mesophytic midlands and lowlands, as well as for subalpine regions. (3) Grassland or forest is the usual terminus of a succession; they predominate in lands physiographically mature. (4) The limit of a succession is determined in large part by the progressive increase in occupation, which makes the entrance of invaders more and more difficult. (5) Stabilization proceeds radiately from the pioneer plants or masses. The movement of offshoots is away from the parent mass, and the chances of ecesis are greatest near its edges, in a narrow area in which the reaction is still felt, and the occupation is not exclusive.

(1) The stages, or formations, of a succession are distinguished as initial (prodophytia), intermediate (ptenophytia), and ultimate (aiphytia). (2) Initial formations are open, ultimate formations are closed. (3) The number of species is small in the initial stages; it attains a maximum in intermediate stages; and again decreases in the ultimate formation, on account of the dominance of a few species. (4) The normal sequence of vegetation forms in succession is: (1) algae, fungi, mosses; (2) annuals and biennials; (3) perennial herbs; (4) bushes and shrubs; (5) trees. (5) The number of species and of individuals in each stage increases constantly up to a maximum, after which it gradually decreases before the forms of the next stage. The interval between two maxima is occupied by a mixed formation. (6) A secondary succession does not begin with the initial stage of the primary one which it replaces, but usually at a much later stage. (7) At present, successions are generally mesotropic, grassland and forest being the ultimate stages, though many are xerostatic or hydrostatic. If erosion continue until the sea level is reached, the ultimate vegetation of the globe will be hydrophytic. Should the heat of the sun decrease greatly before this time, the last vegetation will be xerophytic, i. e., crymophytic. (8) The operation of succession was essentially the same during the geological past as it is to-day. From the nature of their vegetation forms, the record deals largely with the ultimate stages of such successions.

CLASSIFICATION AND NOMENCLATURE

325. Basis. New or denuded habitats arise the world over by the operation of the same or similar causes, and they are revegetated in consequence of the same reactions. Similar habitats produce similar successions. The vegetation forms and their sequence are usually identical, and the genera are frequently the same, or corresponding in regions not entirely unrelated. The species are derived from the adjacent vegetation, and, except in alpine and coast regions, are normally different. The primary groups of successions are determined by essential identity of habitat or cause, e. g., aeolian successions, erosion successions, burn successions, etc. When they have been more generally investigated, it will be possible to distinguish subordinate groups of successions, in which the degree of relationship is indicated by the similarity of vegetation forms, the number of common genera, etc. For example, burn successions in the Ural and in the Rocky mountains show almost complete similarity in the matter of vegetation forms and their sequence, and have the majority of their genera in common. A natural classification of successions will divide them first of all into normal and anomalous. The former fall into two classes, primary and secondary, and these are subdivided into a number of groups, based upon the cause which initiates the succession.

Fig. 69. Aspen forest formation (Populus-hylium), the typical stage of burn successions in the Rocky mountains; it is sometimes an anomalous stage in primary successions, interpolated in place of the thicket formation.

326. Nomenclature. The need of short distinctive names of international value for plant formations is obvious; it has become imperative that successions also should be distinguished critically and designated clearly. From the very nature of the case, it is impossible to designate each formation or succession by a single Greek or Latin term, as habitats of the same character will show in different parts of the world a vegetation taxonomically very different. It may some day be possible to use a binomial or trinomial for this purpose, somewhat after the fashion of taxonomy, in which the habitat name will represent the generic idea as applied to formations, and a term drawn from the floristic impress the specific idea. Such an attempt would be futile or valueless at the present time; it could not possibly meet with success until there is more uniformity in the concept of the formation, and until there has been much accurate and thorough investigation of actual formations, a task as yet barely begun. At present, it seems most feasible as well as scientific to designate all formations occupying similar habitats by a name drawn from the character of the latter, such as a meadow formation, poium, a forest formation, hylium, a desert formation, eremium, etc. A particular formation is best designated by using the generic name of one or two of its most important species in conjunction with its habitat term, as Spartina-Elymus-poium, Picea-Pinus-hylium, Cereus-Yucca-eremium, etc. Apparently a somewhat similar nomenclature is adapted to successions. The cause which produces a new habitat may well furnish the basis for the name of the general groups of successions, as pyrium (literally, a place or a habitat burned over), a burn succession, tribium, an erosion succession, etc. A burn succession consists of a sequence of certain formations in one part of the world, and of a series of quite different ones, floristically, in another. A particular burn succession should be designated by using the names of a characteristic facies of the initial and ultimate stages in connection with the general term, e. g., Bryum-Picea-pyrium, etc. A trinomial constructed in this way represents the desirable mean between definition and brevity. Greater definiteness is possible only at the expense of brevity, while to shorten the name would entirely destroy its precision. The following classification of successions is proposed, based upon the plan outlined above. The termination -ium (εῖον) has been used throughout in the construction of names for successions, largely for reasons of euphony. If it should become desirable to distinguish the names of formations and successions by the termination, the locative suffix -on (-ών) should be used for the latter. The terms given below would then be hypson, rhyson, hedon, sphyron, prochoson, pnoon, pagon, tribon, clyson, repon, olisthon, xerasion, theron, broton, pyron, ecballon, camnon, ocheton, ardon.

I. Normal successions: cyriodochae (κύριος, regular, δοχή, ἡ, succession)

a. Primary successions: protodochae (πρῶτος, first, primary) 1. By elevation: hypsium (ὔψος, το, height, elevation, -εῖον, place) 2. By volcanic action: rhysium (ῥυσίς ἡ, flowing, especially of fire) 3. In residuary soils: hedium (ἔδος, τό, a sitting base) 4. In colluvial soils: sphyrium (σφύρον, τό, ankle, talus) 5. In alluvial soils: prochosium (πρόχωσις, ἡ, a deposition of mud) 6. In aeolian soils: pnoium (πνοή, ἡ, blowing, blast) 7. In glacial soils: pagium (πάγος, ὁ, that which becomes solid, i. e., a glacier) b. Secondary successions: hepodochae (ἕπω, to follow) 8. In eroded soils: tribium (τρίβω, wear or rub away) 9. In flooded soils: clysium (κλύσις, ὁ, a drenching, flooding) 10. By subsidence: repium (ῥέπω, incline downwards, sink) 11. In landslips: olisthium (ὄλισθος, ὁ, slip) 12. In drained and dried out soils: xerasium (ξηρασία, ἡ, drought) 13. By animal agencies: therium (θήρ, ὁ, wild animal) 14. By human agency: brotium (βροτός, ὁ, a mortal) a. Burns: pyrium (πῦρ, τό, fire) b. Lumbering: ecballium (ἐκβάλλω, cut down forests) c. Cultivation: camnium (κάμνω, cultivate) d. Drainage: ochetium (ὀχετός, ὁ, drain) e. Irrigation: ardium (ἄρδω, irrigate)

327. Illustrations. The following series will illustrate the application of this system of nomenclature to particular successions, and their stages, or formations.

INVESTIGATION OF SUCCESSION

328. General rules. The study of succession must proceed along two fundamental lines of inquiry: it is necessary to investigate quantitatively the physical factors of the initial stages and the reactions produced by the subsequent stages. This should be done by automatic instruments for humidity, light, temperature, and wind, in order that a continuous record may be obtained. Water-content is taken daily or even less frequently, while soil properties, and physiographic factors, altitude, slope, surface, and exposure are determined once for all. It is equally needful to determine the development and structure of each stage with particular reference to the adjacent formations, to the stage that has just preceded, and the one that is to follow. For this, the use of the permanent quadrat is imperative, as the sequence and structure of the stages can be understood only by a minute study of the shifting and rearrangement of the individuals. Permanent migration circles are indispensable for tracing movement away from the pioneer areas by which each stage reaches its maximum. Denuded quadrats are a material aid in that they furnish important evidence with respect to migration and ecesis, By means of them, it is possible to determine the probable development of stages which reach back a decade or more into the past. In the examination of successions, since cause and effect are so intimately connected in each reaction, it is especially important that general and superficial observations upon structure and sequence be replaced by precise records, and that vague conjectures as to causes and reactions be supplanted by the accurate determination of the physical factors which underlie them.

Fig. 70. Alternating gravel slides on Mounts Cameron and Palsgrove, from the comparison of which the initial development of the talus succession has been reconstructed.

329. Method of alternating stages. The period of time through which a primary succession operates is usually too great to make a complete study possible within a single lifetime. Secondary successions run their course much more quickly, and a decade will sometimes suffice for stabilization, though even here the period is normally longer. The longest and most complex succession, however, may be accurately studied in a region, where several examples of the same succession occur in different stages of development. In the same region, the physical factors of one example of a particular succession are essentially identical with those of another example in the same stage. If one is in an initial stage, and the other in an intermediate condition, the development of the former makes it possible to reestablish more or less completely the life history of the latter. The same connection may be made between intermediate and ultimate stages, and it is thus possible to determine with considerable accuracy and within a few years the sequence of stages in a succession that requires a century or more for its complete development. In the Rocky mountains, gravel slides (talus slopes) are remarkably frequent. They occur in all stages of development, and the alternating slides of different ages furnish an almost perfect record of this succession. This method lacks the absolute finality which can be obtained by following a succession in one spot from its inception to final stabilization, but it is alone feasible for long successions, i. e., those extending over a score or more of years. When it comes to be universally recognized as a plain duty for each investigator to leave an exact and complete record in quadrat maps and quadrat photographs of the stages studied by him, it will be a simple task for the botanists of one generation to finish the investigations of succession begun by their predecessors.

330. The relict method of studying succession is next in importance to the method of alternating areas. The two in fact are supplementary, and should be used together whenever relicts are present. This method is based upon the law of successive maxima, viz., the number of species and of individuals in each stage constantly increases up to a certain maximum, after which it gradually decreases before the forms of the next stage. In accordance with this, secondary species usually disappear first, principal species next, and facies last of all. There are notable exceptions to this, however, and the safest plan is to use the relict method only when principal species or facies are left as evidence. An additional reason for this is that secondary species are more likely to be common to two or more formations. In the majority of cases, the relict is not modified, and is readily recognized as belonging properly to a previous stage. This is true of herbs in all the stages of grassland, and in the initial ones of forest succession. The herbs and shrubs of earlier stages, which persist in the final forest stages, are necessarily modified, often in such a degree as to become distinct ecads, or species. The facies of the stages which precede the ultimate forest are rarely modified. The application of the relict method, together with the modification just described, is nicely illustrated by the balsam-spruce formation at Minnehaha. Of the initial gravel slide stage, the relicts are Vagnera stellata and Galium boreale, the one modified into Vagnera leptopetala, and the other into G. boreale hylocolum. The thicket stage is represented by Holodiscus dumosa, greatly changed in form and branching, and in the shape and structure of the leaf. The most striking relict of the aspen formation is the facies itself, Populus tremuloides. The tall slender trunks of dead aspens are found in practically every balsam-spruce forest. In many places, living trees are still found, with small, straggling crowns, which are vainly trying to outgrow the surrounding conifers. Of the aspen undergrowth, Rosa sayii, Helianthella parryi, Frasera speciosa, Zygadenus elegans, Castilleia confusa, Gentiana acuta, and Solidago orophila remain more or less modified by the diffuse light. It is still a question whether the aspen stage passes directly into the balsam-spruce forest, or whether a pine forest intervenes. The presence of both Pinus ponderosa and P. flexilis, which are scattered more or less uniformly through the formation, furnishes strong evidence for the latter view.

Fig. 71. Relict spruces and aspens, showing the character of the succession immediately preceding the burn succession now developing.

The lifetime of forest and thicket stages of successions is ascertained by counting the annual rings of the stumps of facies. This is a perfectly feasible method for many woodland formations where stumps already abound or where a fire has occurred, and it is but rarely necessary to cut down trees for this purpose. When trees or shrubs are present as relicts, the same method is used to determine the length of time taken by the development of the corresponding stages.