RS ee te eee ts 5} oy CONTRIBUTIONS FROM THE GRAY HERBARIUM OF HARVARD UNIVERSITY Edited by Reed C. Rollins and Robert C. Foster NO. CLXXVIII SOG FLATS AND PHYSIOGRAPHIC PROCESSES IN THE UPPER KUSKOKWIM RIVER REGION, ALASKA With suggestions for the origin of certain mucks, an examination of the nature of the local plant associations, and a discussion of the implications as to periglacial climates. Pie st SL Pere Fate tbs abe nd aPC ar faetentae® . rat 2. vel toy Met um Sopa Be ie Coe the po eee DS. ea By WittiaM H. Drury, Jr. PUBLISHED BY THE GRAY HERBARIUM OF HARVARD UNIVERSITY CAMBRIDGE, MASS., U.S.A. ISSUED AUGUST 16th Ae sae eee ee Ad CONTENTS INTRODUCTION DESCRIPTION OF THE AREA Geography Climate Vegetation The Bogs Geological en ate a the os Vegetational description of the bogs . PHYSIOGRAPHIC PROCESSES The floodplain surface : The divisions of the floodplain The Bogs . ‘ Formation of ioe Distribution of bog types . Transformation of the bog basin : Geobotanical processes and sedimentary a Explanation of field observations . ‘ oo ere ee soligenous mires. - Base level and graded ie in the c ae. _ VEGETATION Vegetation of the floodplain . Vegetation of the bogs . Successional sequence . The plant associations - vegetation of silt esos we bine Beach Vegetation vegetation of peat environments: Peat fog Verein simple stands . ‘ patterned surfaces eee vegetation under the combined influence of peat and silt Comparison of bog vegetation with Scandinavian studies General comparisons Specific correlations of vegetation types Meaning of the presence of sphagnum . Strangmoor Regularly discritacced slaad DISCUSSION Geological implications SUMMARY . THE FLOR GLOSSARY effect on Man’s use of be pti effect of stratigraphic sequences . production of muck . elimination of forests by soil processes . ical implications : areas of persistence ietay: the ice ede as of the Pleistocene... speculations on the nature af ceciotarsal vegetation in the eastern United States relation of these phenomena to the concept of shia in vegetation A OF THE BOGS . LITERATURE CITED . INDEX v, d. rs ts > ie w he 3 E bs a =| & a” ae c ab I v + pw “iy an aA S S be a " 348 pa 2S 5 S35 a mn y. &e oS eb Bu of, a ee Ne =z Ao bd G Re o 5S w.s eis Boe mis Kou ea os wo . v Ee 0 KWIM RIVER REGION. Place names mentioned in the text are include rock hills using a study outlined in black f (Forest distribution taken from Hopkins and and the bed aska shows the location of the area o ah 1951). | 90S, SEARS Sih RES a? Ilust Tt? by bos “Dowaslo pe moven rent of water Is resp i) “i a o m is] =] mo on i tai) 6 =) mr er fi = ae 4 cé appear the winter and snow cove Army Air Force makes the bogs INTRODUCTION The extensive treeless bogs which cover the broad alluvial lowlands of the interior and southern coastal regions of Alaska are of a far northern type which occurs all across Eurasia and North America’. These bogs are characterized by their great size, by their ron areas flooded by standing water, by their anastomosing patt in which they surround islands and peninsulas of rane as well as by their surface patterns of sinuous bog ridges and regularly spaced clumps of trees. In addition, in Alaska the bogs are actively undermining and engulfing the forests of the lowlands by processes active at their banks. An isolated bog area in the Upper Kuskokwim River Region occupies an area of 250 square miles (Figures 1 and 2), and the bog areas along the Tanana River south of Fairbanks occupy several thousand square miles. Continuous quaking bogs have been seen in Alaska as long as 15 miles and as wide as 5 miles. The bogs and the patchy forest areas in them are locally called “Flats”. These flats are the Aapasuo of Finland. Their surface patterns of nets and festoons of bog eg put them in the category of the Strangmoor of the Ge As this indicates, similar areas have been ake in Europe and a number of separate terminologies have been developed. An attempt is made here to compare and equate these terms and for this reason the reader is asked to show patience with numerous cross references in some parts of this paper. Because the text is in English, equivalents have been supplied, such as bog ridge and bog hollow. These are not intended to be new terms and are entirely synonymous with German, Swedish or Finnish words, as pointed out in the Glossary. Early studies in Europe, such as that of Cajander (1913), described the distribution and vegetation of the three major types of bog in Scandinavia. Recent studies have concentrated almost entirely upon the raised bogs or mosses (Hochmoors) of Central Scandinavia and the shores of the Baltic Sea, and have investigated 1) I have seen them in the Yukon and Tanana low ecg — one o_ — -_ Bristol Bay areas of Alaska, in Northeast Manitoba, and i (ge Se rt 8. inlat ne (Cajander, 1913, Taf. 9, 10; Auer, 1920, Taf. 1-7: Troll 1944, p. gr ‘Sjars, 1948, p. 189 etc.) and in northern Russia and western Sileeria (Troll, 1944, p. 640; Katz, 1926, p. 184 cte.). 6 WILLIAM H. DRURY, JR. the physiology of individual plants or plant groupings. They have been concerned with the growth and formation of peat, and geological studies have been carried out simultaneously. In America, Shaler in 1890 described bogs colonizing a pond and invading the wet soil of a wet woodland. Since then many authors have described bogs in temperate North America as phenomena of special conditions of restricted distribution, such as undrained topographic depressions. Furthermore, these studies are founded on acceptance of successional sequences by which open water is colonized, filled with plant remains and transformed into a forested lowland. The concept of unidirectional succession and and progressive development to a stable equilibrium as a general- ization of vegetation process is inherited from the work of F. E. Clements. Earlier workers did not visualize the succession as unidirectional nor the mesophytic equilibrium as final or climactic. Most bog studies, especially those of the New World, have considered either plants or geology while slighting or not under- standing the limitations of the other processes. A few men, such as V. Auer, have studied the intimate integration of physiographic processes with the growth of groups of plants. There are processes in the river-deposited silts and in the bogs studied in this paper tending to build and compact a frozen soil of mixed silt and peat called Muck in Alaska. On this soil a muskeg forest grows. At the same time processes lead to the destruction of the lowland forests and invasion by bogs. Creation and destruction of forests and bogs is repeated and phases of bog and forest primacy move across the lowland. Bogs of regional extent, which engulf the forests and are seen to replace them in a successional sequence, cannot be forced into the frame of reference set up in temperate regions. The correlation of bog types with ages of the floodplain and the particular phenomenon of the advance of bog margins by thawing and undermining the forest show the close relation of these bogs to geological processes. The patterns of vegetation, of islands, and of networks of ridges show that physical forces are of fundamental importance in the bog surface vegetation. These observations lead 2) Muck is defined by geologists as a mixture of silt and finely divided organic material. To the Alaskan placer-miners muck is any fine-grained material over- lying the gold-bearing gravels. In Interior Alaska the mucks are usually frozen, PERI See Soe ETE ee ETE et BEd Eg UGE Mae ge peeps ECan Gs: Ae pt rae enue oe leering Rar eee namert ea Ud oe OA Fa i alah igi tai BOG FLATS AND PHYSIOGRAPHIC PROCESSES IN ALASKA 7 to the conclusion that these bogs must be studied as a combined phenomenon of geological and vegetational processes acting in concert and not as an isolated and peculiar vegetational feature independent of its surroundings while proceeding to its own predetermined goal. Geological processes are evidently the key to the description of the bogs, and beyond that to the understanding of their formation and development. The contribution of this Paper is maintaining the integration of these processes. In the course of these studies I have reached a number of conclusions which may be of some general application. It has become evident that geological and botanical processes involved in the formation of these bogs alter the sequence of sediments and modify alluvial deposits of different ages and histories, obliter- ating their differences. An understanding of these processes will contribute to the knowledge of the history of the Pleistocene and Recent epochs by application to stratigraphy and to pollen analysis, and by the evidence these contribute to knowledge of the climates of the past. Furthermore, the change of vegetation in which regional bogs supplant forested lowland areas by botanical as well as geological processes is a serious argument against the application of theories of regional or climatic climax vegetation to the vegetation of these areas. The study of this change is a contribution to the understanding of what processes are actually involved in vegetational successions. The processes studied here contribute further evidence sup- Porting the importance of soil instability and the effects of frost action on the patterns of vegetation and the containing action of the carpet of plants. Whereas nearly all studies so far published (such as: Crampton, 1912; Polunin, 1934—35; Hopkins and Sigafoos, 1951; Raup, 1951) have been concerned with physio- graphic effects of frost action on uplands, these investigations are a study of the processes active in alluvial lowlands. Field work in the bogs totalling just less than a month was completed during the summers of 1949 and 1950 coinciglent with other duties in the area. The opportunity and laboratory and herbarium facilities for three years study of field data were supplied by the Society of Fellows of Harvard University. Debt for ideas obtained during discussions in the progress of the field work is acknowledged to William S. Benninghoff, Jr., Robert F. Black, Jr., George $. Corchary, Arthur T. Fernald 8 WILLIAM H. DRURY, JR. and William C. Steere. Debt for critical reviews is acknowledged to them and to John Goodlett, David M. Hopkins, Ernest H. Muller, Troy Péwé, Hugh M. Raup, Robert S. Sigafoos and John R. Williams. DESCRIPTION OF THE AREA GEOGRAPHY The area studied (Figure 1) is in the Upper Kuskokwim River Region which lies between 62° and 63° 30’ north latitude and between 154° and 156° west longitude. This is in the region of discontinuous permafrost (Black, 1950) and within the limits of _ the boreal needle-leaved forest. The Kuskokwim River valley forms a lowland 50 miles wide trending northeast to southwest across the map area. The Alaska Range, rugged “young” moun- tains (in the Davis 1899 and 1909 sense), rising to peaks of 10,000 feet lies across the southeast corner. The “old” hills of the Kuskokwim Mountains enclose the basin on the west, northwest and north. Both acidic and basic rocks, both crystalline and sedimentary are included in the structure of the two ranges. The oldest rocks are Paleozoic in age and occur in three areas: (1) the northern flank of the Alaska Range in the Farewell district; (2) the Nixon Fork district; and (3) along the Kuskokwim River from Medfra to McGrath and south. Limestone is the predominant rock type in the first two districts, and shale along the Kuskokwim. Datings are discussed by Spurr (1898), Brooks (1911), Brown (1926), and Smith (1939). Rocks of Cretaceous age underlie the upland areas to the west. These are mostly interbedded graywackes and shales which have been folded in the area along the Kuskokwim River. Tertiary intrusives, stocks and smaller igneous bodies, are scattered throughout the area (Mertie and Harrington, 1924; Brown, 1926; Mertie, 1936). Glaciers flowing from the snowfields of the Alaska Range have in the past extended northwestward as much as 20 miles beyond the front of the Alaska Range. Outwash deposits from the rivers in their valleys form a pieduaieit surface sloping gently to the northwest which is crossed by the South Fork (Echeatnu BOG FLATS AND PHYSIOGRAPHIC PROCESSES IN ALASKA 9 in Hulrén, 1941), the Middle Fork, Salmon River, Windy Fork, and the Big River (West Fork in Péwé, 1953, pl. 1). These tributaries change from braided to meandering streams as they cross from the piedmont to the valley lowlands and there join with the North Fork to form the major river, the Kuskokwim. The Kuskokwim flows to the southwest between high alluviai terraces 60—100 feet above the river level, then plunges into and across the Kuskokwim Mountains from which it emerges to flow north of the Killbuck Mountains, and into the Bering Sea south of the Yukon River. Deep fine-grained alluvium, modified on the valley sides by depositional-slope-deposits, underlies the floodplain of the Kuskokwim, the reéntrant valleys in the Kuskokwim Mountains, such as that of the Takotna and Nixon Fork Rivers, and the lower one-third to one-haJf of the piedmont slope from the Alaska Range. The alluvium ese ae minor streams onto the lower parts of hill-slopes and terr The town of McGrath, a atde 550 feet, (Figure 1) is at the junction of the Takotna River with the Kuskokwim and is a transhipment site for supplies to the gold mines in the Kuskokwim Mountains. The floodplain of the Kuskokwim is about 4 miles wide there, and the surface of the floodplain at the town is about 8 feet above the average summer river level. The main river averages 1,000 feet across and the zone of active meanders is about 2 miles wide. Most of the active meander belt shows evidence of regular flooding during spring “break-up” of the river ice. Perennially frozen ground (Permafrost) is widespread in fine-grained sediments of the depositional slopes and alluvial lowlands. Its distribution coincides with that of Black Spruce forests. CLIMATE The lowlands of the Upper Kuskokwim River Region (climatic data according to Hopp, 1951) have cold and dry winters and and warm and wet summers. Maximum precipitation is in July, August, and September. The daily summer maximum temperatures average in the low seventies and the frost-free period is about 120 days. Annual cloudiness averages slightly over seven-tenths. According to Képpen’s (1931) climatic classification, three times as much rain falls as is needed to make the McGrath climate 10 WILLIAM H, DRURY, JR. (mean annual temperature 25.5° F.) “moist”. The cold dry winters relate the climate to continental types, but the vegetation shows a definite maritime character (forests and bogs). Winter dormancy must allow the plants of this area to resist extremes of winter temperatures, so that it is the con- ditions of the growing season which control the type of vegetation present. The growing season in bogs probably only approximates the frost-free period, given as 120 days, because the bogs have been found to be frozen on June 9—12, 1950, which is more than three weeks after the last killing frost reported at McGrath (May 18). Frost-nipped plants were seen in bogs a month before the first killing frost reported at McGrath in the middle of September. VEGETATION About 75% of the Upper Kuskokwim River Region is forested. In the lowlands and on the slopes of the hills, well-drained and sunny sites (15% of the area) are forested with a White Spruce mixed forest *, but by far the greatest proportion, including most of the lowlands, the depositional slopes and all the poorly drained gentle slopes (60 of the area), is forested with Black Spruce (Picea mariana). Ten percent of the area is occupied by large bog areas which spot the older parts of the valley bottoms where surface and subsurface water has stagnated. Bog areas are indicated with standard swamp symbol in Figure 1. The hilltops of the Kuskokwim Mountains, the Alaska Range and the upper third of the piedmont slope northwest of the Alaska Range (15% of the area) are beyond the limit of closed forest. On the well-drained sites on the older parts of the floodplain and on warm hillslopes, Birch is prominent in the mixed forest. but Birch does not extend above 1,000 feet. Thus, on the depositing slopes of meandering streams and along the braided streams in the Alaska Range, the forest is largely of White Spruce and Balsam Poplar. 3) White Spruce-mixed forest refers to forests containing deciduous trees as well as White Spruce (Picea glaxes). On uplan i are aspen (Populus iremuloides) or white Birch (Betula papyrifera, usually va umilis). On floodplain surfaces the deciduous trees are Balsam Popla (Populus balsamifera) and White Birch (Betula papyrifera usually variety kenaica). BOG FLATS AND PHYSIOGRAPHIC PROCESSES IN ALASKA 11 The dry ridges of the extensive areas of sand dunes or erosional remnants on the high terraces along the Kuskokwim River are covered with a lichen-floored open forest of Black Spruce and Aspen with a Birch (especially Betula papyrifera var. humilis) and White Spruce admixture *. The low places on these terraces are Black Spruce muskegs through which flow sphagnum-clogged streams. Aspen groves line small pools. The low hill slopes and valley bottoms are forested with Black Spruce and bogs are scattered through this forest in the lowlands. These vegetation types are discussed in detail in the following chapters. Tree-line is at about 1,000 feet on north slopes and at 1800—2100 on south slopes. At tree-line on the piedmont slope is a tundra-forest (Berg, 1950), of Black Spruce, White Spruce or Larch with a dense low-brush growth of heaths and dwart birch and a thick moss mat. Immediately above the limit of trees the ground is covered with a carpet of mosses, lichens and low brush identical with the understory of the wet spruce forests. This mesophytic tundra vegetation is modified according to exposure and drainage, and special expressions of it occupy large areas. For example, on the slopes of the gravel outwash there is a development of cotton-grass tussocks, tussock-tundra (Hopkins and Sigafoos, 1950), (known locally as “nigger-heads”) covering as large an area as that occupied by the mesophytic tundra. Above 2,500—3,000 feet and on well-drained ridges the vegetation consists of a more or less interrupted carpet of prostrate shrubs and mat plants such as willows, heaths and Mountain Avens (Dryas), with some mosses and many lichens’. —, 4) It seems to correspond to the Siberian suigu forest as described by a native of that part of the U.S.S.R., and resembles Raup’s (1946) oan tke spruce forest Northeast of La thabasca. A NE brush vegetation immediately beyond treeline is characteristic of the more mesophytic sites of the treeless areas all across the nort It is the low-lying tundra as Middendorf (1864) eee it in Siberia, and the wet tundra - Sigafoos (1951) described it in Seward Penin — — (1948) _ documen the occurrence of thi this type of vegetation ac ctic. t resembles art “Sphagnetum” and “Blanket Bog” of Tauskey’s (1939) British Tus sock tundra i a vegetation type which predominates in the lowlying tundra regions of Alaska: Seward Peninsula (Porsild. 1941: Sigafoos. 1951) and north of the Brooks Range ee ng to botanists Lloyd Spetzman, Quentin Jones and A. R. Hodgdon). It resembles the“Eriophoretum” of Tansley’s (1939) British bogs. 12 WILLIAM H. DRURY, JR. THE Bocs The older parts of the floodplain, near the valley sides, are nearly entirely covered with vast bogs. The bogs occupy about three-quarters of the mature floodplain area, beyond the zone of active meanders, and exist as a contradiction to classical ideas of plant succession on a floodplain. The classical concept of plant succession on a floodplain holds true in temperate regions and has changed little since Theophrastus first described it in the Aegean region in his Inquiry into Plants. The bogs are described in detail here before going on to interpretation of process, formation and development. GEOLOGICAL DESCRIPTION OF THE BOGs ° From a geological point of view, there are two essential parts of the lowland “flats”: 1) the peat bogs with their included ponds and lakes and 2) the surface of the lowland which surrounds them or occurs in them as islands. In the Upper Kuskokwim River Regions this “solid land” is forested, chiefly with Black Spruce, but E. Muller, Geologist, tells me that in the Bristol Bay Region of Alaska similar patterns are developed without forests. 1) The peat bogs. The bogs are a water-sodden mass of the dead accumulations of past years plant growth, chiefly sphagnum. A more or less closed vegetational cover of mosses, sedges, and low brush grows on the surface. Bogs may be more or less circular or elongate, from a few yards to many hundreds of yards long. If bog-vegetation has colonized a drainage system or taken over a low slope, the bog may extend for several miles. In general the bogs are found on level or gently sloping surfaces underlain by silt, fine sand and organic matter, or other materials or conditions which — subsurface drainage. ins cs carpet of mat plants mosses and lichens corresponds to Hochtundra or Flechtentundra ee Middendorf (1864). This is to dry tundra of Sigafoos (1951), arctic-alpine zone of Tansley (1939) and montane vegetation of Pearsail (1950). It is what Europeans call the Arctic vegetation above the alpine meadow = Geilogical is here used to — ig the botanical or vegetational escription . The geological descrip is of the topographic features or the physiography. In the rey oth agteondy Fi is impossible to separate the ig Sipe a process from the biological but in description it is ¢ to do so. Ficure 3. MUCK EXPOSED IN A RIVER Cl O MILES all across the picture half way up the bank On the left. black a dry, SAS SGRATH. A lay ih ower | t muck extends ; limit is sha bank. } I of € Dank, Dut on th tan silt. The terrace-top Lo riVEL LOY. ME TETUDE PHOTOG RAPH OF THE FLOODPL Ve OF THE KUSKOKWIN etatic on ed lepositi r slo fe the by t i ~ if Hie s} ie rs beton« rig , ht cor “ris } : Te} } ie ee Po ; t plar. J sir iS 21 hese constity WOOL FIGURE 6 sedge AERIAL those areas is of PHOTOGRAPH FROM 10,000 FEET LOOKING EAST UP THE IIT ace upie: s the meadow $ The sand est dunes area, that of the mottled, irregular pattern. on the eal terrace covered with Ranteai: pe bee eta RIVER FROM NEAR MCGRATH. Phase is restricted. where sloughs are parallel to the river and pe deciduous _ gt The de Sop a trees in fall i Se are the arger extent where sloughs are at an angle to the river Ter S t of W hite the darkest tone Spruce gre Sw ale and ox- si ison persist. The nad light tone is chief} of muskeg and Saree 4 groves appear on the right. (Stebcarenk BOG FLATS AND PHYSIOGRAPHIC PROCESSES IN ALASKA 13 The depth to mineral soil is variable where known. In cases in which bogs now occupy former channels of the Kuskokwim River the depth can be assumed to be 30 to 40 feet or more. Beds of peat 30 feet high are at present exposed in the river banks, and the exposures in the high terraces, such as that 70 miles up the Kuskokwim River from McGrath, show mucks 35 feet thick‘ (Figure 3). Auger holes 6 feet deep into the bogs have encountered no perennially frozen material and penetrated only peat or peat with a silt content which increases conspicuously with increasing depth. Deposits which underlie areas of peat bogs exposed by cutting of the river are of the following types: 1) gray, river-deposited, bedded silts alternating with beds of peat; 2) gray, river-deposited silts overlain by deposits of peat of various thicknesses; 3) a mixture of silt and finely divided organic matter, often frozen, which when it thaws has the characteristic fetid odor of muck. Involutions have been found under bogs and forests in the bedded deposits near bed-rock hills, but most of the gray, organic silts show no bedding and no signs of involutions. 2) The non- -bog surface in most of the Upper Kuskokwim River Region is readily separated from the bogs because it is covered with a forest of Black Spruce. Below the living mosses which carpet the floor of this forest is a deposit of peat, usually dry, from 1—25 feet, but usually 2—3 feet thick. Below this are the sediments of the alluvial lowlands which vary according to ae history of their deposition as described in the preceding agraph. The top of the perennially frozen ground (permafrost ae generally lies 18—24 inches under the carpet of mosses. mall ice lenses are widespread in the perennially frozen ground. The characteristics of the contact between the non-bog surfaces of the lowland and the bogs vary according to whether the bog is advancing or is being invaded by peat-forming vegetation. VEGETATIONAL DESCRIPTION OF THE BOGS aS The vegetation of the flats areas is also divisible into oT This muck exposure es sharply separated from the eis tg bedded silt Of about ut 40 s mc 7 feet then grades into a silt bank. There are no traces of n the or a a inside it or along — contact with the silt halon it heicotion: soil mo 14 WILLIAM H. DRURY, JR. quaking bogs and raised areas. The open surfaces of the quaking bogs match in detail the available descriptions of bogs in central Russia (Katz, 1926) and northern Finland (Cajander, 1913; Auer, 1920). The bog surfaces are covered by a variety of vegetational associations. Each association is a unit and is repeated as a unit in the mosaic of vegetation, but so various are the possible combinations, that at first there seems to be no order either to the vegetation or to the shifting patterns. The associations are 1) sedge meadow, 2) areas of shallow water with mud bottoms and scattered emergent aquatics, 3) areas of low or 4) tall heaths and dwarf birches with a dense sphagnum carpet, 5) areas where the water flows across the entire surface and mosses are nearly lacking, 6) small open ponds, 7) areas of net-like patterns of bog ridges and bog hollows. The carpet of vegetation in those bogs which have formed by colonization of a body of deep, open water such as an ox-bow lake is much less compacted than that of shallow bogs. The silt-bottomed, shelving beaches of ox-bow or bar lakes * have a characteristic flora which may persist when the area is invaded by sphagnum. This history of formation of such areas can be read in their vegetational patterns and flora. 2. The non-bog surfaces, which include the elevated islands in the bogs and the main lowland surface, like the isolated patches of forest colonizing on the bog surface, have a heavy moss carpet of Sphagnum and woodland mosses (H ylocomium, Pleurozium and Dicranum), often with an extensive growth of lichens (Cladonia, Cetraria, Peltigera, and Nephroma) underlain by varying thicknesses of peat. There is a dense growth of low heath and dwarf birch shrubs (Betula nana var. sibirica) about 2 feet high, scattered tall willows (Salix Bebbiana, S. pulchra, S. Scouleriana) and alders (Alnus crispa and A. tenuifolia) 8—15 feet high, a few White Birch trees (especially Betula papyrifera var. kenaica) and variable numbers of Larch (Larix laricina var. alaskana) in the forest of Black 8) A bar lake is a body of water enclosed by the deposition of a crescentic. bar or depositional scroll on the inside of a meander bend of a river (Melton, — 1936). BOG FLATS AND PHYSIOGRAPHIC PROCESSES IN ALASKA 15 Spruce. This is by far the most widespread forest of the lowlands and gentle slopes. It covers about 6700 square miles or 75% of the total area of the upper Kuskokwim River Region. Larch is most conspicuous among islands and rows of trees on the surface of the bogs. On the islands formed by slump from the terrace escarpments, there may be dense groves of White Spruce and 2 rat o z WHITE BLACK 2 2 POPLAR | WHITE SPRUCE. BLACK HILL ei ag SPRUC! - SPRUCE SLOPE 8) «9 WHITE MUSKEG BOG FLATS OR S $3 BIRCH TERRACE =| HYPNACEAE SPHAGNUM SPHAGNUM me Px 9 4 F} LS ee “ € t G ¢ oe 4 ¥ s * ep ree 3 é $ t 7 // at, ) Lite rene spies ees [Yl FARTS bs wi A ges OER PRE ee, ER TRytchivAainen whl’ KL SEs WS SS oe SS aed é i MA ESS SS SAM. sage ioe oa uA ELLe < erm LARRR SS RR RS SE Se mans a RS ROS RQ =e ees SS SO SSS SR ARK a = SN Sh QS ERO SSS re SS so Sh URE 5. DEALIZED CROSS-SECTION OF THE FLOODPLAIN. The upper diagram eb eeene, a gs have forme wing, the lower, areas where Ogs have fora — W ale pe w aprd aes and swamping. Vertical hatching is living white tions; diagonally hatched mulati areas (top-right to ; hauieas eft) are reeset cre acoaae hatched areas (top-left to bottom-right) are permafrost 9) Where the river has cut against the high terrace leaving a bank 60- a feet hi h, then —— aw - —_ se gen parts of the terrace escarpm ei _— ty 90 the for channels an — in the bogs which te formed there as low seach w tg i beddin PHYSIOGRAPHIC PROCESSES THE FLOODPLAIN SURFACE The permafrost table gradually rises from the younger to the older parts of the slip-off slope of the floodplain (Figures 5 and 12) because on the older surfaces the mosses are better established and thicker and act as better insulation. Although when frozen the moss mat allows loss of heat by radiation during the winter, when thawed and especially when dry the moss mat and peat preserve frost in the ground during the summer thaw (Muller 1947 p. 53). The oldest and highest floodplain surfaces, with the thickest moss carpet, have the highest permafrost table, both in the Upper Kuskokwim River Region and in the Galena Region, 120 miles to the north on the floodplain of the Yukon River (Péwé 1948). Peat under the forests was largely dried out by the end of July in the Upper Kuskokwim Region. THE Divisions OF THE FLOODPLAIN Péwe in his discussion of the terrain of the Galena area divided the floodplain into four phases. These are clearly marked on the floodplains in the Upper Kuskokwim Region. The criteria he used for distinguishing the phases are drainage pattern, distribution of vegetation, depth to permafrost, elevation, lithology of sediments, character of the river bank, and distribution of drift wood. He has divided the younger surfaces into two phases: I. Linear Phase youngest — Distinct linear lakes parallel to the river, no integration of drainage, large deciduous trees, permafrost absent or low, no large masses of ground-ice. Il. Advanced Linear Phase — Distinct linear lakes, som= integration of drainage, large coniferous trees, slightly higher permafrost table, no large masses of ground-ice. The youngest surfaces of the floodplain (Figures 4 and 6) are the slip-off slopes on the inside of a bend of the meandering river. In its migration, the river is depositing sand and silt, building and extending these surfaces. The surface is ‘rregular, scarred with — the pattern of bars and swales. Floodwaters flow through the larger swales filling them rapidly with alluvium. The first stage of succession on the younger surface is tall willows and alders, BOG FLATS AND PHYSIOGRAPHIC PROCESSES IN ALASKA iz followed by a deciduous forest, Phase I, and then by a tall mixed deciduous and coniferous forest, Phase II. The outlines of the recently abandoned channels and bar-lakes are not modified in Phases I and II except by the colonization of the shallow water by vegetation. The migrating meanders of the river tend to remain within a certain zone, which is the zone of active meanders, rather than wandering over the whole valley bottom. For this reason the surface of the floodplain beyond the zone of active meanders is significantly older. It has not been disturbed by the cutting of the river for a long time. The younger surfaces show the patterns and the forests of recently deposited surfaces clearly, while the older parts show the floodplain patterns partially obliterated. The rise of the permafrost table has eliminated the mesophytic tree species over large areas, and bogs and scattered stunted trees in a “forest-tundra” have replaced them. Transitions do exist and are best shown on the floodplains of the larger rivers such as those of the main Kuskokwim and of the Yukon at Galena, but in general there is a marked break in the conditions of soil and vegetation at the edge of the zone of active meanders. In Péwé’s classification, the older surfaces are also divided into two phases: III. Coalescent Phase — coalescing linear lakes generally at an angle to the river, coniferous trees, some “tundra”, high Permafrost table, and no large masses of ground-ice. IV. Scalloped Phase (the oldest) — irregular shaped lakes with scalloped borders, integrated drainage, stunted coniferous trees, much “tundra”, very high permafrost table, much ground-ice. Péwé’s Phase III is evident on the floodplain of the Kuskokwim behind the tall forest of White Spruce and Birch (Figures 6 and 7). It is the area of transition at the edge of the zone of active meanders. The oldest surfaces of the floodplain are found on the lake- dotted surface near the Big River and Medfra, or again in the Nixon Fork Valley. This physiographically corresponds to Péwé’s Phase IV. On the relatively level surfaces of the main alluvium on the Medfra Lake Terrace lakes are scattered at random and the bogs developed within them (Figure 7). In the Nixon Fork Valley re-entrant, however, the alluvium is gently sloping and there is a down-slope component in the forces influencing the Figg me 7. AERTAL PHOTOGRAPH FROM “AR MEDFRA. Phases I and II are limit 10,000 FEET LOOKING te ted in ase IV is the Medfra ee Tert -, on the fioht i at ee he rolling surface around he bogs is open ‘BI ack Spiace © ; ooo winter and snow CONCE Makes y Air Forces). kes . = bonne S are now occupied by oe de 1€ a“ S the boys EAST UP THE SOUTH FORK FROM muskez. The photograph was appear white. (Photograph courtesy of the FEET L OOKING Tre AERIAL FHOTOGE APH FROM I q and Bsc aawine g of peren ougn th IOS of have expa Movement ra) water Wi d¢ ding ee: ver 07 lightest tone is se rce s) . 18 WILLIAM H, DRURY, JR. formation of bogs. This has led to down-slope orientation of the long axes of the bogs and of the included islands (Figures 2 and 8). Characteristic microtopographic and vegetational features on the bog surfaces have resulted from these same slope components. Because Phase IV does not occur where the Kuskokwim River is actively cutting against the floodplain, the distribution of frozen banks along that river is not as conspicuous as it is along the Yukon in the Galena area. Another difference exists between the Kuskokwim Region and the Galena Region. The ponds and lakes which are open water in the Galena Region, as they are in the Nabesna-Chisana Region described by Wallace (1948), are regional bogs in the Upper Kuskokwim Region. Péwé (1948, p. 24) expresses some doubt that the Phase IV of the floodplain at Galena is simply an older stage and suggests a slightly more complicated geological history. The distribution of this type of surface in the Upper Kuskokwim River Region (Nixon Fork Lowland and Medfra Lake Terrace) agrees. THe Bocs It was pointed out that the rise in permafrost table in the floodplain deposits is correlated with the thickness of the moss mat. Also dependent on the growth of mosses (first Hypnaceaz and then Sphagnaceae) are the establishment and expansion oi the Black Spruce forests and peat bogs. The mosses responsible are highly adapted to soak up and hold water, so that once established they maintain their own water-soaked environment unless exposed to excessive evaporation or fire or are otherwise disturbed. Muller (1947, p. 53) says that peat (without saying what kind) can absorb water up to 300% of its volume. The moisture contained in the bases and dead remnants of sphagnum mosses has a low pH and nearly anaerobic conditions. This has — an antibiotic effect on bacteria of decay so that the remains or plants collect and are preserved as peat. Tansley pote p. 676) has summarized the important effects of sphagnum cording to recent knowledge, based largely on the work of ee (1915) Watson (1918) and Pearsall (1958). Frost in the ground decreases sub-surface ogee and thawing of the annual frost is a source of water. | , for holding moisture by peat mosses decreases esmna run-off. BOG FLATS AND PHYSIOGRAPHIC PROCESSES IN ALASKA 19 The fine-grained materials of the alluvial lowlands also decrease percolation. The low slopes of the lowlands allow the collection of bodies of water, and the short cool summers do not return large amounts of moisture to the air by evaporation and transpiration. Each of these conditions exaggerates the effects of the others, although they are probably nor all operating at any one place or time. The combination of their effects is responsible, however, for the saturation of the alluvial lowlands. Soil movement and alluviation from minor streams has nearly obliterated the break in slope from bedrock hills to depositional slopes and from depositional slopes to the floodplain. Bogs colonize moving water of the minor streams responsible for alluviation at the margin of the floodplain and can occupy slopes as steep 3°. For these reasons there is no sharp separation between the surface of the floodplain and the depositional slopes and it is not possible to say that bogs are restricted to the floodplain and terraces. They appear wherever the collection of water will allow it. FORMATION OF BOGS In the Upper Kuskokwim Region peat bogs form in three general ways: 1) by the colonization of a body of water by peat forming plants, 2) by the invasion of the lowlands by swamp- making vegetation (forest mosses, then sphagnum), and 3) locally by the thawing of perennially frozen ground, depergelation (regional and climatic thawing of perennially frozen ground, Bryan, 1946). The first involves the colonization of a body of open water by immersed aquatic mosses (Sphagnum, Calliergon and Dre pano- cladus), upon whose remains and dependent upon whose growth the vegetative shoots of sedges can find support. Succeeding years’ growth of these plants weaves a tighter and tighter carpet, while their dead parts sink to the bottom and are preserved under the cold anaerobic conditions. Gradually, as plant remains fill the water, less and less hydrophytic vegetation is able to become established and a quaking bog is formed. This process is common in the northern United States, and bogs formed in this way Orrespond to topogenous mires or basin bogs of the von Post (1937) classification. 20 WILLIAM H, DRURY, JR. The second type of origin was described in considerable detail by Cajander (1913). He calls it, the “Versumpfung des Wald- bodens” ®. This process was described in the United States by Shaler (1890) and was well known and earlier described from Europe. In it mosses colonize low areas or depressions and hold moisture collecting there from rainfall, surface run-off or soil- water percolation. The mosses (and the conditions that allowed their establishment) continue and spread, gradually occupying more and more of the forest floor until the sodden conditions kil] off the forest and a shallow bog is formed. A gradual rise in the ground-water level on the alluvial lowlands of the Upper Kuskok- wim is occurring because of inadequate run-off of drainage from the hill slopes and terraces and the other factors mentioned previously. Sphagnum has colonized the wet places and huge areas have become shallow bogs (Figures 6 and 7, Phase IIT). The third method of formation and spread of bogs depends on the presence of perennially frozen ground and the destruction of it by thawing at the banks of a quaking bog. The process is closely related to that described by Wallace (1948), Hopkins (1949) and Black and Barksdale (1949). Taber (1929) and Muller (1947, p. 65) have shown that silts and other fine-grained soil materials tend to absorb moisture on freezing so that they com- monly contain a considerably greater amount of water when frozen than when thawed. This moisture segregates as crystals and lenses of ice of various sizes. The thawing of perennially frozea silt will produce a depression of considerable depth (Muller, 1947, p. 84), 15 feet in some examples in the Upper Kuskokwim. Because the moss carpet acts as an insulation to the frozen ground, thawing can be started by any break in it. In such a depression water will collect, be warmed by the sun and continue to thaw the ground around it. A depression or break in the moss carpet can be caused by a fire, a tree blow-down, the bursting of a frost-built hummock or pingo, a well-used game trail, a bear’s meat cache, the bank of a small brook, or the water that collects in sphagnum invading the forest by the swamping process. Thawing of frozen 10) This is the process for which von Post (1937) has coined the word ification’’, intending to anglicise German term Versumpfung or Swedish wudification , ; Férsumpning. The word has the twin disadvantages of being of combined classical and i i | h origin and meaning no English word more than - common “swamping” can. The word was used by Wenner (1947) referring to the bogs of al usage. | Labrador, however, as if it were in gener, NNR S366 pics coaster nmaacman me eeuie mate ltr ail wa — +4 aes wee by “ . Vis re IGURE 9. LOW ALTITUDE AERIAL FHOTOGRAPH OF THE NIXON FORK FLATS. The ese bogs are eS nding by thawing into the frozen floodplain sediments. On tl left side of al base-level is the river; in t th loccasoand -level is above the channel lace betw ween bogs has on it scattered Black aa and Larch. rmint w 1 shown ts typical of larg raise I d area in the middle effects of soi 2 BOG FLATS AND PHYSIOGRAPHIC PROCESSES IN ALASKA 21 silts destroys their coherence, so that they slump and increase the depression. Once the thawing has started, it expands in directions and the pattern that is produced depends on the direction and strength of the forces that affect the rate of retreat of the banks (Figures 2, 8, 9, 10, and 22). Thawing expands from neighboring centers of origin to coalesce, and the result is a large area of anastomosing bogs with isolated islands or peninsulas of the former floodplain surface. The banks of these bogs are abrupt (Figures 9, 10 and 22), and the moss and heath mat of the forest floor hangs over into the bog or is broken, exposing frozen or thawing silt. At the edge of the banks a depression is formed which fills with water. It is called a moat ere to separate it from similar features of other types of bogs 1’. The moat channels run-off and melt-water from the thawing surface and encourages further thawing of the frozen bank. Aquatic sphagna and sedges colonize the water-filled depression as they do in bog succession in a pond. Less hydrophytic sphagna and sedges follow which allow the growth of heath and birch shrubs as the mat tightens. In this way the bog vegetation occupies the expanding depression. the swamp-lowland of Nixon Fork Valley (Figures 2 and 8) the bogs are oriented parallel to drainage channels, and the islands have their long axes parallel to the line of slope. This indicates that there is a significant down-slope component to the movement of water in the bogs. It shows evidence of the relation- ship of these bogs to the soligenous types of Scandinavia in origin or physiography, as well as in the vegetation and floristics that will be described in detail later. gc SO 11) A moat is a deep-water canal, often 4-6 feet deep, at the edge of a bog that 1s actively expanding by thawing of the surrounding frozen alluvium. A moat 18 Wider than the marginal channel of most American bogs (called Yazoo stream fe Patts of the Pacific States) which is formed by the break in vegetation resulting som the rise of the floating bog-mat when flooded in the spring. A lagg of Hochmoor or ombrogenous mires is formed as a result of drainage of water from a convex bog. ki -» Moat may consist of an area of open water several yards across, with pond ilies and other aquatics in it, but in most 2s it is a broad zone of coarse Stowing in deep water and closely resembles a /ugg. Skeletons of trees he bogs as shown in Figure 10, are common in the ot Most ate Got dniverssl an the Lppes: Kaskokwin River Region. bul ‘tive expansion of bogs is similarly not universal. ho i] WILLIAM H, DRURY, JR. DISTRIBUTION OF BOG TYPES The three processes of bog formation of course do not segregate neatly into regional types, but on each of the three surfaces which show a maximum development of bogs, one of the processes is most conspicuous. First, on the floodplain of the Kuskokwim and many of the larger tributaries of the main valley, especially Péwé’s Phase III, the bogs form largely by colonization of ox-bow and bar-lakes and by invasion of the floodplain forest by a swamping correlated with rise of the water table (Figure 6). Second, on the older part of the floodplain surface, Péwé’s Phase IV, near Medfra (Figure 7), the bogs have formed by colonization of the shallow water of lakes randomly scattered over the surface. Large bog areas on this surface also result from expansion of bog vegetation along minor drainage channels between lakes. Third, on the lowland in the reéntrant occupied by the Nixon Fork of the Takotna, also Péwé’s Phase IV, are those bogs formed by thawing of perennially frozen ground along drainage channels (Figures 2, 8, 9 and 22). Bogs of the type found widely on the younger parts of the floodplain of the Kuskokwim (Figure 6) are large and irregular in outline, often reflecting the patterns of deposition of the alluvium (the ox-bow and bar-lakes, bar and swale topography). The forested parts are usually close to the same level as the bog surface, and the contact is on a gentle slope. Tongues of bog extend into the forest and large portions of the interior of the forest have a thick, soft moss mat where the swamping process is starting to have an effect. In some areas the non-bog surface is higher, has scalloped outlines, and is separated from the bog by an escarpment. Islands formed by slump from terrace escarpments are found in this type of bog. The forested surface has a dense mixed forest of Black and White Spruce or tall Black Spruce. The bogs on the oldest part of the floodplain near Medfra have colonized scattered irregularly shaped basins (Figure 7). In them, striking flat-iron-shaped lakes persist. The lakes were originally formed as a result of processes that are not understood but which involve depergelation. The outline of the lake basin is often scalloped and the basin oriented as if under the control of forces similar to those suggested by Black and Barksdale (1949) perennially — and Hopkins (1949), migration of lakes by thawing of BOG FLATS AND PHYSIOGRAPHIC PROCESSES IN ALASKA 23 frozen ground in their leeward banks as a result of the effects of warm winds and waves. This is clearly not the entire ex- planation, however. Bog vegetation has colonized these lakes and the edge of the carpet subsequently has been shaped into long crescents by the wind and ice. The explanation of these smoothed, triangular shore-lines involves the fact that in the spring when thawing, which takes place first along the shore, has freed an island of ice in the center, this ice is blown around by the wind. It bumps repeatedly against the lee shore and crushes and flattens the pliable vegetation carpet. The waves stirred up by summer winds beat against the same lee shores and erode them. Both forces act only at the edges of the carpet of vegetation and for that reason buckle the edges into the peat ridges which are conspicuous elements of the margins of bog-margined lakes. The triangular lake basin at present is independent of the original outline. In addition to the bogs which surround the lakes, bogs on this Medfra Lake Terrace have expanded up drainage channels and Occupied nearly all low places by swamping. The surface has a subdued rolling topography covered with a thin, stunted Black Spruce forest. Intricately interconnecting bogs extend in and out of the low, rolling surface. The bogs found in the reéntrant valley of the Nixon Fork (Figures 2, 8 and 9) are formed almost entirely by expansion from drainage channels by thawing in the bog banks (Figure 10). They occur on slopes of 1°—4° and accordingly vary in the degree to which down-slope movement of water affects their shape. Those on nearly level surfaces have scalloped margins and shapes similar to the lakes illustrated by Péwé (1948) and Wallace (1948) (Figures 9 and 22). Characteristically, the bogs are elongat- ed down slope, have abrupt margins, no describable shape and are characterized over most of their area by the patterns of bog vegetation called Strangmoor in German (Cajander, 1913). In Parts of the Nixon Fork Lowland they have obliterated the forest from areas measurable in tens of square miles (Figure 2). The forests in these lowlands are of tall Black Spruce and Larch. TRANSFORMATION OF THE BOG BASIN As soon as a wet depression or body of water is colonized by bog vegetation, slow processes start to fill it with plant remains. 24 WILLIAM H. DRURY, JR. They may eventually transform it to a peat-filled forested hollow. Changes in water level or in frost activity in the bog can, however, reverse the direction of this change and return the area to open bog or shallow water. This is not a unidirectional development toward a climax vegetation as recent discussions of peat bogs in Alaska seem to indicate. With the establishment of vegetation, collection of plant remains starts. Dead mosses, sedge roots and rootstocks, culms and stolons are preserved in the cold, anaerobic conditions that inhibit the growth of bacteria of decay. But percolation of soil water through the matrix of this organic mass allows partial decomposition of much of the organic material and penetration of silt. Silt-bearing winds, annual spring or summer floods, water moving through the bogs, and the mineral soil contributed to the bog by the slumping of the banks add mineral material to the growing mass. During the spring thaw the whole lowland is a sea of moving water, sometimes clear, sometimes cloudy with silt. As the peat and silt deposits build up and are consolidated by their own weight, they form a firmer substrate which allows colonization by progressively less hydrophytic vegetation. The firmer substrate provided as the peat builds up allows the estab- lishment, eventually, of woodland sphagna, and brush (heaths and dwarf birches). Finally, a Black Spruce, Larch, or mixed forest can appear on the surface, growing on a consolidated deposit of peat and silt. This organic soil (Radforth, 1952) is protected from loss of heat by the moss carpet, and, as it is increasingly com- pacted, is less and less subject to the continued flow of bog waters which tend to unify its temperature. Finally, water no longer percolates freely and under these conditions the ground may freeze and remain frozen, forming perennially frozen ground. Freezing leads to the establishment of discrete ice lenses in the deposit, increasing the volume of peat and silt (Muller, 1947), and the result of all these processes is to raise the level of the forest floor. The drier and more stable conditions in the floor which result are usually marked by expansion of the forest and extensive growth of lichens. The forest and its underlying deposits are similar to those originally destroyed by bog advances. The advance of bogs into the frozen alluvium by thawing is in action at the same time that the collection of organic remains is filling sodden depressions. The two antagonistic processes form BOG FLATS AND PHYSIOGRAPHIC PROCESSES IN ALASKA Pass a cycle which is in action on the alluvial lowlands at present and has had wide-reaching effects on the sediments of the lowlands in the past. GEOBCTANICAL PROCESSES AND SEDIMENTARY DEPOsITs Field observations which must be accounted for include: 1) the Present patterns of bogs, of forests being undermined, and of patches of developing forest; 2) the deposits exposed in river-cuts under present areas of bogs. The vegetation patterns are to be explained as dependent on microtopographic features produced by thawing of frozen silt at the advancing margin of the bog, and by other geological processes in the substrate, rather than simply “vegetational development”. They will be discussed in the chapter on Vegetation. EXPLANATION OF FIELD OBSERVATIONS The deposits found under present bogs can be explained in terms of processes visible on the floodplain surfaces. Strata of peat are explained by the presence of peat-forming vegetation on the floodplain during aggradation. The interbedding with river-borne silts points to periods of shifting channels of the river and to floods. Peat layers near the surface underlie living peat-forming vegetation. An explanation of the formation of mucks remains. Tuck (1940), Eardley (1937) and Taber (1943) have discussed the formation of muck and Péwé has in preparation an interpretation of muck deposits in the Fairbanks district. All of these inter- Pretations indicate that muck formation occurs on valley sides. In general they suggest that clouds of wind-borne silts have been carried from the bare river beds of braided streams and deposited On the hills; these silts are moved by solifluction and alluviation down to the val y sides, incorporating organic material into the deposits as they gather. Péwé (1950, 1951) has shown that much of the silts of the uplands in the Fairbanks region is wind blown (loess). He has explained muck formation there to me as transport by small streams of loessal deposits from the hills into the valley Sides where they are incorporated as minor delta deposits with finely divided organic material, and then frozen. These ex- 26 WILLIAM H, DRURY, JR. planations do not satisfactorily account for muck deposits in valley centers. In the 80-foot terrace, 75 miles upstream from McGrath, there is a deposit of muck about 35 feet thick (Figure 3), but the Kuskokwim River lies between this muck deposit and the nearest bedrock hills. It may be that the Kuskokwim River once drained South between the Alaska Range and the Lime Hills, but until this is proven to be the case, solifluction or minor streain deposition cannot have deposited these sediments across the river, and the nearest other source of the solifluction materials is the Alaska Range, 40 miles away. I believe that the geological and botanical processes in the regional bogs are adequate for the formation of such deposits. The materials called muck are nearly always frozen. For their formation it is assumed that perennially frozen ground plays an important part and it does in the bog-forest cycles in the Upper Kuskokwim Region. Climatic conditions have to be such that finely divided organic material is preserved in a semi-decomposed state (peat) while being thoroughly mixed with fine-grained mineral material. Maddren has pointed out what is conspicuous, that peat is the most widespread and characteristic of recent deposits (1910) *. Two processes are available for the formation of mucks: 1) repeated cycles in the advancing of bogs by thawing, (which incorporate slumped materials into the bogs) or swamping (Ver- sumpfung); 2) collection of aeolian and alluvial silt on the sur- faces of sedge-meadow bogs. The soligenous nature of these bogs will lead to the incorporation of the mineral material into the peaty material below. Both of these processes must take place during a period of stream aggradation. Cycles 12) The quaking bogs involved in the cycles are often deep and can act as raps . entomb large . The inclusion dismembered, often unbroken, fossilized large-mammal remains is a characteristic of mucks in some areas ages Aetna ewan sti nina vimer~ren =the vr oo enn smectite BOG FLATS AND PHYSIOGRAPHIC PROCESSES IN ALASKA 27 chance. The result is a thorough stirring of the various strata, obliterating their original sequential stratigraphic positions. Successive cycles will disrupt stratigraphy throughout the sedi- ments and produce a more or less homogeneous mixture of sediments with organic material. A secondary possibility is one of a long series of cycles of forest and bog with the destruction of the forest by swamping (Versumpfung). Bog processes restore the forest, and a subsequent or continued rise in ground water could drown the forest again. This would produce a varying pattern according to the various levels in the area, and according to the rate of plant growth. Some areas could keep ahead and be in forest while others would lag behind and be incorporated into the bogs. A cycle similar to the freeze-thaw cycle would result, and it would not be directly dependent on the behavior of perennially frozen ground. But in this case the deposits would be alternating strata of peaty and silty materials or stratified organic silts. Soligenous Mires Or areas of Carex rostrata, Eriophorum angustifolium or E. Chamissonis). Flow of water through bogs is obstructed by the Vegetation, and any load the water may carry is soon deposited. ne tlow of water through the bogs can be considered as an agent of transport of materials from the hill slopes, carrying the fine- grained materials of residual soils or loess. Spring melt-water floods spread considerable quantities of river-borne materials over the lowlands, and wind-blown silts contribute directly to deposits in the bog. In fact, loess deposition on the surface of an active bog in itself is adequate to create muck by incorporation of the silt into organic remains. These inorganic deposits mix with the Plant remains, settle or are carried by water into the bogs, and are compacted with the peat. Repeated tests by borings to depth; 28 WILLIAM H. DRURY, JR. up to 6 feet in bogs that are part of the flats system indicate a conspicuous rise in mineral content with increasing depth. This is not the case in typical raised bogs (ombrogenous) where borings indicate only peat at depth. Cycles of disturbance of the strata laid down in the bogs, and repeated exposure to circulation of water carrying some oxygen and greater or lesser amounts of silt, will break up and distribute the organic material and mix it with inorganic, producing an unstratified deposit of finely divided organic materials and silt. Circulation of water will also allow partial decomposition of the organic material. The rivers and streams may sweep back and forth over the valley, seeming to erode. In the same way the thawing activity of the bogs may seem to be grading the lower slopes to the stream banks, but if the overall pattern is one of stream deposition the sediments in the valley will be thickening. The development of deep deposits of such muck can be visualized in terms of repeated cycles of bog and forest, freeze and thaw, or of silt collection in a soligenous bog during a cycle of stream aggradation. These processes explain the formatiog of muck deposits in valley centers such as those along the Kuskokwim and along the Yukon at Galena (Péwé, 1948). BASE LEVEL AND GRADED SLOPES IN THE BOGS Bogs form in flooded hollows and along water-courses and advance by swamping or slumping due to thawing. The minox drainages which control the water level in such places also control the bogs and it is to be expected that the bogs are adjusted to the base levels of those streams. In general, the processes of erosion accompanying bog development lead to the formation of graded slopes and eventually the bog surfaces will be graded to the master streams. Actually, over most of the Nixon Fork Lowland for example, as aerial photographs show (Figures 2, 8 and 9), streams meander across the lowland in channels incised into the alluvial surface, and the bogs are in adjustment to temporary base levels on the stream banks rather than to the main streams. In some parts of the valley, such as near Washington Creek (Figure 1), the bogs — are connected to such through-stream channels by a number of BOG FLATS AND PHYSIOGRAPHIC PROCESSES IN ALASKA 29 brooks. In this case the temporary base level is high enough to allow development of a new series of minor streams and the bog surface must become readjusted to the new base level. Figure 9 shows a series of bogs which are graded to at least two levels, One group to the river bank and the other to the river level. By thawing into the surrounding deposits the bog surface graded to the new, lower base level will migrate upslope producing a knick point and establishing a new graded slope. Although graded slopes exist in the general terms just described, in detail they are interrupted by micro-relief developed by plant growth. First, the bog phenomena are of short duration when compared to the cycles of aggradation and degradation in the main stream. Redevelopment of frozen ground and forest cover Prepares the way for another cycle of invasion by bogs while the base level at the river bank is unchanged, and the advance of the bogs into the frozen surface breaks the continuity o7 graded slopes. Second, as will be discussed later, bog ridges act as dams and produce step-like low terraces in the graded swamp surface. The bog hollows between ridges are flat, and the next bog hollow is on a lower level as if on a low step. Third, when the ground freezes its volume increases. In an area where parts of the surface are underlain by permafrost and others by thawed materials, there will be irregularity of the surface according to the conditions of frost in the ground. The forested ground, underlain by permafrost, is graded to a previous base level, while several new base levels control the slopes of the bogs. If the bog advancé is completed, the bogs will undermine and destroy the forested surface. But before the forests are eliminated and the whole alluvial lowland graded to the bogs’ temporary base level, the bog-filling processes re-establish a peaty substrate which freezes and is no longer a part of the graded surface of the bog. In these ways the botanical processes have interfered with the details of establishment of graded slopes. Furthermore, even Seologically, it is rare for a cycle of stream erosion to run its course fully and to produce slopes graded to a base level so that it fits the definition of the proximal portion of a peneplain. Nevertheless, the slopes of the lowland of the Nixon Fork Teentrant are the closest to graded slopes of any in the Upper Kuskokwim River Region. VEGETATION THE VEGETATION OF THE LOWLANDS The changes in the floodplain surface that have been described geologically are directly reflected in the vegetation. The vegetation is closely dependent upon the drainage, the sediment size and the condition of frost in the ground and because of this accurately indicates the changes in these by zonations on the floodplain surface. It is most convenient to describe the vegetation in the categories Péwé established for the floodplain at Galena. Figure 5 is a diagramatic representation of the vegetation zonations on the floodplain of the Upper Kuskokwim River, and Figure 12 a summary of the vegetational changes. Phase I (Linear Phase) of Péwé’s classification of the flood- plain (Figures 4 and 6) is divisible into three vegetational sections. 1) The recently exposed mud and sand bars are colonized by Horse-tails (Equisetum arvense) and Felt-leaved Willows (Salix alaxensis var. longistylis). 2) The first mud banks and levels just above midsummer high water (about 4 feet above the average river level in the summer) are populated by dense stands of Felt-leaved Willow which may be 30 feet tall. Slightly higher surfaces are populated by other species of willow (Salix arbusculoides, S. Bebbiana) and alders (Alnus tenuifolia) in addition to the Felt-leaved Willow, which is still the most conspicuous. Beneath these, Horse-tails (Equisetura arvense) and some grass (Calamagrostis canadensis) cover the ground. : 3) The first floodplain surface on top of the banks, is forested with Balsam Poplar (Populus balsamifera). The floor is open and on it scattered stands of the following occur: grass (Calamagrostis canadensis), Horse-tails (Equisetum arvense, E. pratense), herbs such as Pyrola (Pyrola secunda, P. grandiflora, Moneses uniflora), Bunchberry (Cornus canadensis), Fireweed (Epilobium angusti- folium), Twinflower (Linnaea borealis var. americana), ar (Arenaria lateriflora), Mountain Cranberry (Vaccinium Vitis- Idaea var. minus), low shrubs such as rose (Rosa acicularis), Currant (Ribes triste), High-bush Cranberry (Viburnum edule), Osier (Cornus stolonifera) and varying numbers of willows and CUT-OFF MEAN PHASE | STANDING WATER gonses ST, E NUPHAR POLYSEPALUM get - ape NYMPHAEA TETRAGONA FILLING] OXBOW MENYANTHES TRIFOLIATA POTENTILLA PALUSTRIS -- aremnweaen 4: - — _ COLONIZATION BY Pies eho as tape retee s ee Eat Shee: Mote, MOSSES AND WHITE SPRUCE EQUISETUM FLUVIATILE we . . CAREX AQUATILIS CAREX ROSTRATA ; ee rs CALAMAGROSTIS CANADENSIS ‘Se “7s PLEUROZI Im ScHREBER!. PHASE |! CAREX PHYSOCARPA Om %, HYLOCOMIJM SPLENDENS ‘Sy WILLOWS @ ALDERS. ay, We > FLOODPLAIN WHITE SPRUCE MIXED an SPHAGNUM OBTUSUM Ss, SPHAGNUM SUBSECUNDUM % Og % SPHAGNUM SQUARROSUM c spHaGnuM *FUSCUM %e, SPHAGNUM RUBELLUM PHASE II! STANDING A SPHAGNUM neeorve SPHAGNUM Mim wraiene

+ ACROCLADIUM cuserD- SPHAGNUM BALTICUM * saa races cae ee , REPANOCL: VERNICOS:| SPHAGNUM RE [Cc URVUM WARNS, Taken from Journal of Ecology. 14 (1926) 40 WILLIAM H. DRURY, JR. Classification of vegetation on this basis emphasizes coincidence of requirements and deémphasizes any assumption of intrinsic relations between the plant species requiring that they attain an equilibrium amongst themselves. This does not assume cither that plant communities are purely random aggregations or that there is no relationship between species. The comparisons made later between Alaskan bogs and Scandanavian bogs argue against this. This vegetation cannot be considered under systems based on climax concept nor should it be rejected as weedy (Griggs, 1934). This problem is considered further in the Discussion. Three major divisions of the vegetation of boggy and marshy places are made according to the conditions of the water table (Figure 13): 1) silt-bottomed, shelving beaches, 2) fully developed peat bogs with a high water-table, and 3) silt-shored ponds invaded by bog vegetation. Fully developed peat bogs are similar to ombrogenous and soligenous mires of European workers, and similarly possess a flora poor in species. Their high water-tables are maintained by rainfall, stream flow, or groundwater flow. According to strict circumscription, the vegetation of silt- bottomed, shelving beaches is not a part of the bogs system. However, wet sloughs or lake shores can be invaded by bog vegetation and become part of the bogs system as division 3 of this classification. Bog areas which have had this kind of history show the original beach flora and additional species typical of disturbed areas. The flora and individual species behave in the way that Anderson has described for /Jris in the delta of the Mississippi (1949, esp. chapter on Ecological Basis). The sequences of vegetation under the two main regimens active on the floodplain surfaces are diagramed in Figure 20. The left column shows the effects of bog expansion by thawing and formation of a quaking bog carpet by plant growth at the same time. The right column shows the filling of a slough with silt and coincident changes in shore plants. Figure 16 shows maps of the flats area most intensively studied. Map A, small scale, shows about half of the whole of the particular flats and on it are marked the portions enlarged in the others, on which are mapped the major vegetation types. The flats area is on Phase III of the floodplain and is modified by extensive slope-wash and alluviation from Roundabout Mountain and Candle Hills. Because BOG FLATS AND PHYSIOGRAPHIC PROCESSES IN ALASKA 41 the scale is small, details of Strangmoor patterns have often had to be omitted. Vegetation of Silt-bottomed, Shelving Beaches: ach Vegetation This concerns silt shores where there is gradual lowering of the ground-water table or filling by river-born silt. Such pond shores are characteristic of the deciduous and White Spruce-mixed forests of the young floodplain surface, Phases I and II. Most of the lakes are old ox-bow or bar lakes. Many such ponds, however, exist in dune hollows and similar depressions on terrace surfaces. The vegetational zones on the beaches of these lakes are diagramed in Figures 12, 13 and 20. They are as follows: 1) shallow water vegetation, essentially floating; mosses: Calliergon cordifolium, C. giganteum, C. stramineum, repanocladus aduncus, D. badius, D. exannulatus, D. fluitans, Sphagnum Lindbergii. higher plants: Potamogeton epihydrus var. Nuttallii, P. gramineus var. maximus, P. natans, Nuphar polysepalum, Nymphaea tetragona. Ranunculus Gmelini var. yukonensis, Hippuris vulgaris. -_ 2) shailow water vegetation, emergent; “ < Equisetum fluviatile. : Glyceria borealis, G. maxima var. grandis, G. pulchella, Carex aquatilis, C. rostrata, Potentilla palustris. 3) wet shores; Mosses: the same with the addition of Sphagnum obtusum, $. squarrosum, S$. subsecundum., higher plants: Calamagrostis canadensis. Salix arbutifolia, $. pulchra. On the margins of this, next to the White Spruce-mixed forest, there are willow and alder thickets, most conspicuously Salix ll Hip pedal F pibtetass Dye es, re, aa MCE aves: a eRe Bsa A ee ; a ten S09 SS rises Agee ae ae 4 Ay & Yer a itrt ana ey Ake y feet} 100 YARDS Nenana 50 YARDS ————t———j 100 YARDS F"%]] PICEA GLauUcA [| CALAMAGROSTIS CANADENSIS TE] PICEA MARIANA («= «CAREX PALUDIVAGANS SEDGE MEADOW FLOODED 80G : EMERGENT OPEN [] Low erusH CAREX AQUATILIS ["] Carex RosTRaTA oe AQUATICS” waTER 16 Fic 16. TATION MAPS OF A FLATS AREA, ee pi 8 are of part of the bog area across the river, west of McGrath. The entire flats at this face | is * about ues times the area we in Ma s bordered on the North ae in Map A) by the Tatalina hye on the East and South by ge Kuskokw cei and on the West by the done: # gs dabout Mountain. = the small scale of these maps it impossible to show details hog ap pas . Areas of these (Strangmoor) occur on ota top-left of Map B and at bie top of Map A. Areas eee in black at the top-left of Map B and in the ‘sedge meadows on the right of Map D are the wet centers of i hollow BOG FLATS AND PHYSIOGRAPHIC PROCESSES IN ALASKA 43 Bebbiana and Alnus tenuifolia, but with scattered patches of S. alaxensis, S. arbusculoides, and A. crispa. Then finally there is the forest of the floodplain, whether it be White Spruce in pure stands, White Spruce-Poplar, White Spruce-Birch or a mixture of all three. The zones are established primarily by extrinsic factors dependent upon the water-table. A rise of water-table will shift all the zones toward the wetter types and a lowering of the water-table or addition of silt deposits leads to colonization by less hydrophytic vegetation. Vegetation of Peat Environments and High Water-Table: Peat Bog Vegetation The water-table of peat environments is high and steady, and Peat-forming mosses have invaded the stagnant or slowly moving water. The conspicuous development of Carex communities and the patterns of vegetation on valley slopes indicate, according to the Scandanavian studies, that there is a significant soligenous element to the bogs of the Upper Kuskokwim. The soligenous water (soil-water-flow) includes mineral elements not usually found in the rainfall collecting on the surface, which is the only source of water for a raised . The structure of soligenous mires is correlated with lower precipitation and more down-slope Water movement. Changes in vegetation under the regimen of Peat habitats are diagramed in Figures 12, 13 and 20. For purposes of description, the vegetation of the peat bogs will be divided into three headings: 1) areas of relatively con- Sistent simple stands; 2) areas of patterns of anastomosing bo; tidges and intervening bog hollows; 3) areas where raised peat tidges form beaches along the edges of ponds and lakes. All of these occur in the bog area mapped in Figure 16. Areas of relatively consistent, simple stand. This heading is further subdivided into six categories: moats, hollows, sedge meadows, Scirpus-Myrica, low brush, and high brush. These are life-form categories as seen in the field, and in Many bogs these simple stands occupy most of the surface. Although this vegetation has the weedy characteristic that it Consists of large expanses of a few species, a regular internal Structure is present, and species groupings are repeated with 44 WILLIAM H, DRURY, JR. regularity within the same bog, between bogs, and between areas. The simple stands fall into three aspect types: a) Conspicuous vegetation meadow-like; includes moats, bog hollows, sedge meadows, and Scirpus-Myrica areas. b) Conspicuous vegetation low heaths and Dwarf Birch shrubs, not more than 6 inches high. c) Conspicuous vegetation of tall shrubs, about 4 feet high. In all cases, the associations may contain a moss carpet or have a mud bottom. In either case, the bottom may be firm or yielding, but in general the bottom is more dependable where there are shrubs, even though there may be a moss carpet that yields to a depth of two feet. Some sedge areas are nearly as firm as the sphagnum areas in Black Spruce forests; others yield to a depth of 4 feet. Where the sedge areas have a mud bottom, as in the bog hollows, they are especially insecure. a) Conspicuous vegetation meadow-like. 1) The wettest areas of simple stands (which often are found in the moats) include Carex rostrata, Carex lasiocarpa, Carex aquatilis and a very few mosses, some Calliergon, Drepanocladus, and Sphagnum Dusenii and S. Lindbergit (Figure 10). 2) The next wettest areas have mud false-bottoms at about 3 inches and scattered emergent aquatics and sedges. They are shown in black on Map B of Figure 16 and those black areas in the sedge meadow in Map D. Scorpidium scorpioides and Drepanocladus revolvens occur sca in the shallow water, as do Carex limosa, Carex rotundata, Carex chor- dorrhiza, and, less conspicuously, Eriophorum angustifolium, Carex livida, Drosera anglica, Menyanthes trifoliata, and Utricularia tnter- media. In some places the bottom may be carpeted with Campylinm stellatum, Hypnum Bambergeri and Tricostemum cuspidatissimum as moss layer; Scirpus cespitosus and Myrica Gale are characteristic of such a place, or Scirpus and Myrica with Scorpidium and Drepanocladus. areas are called bog hollows here (Flarken or Rimpi by the Scandinavians) and are characteristic of the spaces between bog ridges in a Strangmoor. 3) Next, the sedge meadows are the driest of the meadow-like vege- tation. They consist of Sphagnum balticum, or S. Dusentt, (S. Dusenti in wetter areas) with Carex limosa, C. chordorrhiza, C. rotundata, and less often C. rostrata, Eriophorum Chamissonis and Carex paludivagans. They are marked as sedge meadow in Figure 16. b) Conspicuous vegetation low heaths and Dwarf Birch shrubs ubs. A firmer carpet is made by colonization by Sphagnum papillosum BOG FLATS AND PHYSIOGRAPHIC PROCESSES IN ALASKA 45 or S. fuscum, heaths and Dwarf Birch. The first arrivals among the are Andromeda Polifolia and Myrica Gale; then Betula nana, Vaccinium uliginosum and, less often, Chamaedaphne calyculata. Vac- cinium Oxycoccos is an abundant plant at this stage, but is inconspicuous because of its size. At the same time Vaccinium Vitis-ldaea and Rubus Chamaemorus appear. This type usually grows in a tight matrix of sedges and sphagnum. This is shown as low brush in Figure 16. C) Conspicuous vegetation of tall shrubs, about 4 feet high (Figure 11). The high brush areas become established coincident with Sphagnum papillosum and S. fuscum in which are growing shrubs of Myrica Gale, Betula nana, Betula glandulosa, often Potentilla fruticosa, Ledum decum- bens, Chamaedaphne calyculata, Andromeda Polifolia and Vaccinium uliginosum, all about 3 feet high. Among these are a few Black Spruce and Larch. Following the thickening of the Sphagnum fuscum mat, a forest cover eventually is developed, if the water level has been lowered enough. is type is not mapped in Figure 16. In many places where a shallow layer of water from a hillslope runs across the bog surface (Uberschwemmungsmoor) the 4-foot shrub vege- tation of Scirpus-Myrica grows with scattered mosses or no moss carpet. Large areas are like this at the heads and margins of streams that run through the bogs. This type is shown as flooded bog in the top left-hand corner of A, and ar the top of D in Figure 16. The bog surfaces at some distance from wet centers and especially bogs where there are no peat ridges have a very wide- spread high brush cover of heaths, Myrica and birch. These parts seem to have reached comparative maturity or stability of Vegetation (a subjective impression). Near stream channels, moats, regressive areas or on “new” bogs, sedge meadows and low brush are found. The high brush type seems to converge with the brush tundra or wet tundra type of vegetation and may reflect an €special success of that life form and group of species. All across the north the heath-birch brush seems to be especially successful in and out of the forested areas and may represent the closest approach to a climax concept, at least as DuRietz (1936) uses the idea, developed first by Gams (1918), of considering vegetation by one-layered communities. Brush constitutes the stable bog Vegetation until drying out allows occupation by Sphagnum fuscum-Black Spruce or further bog cycles invade the area. Patterned Surfaces (Strangmoor) Patterns of bog ridges are found in all parts of the bogs, but they usually are best developed at the upper ends, lower ends or 46 WILLIAM H, DRURY, JR. around the edges and away from central bodies of water. Their distribution seems to depend upon the presence of slope in the surface of the bog. There are rwo parts of the pattern, the bog ridges (Strange or Pounut) and the areas between or bog hollows (Flarken or Rimpi). Related to these nets in their vegetation and presumably ii the physiographic process which creates them is a pattern of Black Spruce trees spaced as regularly as if in an orchard. The two patterns are discussed in the section about vegetation patterns on the bog surfaces. The patterns of both types, ridges and patches, show especially well on aerial photographs. The bog ridges may form into a net-like pattern (Figures 18, and 19 bottom diagram) or into festoons and parallel ridges like waves on the ocean (Figures 14, 15, and 19 top diagram). The festooning bog ridges are found in the wettest parts of the bogs. They “drag” against islands and “belly out” in the zone between. A large area of this type of bog ridges was found just at the top of Map A, Figure 16. Bog ridges are more numerous and the bog hollows narrower on the margins and lower edges of bogs. In such places the pattern is often expressed as a net rather than as festoons. A net pattern is developed in the top left corner of Map B, Figure 16. The bog ridges are narrow (often about a yard across althouga they may be ten or more) and may be several hundred yards long. The areas between (bog hollows) are wetter with less compact vegetation; often their centers are shallow, mud- bottomed pools with scattered emergent aquatics. They are often 30 to 40 yards across. The length of their long axis, like that of the bog ridges, depends on the bog in which they are found, because they often extend from shore to shore. The long axes of the bog ridges and bog hollows run across the slope or lie at right angles to the direction of flow of water. The bog ridges act as dams and resemble terracette margins, while the surface of the intervening bog hollow is level. Although there is consistently a vegetational difference between the bog ridges and bog hollows, the difference is not always the same. The only safe generalization is that more mesophytic species of sphagnum, sedges and heaths occupy the ridges, more hydro- phytic ones the hollows. There is a sharp segregation of species Figure 15. ete. Ficu RE 17. FROZEN POOL . T ROCKPORT. MASSACHUSETTS, FEBRUARY 1952. The surface plan of ridges is similar to that of a net of bog ridges. Wat tank through the outlet in the left foreground | flowed diagonally away toward the figure in the backgr ann freezing as it mov no indication of the collapse of a temporary sheet of ice _ FIGURE 18 BOG RIDGES ON A STRANGMOOR. Bog ridges are shown on a sedge meadow in the bogs west of McGrath. The sedge aie is very poorly consolidated and will support the weight of a man only briefly. A man’s weight on one of the ridges causes all of the trees on that ridge to tip towards him. The packboard supplies scale. This pi ff the i map, above and to the right of Map A, Figure 16. BOG FLATS AND PHYSIOGRAPHIC PROCESSES IN ALASKA 47 involved. Only rarely is the pattern of the ridges due only to differential species abundance, which seems to be often the case with vegetation patterns such as rings and polygons in the tundra. Occasionally, within a large scale pattern of bog ridges and hollows at the edges of bogs, there may be found a pattern of minor ridges across the long axis of the hollows of the larger pattern. This is described later. Vegetation of bog ridge and hollow areas arranged in sequence from the wettest to the driest. (a) The wettest centers of bog hollows (the Braunmoor, of Auer, mud or Carpet vegetation of Sjors !*) have a sparse moss flora of Drepanocladus revolvens and Scorpidium scorpioides; usually Utricularia intermedia is consistently present, and, occasionally, a few plants of Utricularia vulgaris. The emergent vegetation is a scattering of plants or clumps about 10 inches apart growing in about 4 inches of water. The bottom is mud and yields to a depth of 4 feet, more in some Cases. The most usual higher plants are Eriophorum angustifolinm, Carex livida, Carex limosa. Carex rotundata, Carex rostrata, Myrica Gale, Drosera anglica, and Menyanthes trifoliata. Sometimes the centers Contain Scirpus cespitosus, Myrica Gale and Campylium stellatum. (b) The margins of the centers (Weissmoor, or carpet or lawn com- Munities), are characterized by the presence of Sphagnum Dusenii, Sphagnum recurvum, or S phagnum balticum and sedges whose decum- bent stems and other underground parts weave together into a carpet, holding the sphagnum more firmly. Sedges found here are Carex chor- dorrhiza, Carex limosa, and Carex rotundata. Occasionally some of the elements of the wettest centers are important in this zone, and often there are scattered shrubs of Andromeda Polifolia. Less often Betula nana, Ledum decumbens, Chamaedaphne calyculata and Vaccinium uliginosum colonize from the drier sites. The shrubs are found as scat- tered individuals or in patches. Such patches are often marked by a (A #liginosum and Betula nana). Once the Sphagnum papillosum is estab- 15) og terms are included as examples of their application to Alaskan bogs. d : They wil iscussed in the section comparing the bogs with Scandinavian f€s¢arches. 48 WILLIAM H. DRURY, JR. lished, scattered plants of Eriophorum Chamissonis, Drosera rotundifolia and patches of Sphagnum fuscum and Sphagnum rubeilum are found in it. (c) The bog ridge in the area studied is usually a high brush (Auer’s Reisermoor or Bruchmoor or Sj6rs’s hammock communities). The brush grows on moss ridges of dense Sphagnum fuscum with lesser amounts of Sphagnum rubellum and in some places S. robustum. They are super- posed upon peat or dead mosses of the vegetation zones described in (a) and (b). With the Sphagna appear Pleurozium Schreberi, Ledum decum- bens, Chamaedaphne calyculata, Vaccinium Oxycoccos, Vaccinium Vitis- Idaea, Vaccinium uliginosum, and Rubus Chamaemorus. With the above, but less extensively, there appear Carex lugens, Geocaulon lividum, Betula nana, Drosera rotundifolia, and Empetrum nigrum Lichens appear as the substrate becomes firmer and drier: Cladonia rangiferina, Cladonta sylvatica, Cladonta amaurocraea, Cladonia cyanipes, Cladonia deformis, Cetraria islandica, Cetraria cucullata, and Nephroma arcticum. Although they are relatively few in actual numbers of indivi- duals, where they appear perhaps Picea mariana and Larix laricina are the most characteristic plants of these ridges. In at least two areas, the ridges have included as their important element Eriophorum vaginatum with Myrica Gale, Vaccinium uliginosum and Spiraea Beauverdiana. One such area is illustrated in the top diagram in Figure 19. The two diagrams in Figure 19 are included as examples taken from field notes of how the vegetation types just discussed fit into a Strangmoor pattern. The notes were taken around the lake west of McGrath. The upper diagram is part of a long continuous bog nearly one mile long crossed by bog ridges 100—125 yards long and 15—25 yards apart. The bog ridges are 5 yards across, 2 feet high and support a few Black Spruce and Larch. This area is just at the top of Map A, Figure 16. The wet centers are Carex limosa, Carex paludivagans, and Menyan- thes, often with scattered individuals of Eriophorum angustifolium, Carex livida and Myrica Gale. The wet meadows in these areas consist of Sphagnum Dusenii (or S. balticum or S. pulchrum), and Carex limosa with the following secondary species: Calliergon sp., Sphagnum compactum, Eriophorum angustifolium, Scirpus caespitosus, Carex pauciflora, Carex chordorrhiza, Carex rotundata, Carex rostrata, Scheuch- zeria palustris, Myrica Gale, and Drosera anglica. The bog ridges have a sphagnum matrix of Sphagnum papillosum or, in part of the area illu- strated, Sphagnum balticum and S. fi: bushes are Myrica Gale, Spiraea Beauverdiana, and Vaccinium uliginosum. There are the fol- SARE ASSES PN TEENS BOE TC EE OL NE DA AO RE oR ORR ao ON OT RR SE a EEE RA SERPENT Seen A rs Don Ea nN AER eRe Ee a EE SE a a AIS NT SES ne Ee NC TRS RN RT N Te a a LONE TRA Ea ONS CN SAME MS DOT RNA eM ee NN STN ES A Te Rae ah Pe oe Tae BOG FLATS AND PHYSIOGRAPHIC PROCESSES IN ALASKA 49 19 Figure 19. VEGETATION MAPS OF BOG RIDGE AREAS. The upper map repfe- Ma Figure 16. This area ts ‘0 each other. p B, at the at a minimum and many bog ada 2 formed at various angles t 50 WILLIAM H. DRURY, JR. lowing secondary species: Awlacomnium palustre, Sphagnum robustum, Sphagnum recurvum, Picea mariana, Larix laricina, Eriophorum vagi- c daphne calyculata, and Vaccinium Vitis-Idaea. In some wet 5 ts, the ridges are built of Sphagnum fuscum, Eriophorum vaginatum and An- dromeda Polifolia, and locally Ledum decumbens and Vaccinium Oxy- coccos are secondary ty species. The lower diagram in Figure 19 was taken from a net-pattern southeast of the lake west of McGrath, in the top left-hand corner of Map B, Figure 16. a) Wettest centers (Rimpi or Braunmoor of Auer; have the elements of both mud bottom and carpet communities of Sjors). Eriophorum angustifolium Carex limosa Carex Itvida € most widespread vegetation of the centers also occurs as a concentric ring around the wet holes and inside the ridges: (Wetsmoor or sedge meadow, a lawn community). Sphagnum cn i Sphagnum éabiggs Carex rotun c) The dae are here expressed as stringers of the Black Spruce forest, (Bruchmoor or hummock community). Ledum decumbens 2 — calyculata Picea mari. d) Wer ‘Serger: -Myrica area which in 064 places replaces b) type (included in Weissmoor or lawn communities Sphagnum eae in clumps — ces pitas Mprica Gale e) Wet sedge meadow: occupies the margins of a) and may replace Sphagnum balticum phagnum Dusenii Carex limosa In a considerable part of the area the pattern lacks the a) Pe. The of Strangmoor areas toward the forest and away fro : water tend to lack it while those nearer to the lake or open water ay it. Long avenues of sedge meadows exist near the margin of the BOG FLATS AND PHYSIOGRAPHIC PROCESSES IN ALASKA 51 > S RRR S *. > > SS ROR Oe: Y - SR SS me RRR OR S SS neva Psa) betes Se Os ee s tof, S S S Po bore eae S oe PROS COS ke» Ne KD Set CKO 6, Os FiGuRE 20 DIAGRAMS OF BOG AND POND ZONATIONS. The left column indicates the process of bog expansion by thawing. While slump is going he bz ildi and i . The right column if S 25a.) : with silt. Symbols for living mosses, peat. sediments and frozen ground - i + S are tho same as in Figure 52 WILLIAM H, DRURY, JR. Strangmoor across which a small pattern of ridges has developed at right angles to the long axis of the meadow. One of these may be seen at the center of the top of Map B, Figure 16. The ridges of the large pattern are about one yard across, 6 inches high and 25 yards apart. The meadow is Carex rotundata, Sphagnum balticum and S. papillosum. The transverse ridges are Sphagnum fuscum, Betula nana, Rubus Chamaemorus, Andromeda Polifolia and Vaccinium uliginosum, with lesser amounts of Ledum decumbens and Chamaedaphne calyculata. Raised peat ridges forming beaches lining open water Wind-driven waves pounding against the edge of the pliable bog carpet, or the drifting ice in spring (after the thaw has freed the main mass from the shore) raise beach ridges. In some cases, these ridges are large and support Black Spruce trees; in others they are most noticeable in the change of vegetation and the sharp drop of the shore at their edge. Their vegetation is, as would be expected, a telescoped transect of the wet-dry sequence. The ridge often has a few Black Spruce or Larch trees, alder bushes and the vegetation of a bog ridge. The margins show Sphagnum Lindbergii, Sphagnum Dusenii, Equisetum fluviatile, Carex limosa, Carex chordorrhiza, Myrica Gale, and Menyanthes trifoliata; that is, the plants of wettest environments. Vegetation of Areas Under the Combined Influence of Silt and Peat Bogs of this type appear closely correlated with Phase III of the floodplain. Examples are shown in Figure 16. It is suggested that vegetation originally became established on the shelving silt shores of a pond or lake. Later the pond became invaded by peat bog forming vegetation which clogged the outlets and impeded the drainage system. Such lakes show vegetation largely of peat bogs, but also traces of vegetation which became established on a silt shoreline. The influence of the minerals in the silt is still effective because the peat-forming vegetation is so thin. The amount of silt shore vegetation in the bog is a measure of the conditions at the time of invasion by bog flora or the shallowness of the bog. Because over most of its area this type of bog results from the occupation of relatively deep water, the a eee ete BOG FLATS AND PHYSIOGRAPHIC PROCESSES IN ALASKA 53 carpet of mosses, sedges and grasses is usually very loosely woven and yields to great depths. The vegetation of these areas is divisible into three headings: ox-bow bogs, pot holes, and Aspen-lined pools. Ox-bow bogs show best the characteristics of the rich flora of disturbed areas and are vegetated by large stands of relatively simple types. These are indeed weedy in the real sense of the word. Where these bogs are found close under the terrace escarp- ments, there are often a number of islands formed by slumping from the terraces. These are vegetated by Pleurozium Schreberi, Calamagrostis canadensis and Betula papyrifera var. kenaica. There is often strong indication that the edges of these bogs are invading the floodplain forest and killing off the trees by the swamping process. Pot-holes and the bogs that occupy them are circular, oval or oblong areas that have developed in the midst of the forest of Black Spruce in natural hollows or as a result of thawing. Initially they are not connected with the main channels of the “flats”, although many evidently have been subsequently invaded by the advancing- margins of the bogs. The flora of pot-holes is characteristically that of an invaded ox-bow or bar lake with emphasis on Sphagnum obtusum, Sphagnum subsecundum, Sphag- num squarrosum, Calamagrostis canadensis and Carex physocar pa- The term “Aspen-lined pools” is used for the depressions Occupied by a pond more or less overgrown with vegetation and surrounded by a narrow ring of Aspen trees on the beach ramparts. Such groves and pools are common features of the Black Spruce forests between the dunes or erosional remnants on the terraces south of McGrath. The ground under the Aspen groves has no Permafrost to at least 4 feet, which is also true of the bogs that have invaded the ponds. The surrounding Black Spruce muskeg, wever, may be frozen at a depth of 18 inches in August. Under present conditions these Aspen-lined pools are largely being drained by regional lowering of the water table, and for is reason most have the familiar silt-shore zonation: Equisetum fluviatile — Carex rostrata — Glyceria, Calamagrostis canadensis, and Dwarf Birch — heaths. Some, on the other hand, have become Involved in drainage channels, since they are largely features of small valleys and interdune areas. These channels have become choked with bog vegetation, and the shore-line zones modified 54 WILLIAM H. DRURY, JR. by the colonization of the peat-forming species. Once peat bog vegetation has invaded a minor channel it often starts to destroy the channel banks and expand laterally, engulfing the surrounding Aspens and then the Black Spruce forest. By this process the flats have invaded and come to occupy much of the interdune areas between Blackwater Creek and Middle (Pitka) Fork. These invasions produce bogs that over large areas are purely of the peat bog type and that in other places are a hybrid of the peat bogs with old ox-bow bogs. These two influences are clearly expressed in bog areas examined at the mouth of the Middle Fork. The sequence of events just described has allowed bog systems to advance up drainage channels, expand on new surfaces, and incorporate into the regional bog systems both floodplain and depositional slope surfaces of quite different physiographic age and history. For this reason, regional bog types discussed in this aper cannot be correlated with physiographic provinces more particularly than the general term alluvial lowlands. COMPARISON OF THE BOG VEGETATION WITH SCANDINAVIAN STUDIES GENERAL COMPARISONS The flora and vegetation of the bogs in the upper Kuskokwim correspond closely to the flora and vegetation of the bogs in the boreal or subarctic parts of Europe around the Baltic and in Russia (Cajander, 1913; Auer, 1920; Katz, 1926; von Post, 1937; Tansley, 1939; Sjors, 1948). The bogs seem to be a combination of the soligenous and ombrogenous mires of the Swedish system (von Post, 1937; Sjdrs, 1948) and correspond to the type of bog found in Europe in northernmost Sweden and Finland, north of the Hochmoors and intermediate types of southern Scandinavia and farther south. Unfortunately, nearly all of the European studies so far published have been made in regions where the vegetation is of a less northern type than that studied in the Upper Kuskokwim River Region. _ : Cajander (1913) separated the bogs of Finland into three major types according to natural plant groupings. The most northerly type, which he called the Aapamoor complex (Finnish, Aapasuo), was then as now the least well-known. The bogs of the Upper BOG FLATS AND PHYSIOGRAPHIC PROCESSES IN ALASKA 55 Kuskokwim River Region correspond most closely to this type. Cajander’s definition of the Aapamoor complex in general terms is translated below. Bracketed insertions are mine. “I. It is a bog type of huge areal extent, whose borders can scarcely be recognized except where they join to the tundra. Bogs anas- tomose widely. “II. There is characteristically a greater development of open bogs rather than forested bogs. The latter are very scarce. “III. The open bogs are full of wet areas, rimpis, which often appear in more or less regular patterns separated or outlined by sinuous ridges, strangs. “IV. Only exceptionally is there a high central area from which water flows outward [as it does in a Hochmoor type}. Water flow is early always unidirectional and can therefore have great force. [Equivalent to soligencus mire of Sjors, 1948}. are found in bog complexes of other types, but the open, most co n areas, the proper Aapas, seem stable. Variations in development [toward firm vegetation mat} and retrogression {towards hollows] are common enough in the centers. “VI. The bogs correspond in distribution to the area of a thin forest found on peaty, partly mineral soil (Rédseikét or Vestkankaat). Cajander remarks that bogs of a similar sort, characterized chiefly by the conspicuous development of bog hollows (he calls them Rimpis) are circumpolar. He mentions Iceland and the Faroes. Von Post (1937) classified Swedish bogs on the basis of the Physiological and peat-forming processes involved in their formation. This classification does not include the Aapasuo type of bog, and I have found it less useful than the classification Proposed by Cajander (1913). In the most recent studies (Sjors, 1948), the vegetation in the Swedish bogs is classified on the basis of “indicator species”. Indicator species show marked physio- logical requirements, such as pH and mineral content, and they are used to label vegetation associations in the Scandinavian- Montpellier systems of plant sociology. According to this system, for instance, all species of Carex (except C. limosa), Eriophorum angustifolium, and Menyanthes are indicators of non-peat-bog 56 WILLIAM H, DRURY, JR. influence or the presence of mineral soil water. Furthermore, when the Scandinavians refer to a community rich in species, such as Sjors Alliance Scorpidion, they mean rich in indicator species without regard for the total number of species present. Most Scandinavian bogs are found in recently glaciated regions or on fresh glacial deposits derived largely from acidic or crystal- line bedrock. The Scandinavians are not concerned with bogs that are extensive on fine-grained alluvium, like the Alaskan bogs. In addition, the Alaskan bogs have formed on alluvium which has been derived in large part from limestone or limey bedrock. For this reason, classification of Alaskan bogs according to the European system — topographic bogs (topogenous), rainfall] bogs (ombrogenous), and soil water bogs (soligenous) — is difficult. The bogs of the Upper Kuskokwim River Region rarely show pure “moss bog” structure, presumably as a result of four factors: the lime content of the underlying silt deposits, the temperature, the amount of rainfall, and lack of any real topo- graphic bogs. All of the bogs are under the control of through- soil movement or percolation (soligenous), and do not show true Hochmoor structures. For these reasons, sedgy areas are more conspicuous in Alaska. These are close to the fens of the Scandinavians (Sj6rs, 1948), but not the fens of the English (Tansley, 1939). Within the bog complexes Cajander (1913) recognized bog types as Wezssmoor, Braunmoor, Reisermoor, and Bruchmoor. The bog types of the Aapamoor complex are chiefly of the Weissmoor type with some development of Reisermoor. This means that the bogs are mostly sedge meadows with some heath- covered areas. It is hard to understand exactly what Cajander’s bog types mean without actual experience in the field. Certain Alaskan vegetation types considered in this paper have been found in all three major physiological divisions of Cajander’s system (Weissmoor, Braunmoor, and Reisermoor). Sjérs divides bog vegetation into communities: mud-bottomed, carpet, lawn and hummock. According to his system the Kuskokwim bogs are chiefly lawn communities with some mud-bottomed and carpet communities in wet areas and some hummock communities in areas. It is impossible to resolve a number of inconsistencies that appear, but these are probably the result of different bases of REET SS oe Tat Reh Pe ERIE: SOUS SY DSRNA Eo IO ear MD PRAMS DS NSE I Pe ta A ea aoe MR ea bs aalasalaesee BOG FLATS AND PHYSIOGRAPHIC PROCESSES IN ALASKA a7 classification. A comparison of plant communities in the field in both areas would contribute much to a consideration of the real nature of plant associations. Furthermore, the physiology of individual plant species probably varies between Scandinavia and Alaska. Different amplitudes produce different areas of overlap in optimal requirements and thus different species combine to form new associations; for instance, the combination of Scirpus cespitosus-Sphagnum papillosum (fen communities, mire ex- panses, and lawn communities) is unusual in the Upper Kuskok- wim (although it is found with Carex rotundata, C. limosa and Myrica Gale in meadows in Strangmoor), but forms an important part of the fens described by Sjors. There are extensive areas of mud-bottomed Scirpus cespitosus-M yrica Gale communities in the Upper Kuskokwim Region which Sjérs does not even mention. The real problem probably lies with the expectation that associations will be identical in these two areas. This subject will | be discussed later, but it is sufficient to say here that this Plant sociological basis disappear when it is regarded from an individualistic point of view (Gleason, 1926). On the other hand the overall characteristics of the bog com- Plexes in Scandinavia and Alaska are identical, and many of | the species are identical. Cajander lists 84 conspicuous species that Stow in bogs in Finland, 60 of which are found in the Upper Kuskokwim. In addition, 8 of the 16 species found in forested _Tegions in Scandinavia are also found in forested regions in the Pper Kuskokwim. These 100 Scandinavian species are the |Plimary species, the most conspicuous in the society. Similarities een the two floras diminish when secondary species are considered. A comparison of secondary species between three of _ (Cajander’s bog types (Weissmoor, Braunmoor, and Reisermoor) _ and Alaskan bogs shows 121 identical species and 164 dissimilar. _ However, I have been unable to find an Alaskan equivalent of raunmoor (if I understand Braunmoor correctly), and if these Species are eliminated from consideration there remain 109 a identical species and 124 dissimilar. The Bruchmoor, the most 58 WILLIAM H. DRURY, JR. mesophytic bog type, shows an additional 34 identical species and 102 dissimilar species, emphasizing the point that as one gets away from the characteristic conditions of the open bogs the similarities diminish. This indicates very strongly that we are dealing with the same phenomenon in two areas. The surface expressions and the primary species are the same. When we consider the “specialties”, less successful in the widespread vegetation — the plants that require more mesophytic or special calcareous conditions — the inter-continental differences appear. The primary species in the bog are highly successful in Finland as in Alaska; they are the circumboreal species, and this is anothec illustration of the familiar experience that the great mass of boreal vegetation (those species with large numbers of individuals) is remarkably uniform. SPECIFIC CORRELATIONS OF VEGETATION TYPES The emergent vegetation, sedge and grass, on the shelving silt beaches of ox-bow lakes, corresponds to fen communities as Tansley uses the term and to those fen communities of the Scandinavians which are found on lake and river shores. e vegetation of peat areas as classified in the field is in many ways out of step with the Scandinavian classifications but the evidence of similarity of major types is clear to anyone familiar with bog vegetation. Simple Stands Areas of relatively consistent, simple stands consist of sedge- meadows, low brush and high brush. a) Low sedges in a meadow-like area are fens of the Scan- dinavians or Weissmoor. 1) The wettest areas (coarse sedges) resemble British fen vegetation in the presence of coarse sedges and are similar or identical to the /agg. The lagg of Scandinavian bogs is occupied by a fen community of coarse sedges which show strong mineral soil influence. These correspond to the vegetation of the moarts. Sjérs (1948) describes fen carpet communities similar to those that have a wide distribution on margins of regional bogs. He describes them as topogenous (such as bar lakes) and limnogenous (edge of lake) fen sites subject to repeated inundation. The edges of bogs in the Upper Kuskokwim River Region which are next BOG FLATS AND PHYSIOGRAPHIC PROCESSES IN ALASKA 59 to stream courses are occupied by a similar vegetation, and on the edges of lakes this tall sedge type may include Sphagnum papillosum (site near Lone Mountain) thereby approaching the Scandinavian type even more closely. 2) The next wettest sites, (typical bog hollows) have mud bottoms, scattered emergent aquatics and sedges. They are referable to Cajander’s Rimpis, or are close to his Braunmoor. Perhaps only the Scirpus cespitosus-Myrica Gale and the Scorpidium scorpioides-Carex livida types should be classified as Braunmoor. Even the last can be correlated with samples Cajander included in Weissmoor and Sjérs in fens (Swedish Kdarrsambdllen, German Flachmoor gesellschaften). Braunmoor is characterized by the presence of Dre panocladus-Scorpidium, and no Sphagnum; it is calcicolous. It may be that in the Upper Kuskokwim River Region e only places where the calcicolous tendencies in the alluvium derived from limey bedrock show through are in those bogs that are nearly free of mosses because any peat concentration creates acidic conditions. These wet mud-bottomed sites correspond, according to the details of their vegetation, to Sjérs’s mud- bottomed communities (/osbottensamballen) if indicator species Utricularia sp., Carex chordorrhiza and Drosera anglica are Present, and to carpet communities (mjukmattsamballen) if indica- tor species Scorpidinum scorpioides, Carex limosa and Carex livida are present and Carex chordorrhiza, Utricularia and Drosera absent. All of the indicator species which Sjérs uses to separate the two communities occur together in Alaska in mostly mud-bottomed areas in the wettest parts of the bogs. These places are usually old lakes, and there are no ridges developed. 3) Sedge meadows seem to correspond to Weissmoor of Cajander and some of the fens of the Scandinavian literature, but not the fens of the English writers. e closest to this in Sjérs’s classification is the lawn community (fastmattsamhallen) with indicator species Myrica Gale, Erio- Phorum vaginatum, Sphagnum balticum and S. papillosum. When Sedge meadows are combined with areas of low brush as they are for example in some parts of the bogs west of McGrath they correspond closely to Weissmoor and lawn communities. b) Low heaths and Dwarf Birch in Sphagnum balticum or S. papillosum are not separated as communities as such by the 60 WILLIAM H. DRURY, JR. Scandinavians. They seem to correspond to Weissmoor and lawn communities. c) Dense brush up to four feet high growing in sphagnum on an open bog corresponds closely to the Reisermoor of Cajander’s classification. Where Black Spruce colonizes, the brush growth is usually looser and corresponds to the Scandinavian Bruchmoor. These both seem to correspond to Sjors’ s hummock communities (Ristuvsamballen) with indicator species Vaccinum spp., Em- petrum sp., Betula nana and Sphagnum fuscum The tall heath communities with a dense Sphagntin carpet relate those bog areas to ombrogenous mires. Areas of brush which are mud- bottomed and vegetated with Scirpus cespitosus-Myrica Gale, giving the impression in the field that they are subject to repeated floods, are equivalent to Scandinavian Uberschwemmungsmoor and oo types are considered as important in soligenous mires by Sjérs. Small open ponds occur in the Alaskan base as ar as in those in Scandinavia (Kélke). Their presence and the processes producing them are sufficiently complex to offer material for a separate study. Patterned Surfaces Patterned surfaces consist of islands of mesophytic vegetation or the bog ridge-bog hollow complex. Vegetationally the islands are identical with bog ridges. ridges are called Strange in German, Strangar in Swedish, and Pounut in Finnish; Auer refers to the word Revel without indicating source. The ridges can usually be put into one of two categories: a) low brush associations which correspond to lawn communities of Sjérs and in part to Weissmoor of Cajander; b) high brush which corresponds to Sj6rs’s hummock communities or to Reisermoor. If there are scattered trees they correspond to Bruchmoor. The wet centers or bog hollows are called Flarken in German, Flarke in Swedish, Kulju or Rimpi in Finnish; again Auer, with- out source, refers to Dykarr. The hollows in their wettest portions correspond to mud-bottomed communities and carpet communities of Sjérs and Braunmoor of Cajander. The sedge meadow centers and low brush margins correspond to carpet or lawn communities in Sj6rs’s system or Weissmoor. The low brush areas which are common and often a conspicuous BOG FLATS AND PHYSIOGRAPHIC PROCESSES IN ALASKA 61 part of the pattern between wet centers and ridges in Alaskan bogs are not reported to be of such general distribution in the Scandinavian Strangmoor. The absence of this community as a conspicuous part of Strangmoor vegetation may account for the ambiguous position of Weissmoor and lawn communities. The sedge meadows and low brush types seem to occur together as one community in Scandinavia instead of in two parts, as in Alaska. Examples of these communities as they are found in the field are included with their equivalent European names in the description of the vegetation under Strangmoor. The combined influence of silt and peat The large areas of the loosely compacted sedge meadows under €se mixed conditions are equivalent in composition and aspect to Weissmoor or lawn communities. The tall brush areas corres- pond to Reisermoor or hummock communities. But mud-bottomed communities and carpet communities rich in species that Sjors calls Alliance Scorpidion both occur. The Black Spruce forest of alluvial lowlands and depositional slopes corresponds in form to what the Finnish writers (Auer, 1920; Cajander, 1913) refer to as Bruchmoor, Reiser-Féhrenmoor, or especially “Anmoorig Wald”. t seems to correspond in form also to what is called Raaseikot or Vesikankaat in Finnish. Bruchmoor is a swamp forest. Reiser- Foébrenmoor is an open stand of stunted Pines (Pinus silvestris) on a heath bog, which probably corresponds to our word muskeg. Anmoorig Wald is a special term that was used for the forests in the region of Aapamoor (Aapasuo), “a thin forest on partly Peaty and partly mineral substrate” (Auer, 1920), probably Corresponding to what we would call dry muskeg or park-like Black Spruce. “Muskeg” in technical use (Radforth, 1952), ugh not in common use, best corresponds to the European term, peat lands (Tansley, 1939). Of course, the important line of dif ferentiation is that the European forest does not contain Black Pruce which is the single plant species that characterizes the Present forest and separates it from other vegetation types su aS wet tundra, or high brush bog types. ANALYSIS OF VEGETATION PATTERNS OF THE BOG SURFACES On the bogs are orderly patterns of patches and ridges of more Mesophytic vegetation superimposed on a general meadow-like 62 WILLIAM H. DRURY, JR. cover of sedges and low heaths growing in sphagnum. The regular patterns take two forms: 1) the clumps of Black Spruce almost as regularly spaced and ordered as trees in an orchard (Figures 14 and 21), 2) the bog ridges, often lined with Black Spruce trees, which occur as festoons or nets. These bog ridges are the strings of the Strangmoor of European authors (Figures 14, 15 and 18). The terms “Streifensumpfe’, “Streifensumpftaiga”, “Polygonsumpfe”, and “Polygonsumpftaiga” have also been used for particular features similar to these. he explanations of these vegetational phenomena are included in the effects of frost action broadly conceived. As Troll (1944, page 639) says in his review of frost action and structural soils: When you pass south from the general continental tundra of north Europe into the region of needle-leaved forests, the frost forms on bog surfaces change similarly s Auer has pointed out for Lapland, palsen [ice-centered peat ummocks, WHD] me lower and oriented into groups and at the same t they transitions to peat waves that take on the net-like, sinuous patterned appearance of the bogs in southern Lapland. Strangmoor, however, are really only a special form of the more inclusive bog type called Aapamoor bog surface is horizontal or only very slightly sloping, then the hummocks are more or less isodiametric and scattered throughout the marshy, grassy bog like little sslands. On the other hand if the the mounds become stretched out lengthwise into cords with their long axes perpendicular to the direction of slope like the solifluction ramparts...... is My observations at tree-line in Eastern and Central Canada and in Alaska agree with Troll’s conclusion that Strangmoor are a phenomenon of the region just south of continuous permafrost. At least their distribution coincides with the northern edge of the boreal forest, which, within 200 mile limits, corresponds with the limit of continuous permafrost. The vegetation patterns of the Upper Kuskokwim River Region can probably be explained in the terms Troll suggests. MEANING OF THE PRESENCE OF SPHAGNUM That the growth of the bogs and of sphagnum is ultimately dependent upon conditions of climate in the broad sense is clearly indicated by the fact that Sphagnum is less abundant in specics and individuals in the tundra regions of Alaska and Canada than it is in forested areas of these regions. The number of arctic Sphagnum species recorded by Steere in Polunin (1947, p. 377— 382) is about half of the number known from the Kuskokwim _ BOG FLATS AND PHYSIOGRAPHIC PROCESSES IN ALASKA 63 Valley. J. B. Tyrell (1910) used the absence of Sphagnum bogs in the tundra in the regions east of Athabaska and Great Slave Lakes as evidence that forests had never grown there, since these bogs according to him are limited to the forested regions. Middendorf (1864, p. 737) pointed out that Sphagnum is notice- ably less common on the tundra of Siberia than in the forested regions, and as he says, it is not ‘“‘at home” there; “climate relations in the far north are not suitable for peat formation”. In the Upper Kuskokwim River Region, however, Sphagnum moss carpets about 75% of the land surface not including the bogs. “The peat derived from the heavy growth of Sphagnum mosses over all the swampy lands may be considered the most characteristic and widespread of recent deposits” as Maddren (1910, p. 61) says in referring to this region. My observations from the air near Fort Churchill, Manitoba, north of Fort Chimo on the Ungava Peninsula of Quebec, and in Alaska all indicate that Strangmoor are a phenomenon very closely associated with tree line. Where I have seen them they extend only a very short distance onto the tundra beyond tree line, and they extend perhaps only 200 miles south into the forest. The distribution that Troll (1944) reports from northern Europe and Asia agrees with this. The reason is probably that the con- ditions suitable for the luxuriant growth of sphagnum occur in that area and, as will be discussed later, that the same conditions Produce characteristic patterns of frost activity responsible for the ridge pattern. The relative scarcity of Sphagnum mosses in the “flats” of the Yukon and Tanana rivers requires some explanation if these statements are correct. One reason is correlated with the fact that, as Katz (1926) and others have pointed out, fire destroys the identity of a peat bog. A marsh of sedges and grasses (a fen in British and Scandinavian literature) usually takes its place. W. S. Benninghoff has frequently told me of the conspicuous evidence of repeated fires in the flats in the Yukon Flats area and has remarked on the almost complete absence of evidence of fire in the Kuskokwim River Region. In the few areas of the bogs in the Upper Kuskokwim River Region that have been burned, there is a conspicuous increase of 8tasses and sedges and decrease of sphagnum. A second reason is concerned with the percolation of soil water. 64 WILLIAM H, DRURY, JR. In the Scandinavian classification of bogs (von Post, 1937; Sjérs, 1948), those bogs which derive their moisture from water per- colating through the soil (soligenous mires) are characterized by marked decrease of Sphagnum and increase of grasses and sedges. Fully developed peat bogs, according to the Scandinavians, depend largely on water falling as rain (ombrogenous mires) and not that derived from the water table because of topography (topo- genous mires) or movement through the soil (soligenous mires). It is my understanding from W. S. Benninghoff and J. R. Williams, who have studied that area over several years, that the sediments of the still-braided rivers, Yukon and Tanana, in the region where the flats are found, are coarser than those of the Upper Kuskokwim. Perhaps a combination of both the coarser sediments allowing more water percolation and the widespread occurrence of fire has prevented the full development of peat bogs and has contributed to transformation of them into fens over wide areas where peat bogs may have occurred in the past. STRANGMOOR The most conspicuous vegetation pattern is that of the serpent- like moss ridges. These are evidently a response to down-slope forces. Their vegetation has already been described in detail. Auer (1920) has presented an exhaustive study of Strangmoor in Finland, and nearly all the references to theories of origin are taken from his publication. Troll (1944) refers to it as the most satisfactory and up-to-date explanation of bog ridge formation. I have translated his summary of Auer’s discussion of the various theories on ridge development as follows: “V. Auer en has sea a precise apie a into the nature of Strangmoor and their meaning in earth climates, that brings together all earlier- pai sented theories. Sang paar climatic conditions va frozen ground on gently s!oping bog su in boreal climate (needle-leaved forest). To menti briefly. Fellman ( ob ie and Norrlin (1873) find ridges (Stramgs) similar to Palsen. Hjelt and Hult (1885) recognized that linked to sloping bog surfaces. The essential process of the ge nesis was seen by G. Anderson and Hesselman (1907) 1n the en flow of the peat, especially in spring through the strong action of meltwater over the still frozen substrate. The formation of bog ridges is accomplished ‘wholly analogous to the formation of solifluction terraces in arctic regions that ce alr been interpreted at that time. Tanttu (1915) further recognized the m 2 of the differentiated — ice structure. “During the process of tt e autumn, ice first aj in the bog BOG FLATS AND PHYSIOGRAPHIC PROCESSES IN ALASKA 65 hollows (Flarken), then later in the ridges due to the poorer heat conductivity dry hummock peat. Thereby, at one time, a squeezing of the peat occurs in the ridge, entirely similar to the case in Palsen. In mid-winter, on the other and, the ridges freeze more solidly, as they have a thinner snow cover and are an readily blown clean, which produces a further vaulting effect. In spring the importance. The hypothesis for Strangmoor is a strong expression of winter ground frost and summer thaw. They are a true frozen ground pattern, but without permafrost in contradistinction to Palsen.” This explanation is presented from a geologist’s point of view. Troll has not considered that in these features there are living plants of different growth rates. He has considered the hummocks and ridges as physical features and has omitted L. von Post’s suggestion that differential growth (hydrophytism and meso- phytism) in wet or dry places, referred to by Auer, is a genetic factor in the formation of these micro-relief features. H. Ranken (referred to in Auer, 1920) suggested that the plant growth expands laterally from hummocks, and emphasized that know- ledge of the formation of the patterns must involve knowledge of the underlying peat structure, i.e. the history of formation of the irregularity. Troll’s emphasis is upon the subsequent development of ridges by frost action exaggerating the differences initially established. Neither he nor anyone else has presented an adequate explanation of the initial establishment of the bog ridge and the intervening bog hollow. Troll says that winter freezing in the hummocks causes additional doming although ice does not persist over the summer. This is hard to consider of more than temporary effect from the Point of view of plant growth, as the doming occurs during the Period of plant dormancy. Troll does not consider the details ot Auer’s theory, and so refers only briefly to the critical feature of persistence of bands of ice leading to the lateral growth of ummocks. Auer’s 1920 paper remains the best discussion of the develop- ment of this geobotanical feature. Although he does not adequa- tely explain the initiation of the pattern, he discusses the existing theories, their merits and disadvantages and contributes his own 66 WILLIAM H. DRURY, JR. discoveries, amalgamating the whole into a paper which should be recognized as a classical contribution to the study of ecology, I have translated his review of the existing theories at the time of his writing as follows: “I) They are the normal result of a struggle between those processes a) constructive and building up the bogs [including Scirpus cespitosus in Finland, WHD)}, and bh) those of regressive nature making hollows “IT) The bog is raising itself above the water level, and bog ridges devel by formation of extensions of hummocks, Both nt ype and bog ridge represent results of progressive processes, and the topography of the basin occupied by the bog has an especial influence over ke pattern of low. “TEE vie features depend in on slipping of the peat. When peat is soaked slips slowly and stops when the ramparts reach a certain height.” This is the most readily concluded explanation and many still subscribe to it although no one has presented adequate evidence of any movement to support it. “IV) Bog ridges are the remains of a Sphagnum fuscum Weissmoor (sedge meadow) which has been torn by expansion and the holes are occupied by bog hollows. The advance or growth of bog hollows is due to climatic or orographic forces Sohju theory — that plant remains and slush in early spring are gathered into lines (windrows) and freeze irregularly; the snow meltwater é _— exaggerates t [This theory would explain the pattern of ridges in relation to slope and islands, and depend on the other theories for their spread and development. It has the disadvantage that at the time Auer published no one had been able to find windrows of size sufficient to offer an adequate start for the ridge. I believe, however, that my observations on Bylot Island, Canadian North- west Territories, in 1954 indicate that this is very close to the actual mechanism for the establishment of the pattern. These observations are discussed later. "¥I) — develop by: a) freezing =_— of rifts in the bog made by water, at the autumn freeze-up, b) freezing in the breaks caused d ice.” ue cake through and upw elling ey ‘water from under th In his paper Auer (1) establishes the critical importance of relative wetness and dryness in ridge development, (2) emphasizes that the kernels of spread are sphagnum tussocks, (3) nowhere emphasizes that ridges are a simple su | hy phytic species of sphagnum on more hydrophytic, although his illustrations point it out, (4) concludes that neither hollow nor ridge can exist without the other, and (5) believes that flooding BOG FLATS AND PHYSIOGRAPHIC PROCESSES IN ALASKA 67 and drying are indirect in their effect on the extension and interconnection of hummocks. ere are many processes and forces involved in the develop- ment of ridge patterns according to Auer. Six processes are emphasized by him. “1) In the north, spring thaw is rapid and gives meltwater great force. The layer above the frozen ground is thin, and becomes lifted ma broken and slip. The products of this high water and snow debris collect in lines. 22 records isolated data on actual ridge movement with Scirpus caespitosus tussocks, and suggests that the distortion of ridges is strongest of al! during fall flood before the ridges are frozen in. In spring ridges are still anchored so that they are compressed and elevated by the pressures rather than being moved.”’ [There seems to be no evidence of any extensive fall floods in the Alaskan bogs. Observations indicate, on the contrary, that there is a progressive drying from July to October. ] 3) Cases in which ridges result from the actual overthrust of peat due to movement of the whole bog seem to be restricted to the borders of large lakes. g. ize, number and spacing of bog ridges does not seem to reflect the angle of slope. On nearly level bogs, there is correlation of the topography of the bottom of the bog basin with ridges, but this does not seem to hold if there is any slope. “5) Winter freezing pushes against the bog ridges from the bog hollow; Persistent frost in he mdges protects them from spring : “6) In spring, water percolating across the bog is warmest after passing Over a hollow and coldest just after emerging from a frozen ridge. In fall the Water is coldest after crossing the hollow. Therefore ice forms near the ridge squeezes it.” Auer establishes, as field observations indicate, that because there is no surface displacement of peat strata, there is no horizontal movement of peat. However, his chief contribution is that he was able to demonstrate that in bogs at high latitudes the last remnants of ground ice persist under a hummock in strips elongate at right angles to the direction of slope, forming a Peninsula to the sides of a hummock. Strips of ice similar to these can be found in spring in temperate regions when the ice melts from ponds or especially shallow pools in swampy places. They are of short duration and soon blow to the shore. Auer goes on to say that drifting slush and snow debris, torn up plant remains, and silt borne on spring floods will collect against areas raised by the persistence of ground ice. These will be elongate across the slope. Such debris contributes to the dryness 68 WILLIAM H. DRURY, JR. locally by formation of a tide-line-like collection, and encourages colonization by less hydrophytic mosses. Any slight advantage ridge. Once started, such a process will accelerate by its own effects, The initial elevation allows the formation of hummocks which grow by lateral extension. Hummocks coalesce and act as further dams for slush and floating debris. At this stage the element of frost-push from winter ice in the hollows enters, narrowing and elevating the ridge and preventing its expansion downslope. Because in the autumn the water is coldest after crossing a hollow, ice forms downslope first and then water movement pushes to the sides. Mid-winter ice-push thrusts hardest along the long axis. These forces tend to elongate hollows across the slope too. Auer allows that freezing in the hummocks is important at a late stage, but disagrees with Helaakoski and Tanttu who considered it of fundamental significance in the initial establishment of the pattern. im my excavations ridges are formed by the simple superposition of one sphagnum species upon another, Sphagnum balticum or S. recurvum growing on S. Dusenii; S. papillosum growing upon S. Dusenii, S$. balticum or S. recurvum; S. fuscum growing on 16) Tansley (1939) shows (after Osvald, 1923) the same three stages of sphagnum sequence from w occur in Eur that occur in Alaska. In Britain S. cxspidainre is followed by S. papillosam which is overgrown by S. rubellum. Sphagnum cuspidatum is a more southern representative of the Cuspidatae group of Sphagna (S. Lindbergii, S$. Dusenii, S. balticum, S. recurvum), to which the sphagna wet areas in ska belong. Sjé comments that Sphugnurz Dusenii and §. Lindbergii belong to more northern bog types than those Scandinavian bogs he ied. Sphagnum papillosum seems to occupy the ition i of the bogs of Scandinavia and Alaska ve ion but §. fuscum in Alaska fills the sites on tops of the hi occu that site in Great Britain. §. rubellum in Alaska forms a deep carpet on damp, v ther despread on gy Britain than S. rabellum, and is more important in hummock formation in dinavia than S. rubellunz. BOG FLATS AND PHYSIOGRAPHIC PROCESSES IN ALASKA 69 As a specific example, a vertical section was dug across the right- hand ridge at the top of the lower diagram Figure 19. The sides of the ridges are steep on the sides away from the lake, and gentle on the side toward the lake. The section showed 1) Sphagnum fuscum and its dead bases to a depth of 10 inches, 2) a 7 inch layer of compacted Sphagnum papillosum; 3) a 15 inch layer of compacted Sphagnum balticum. The water table was at a depth of 15 inches, near the base of the Sphagnum papillosum layer. Tussocks of Eriophorum vagina- tum were rooted at 10 inches, right at the top of the S phagnum papillosum layer, which seems to be the rule for these tussocks. There is no indication of slipping or horizontal movement of peat or surface displacement of strata, although bog ridges are consistently higher and steeper on their downslope margin than on their upslope edge. The horizontal bedding of the peat layers precludes any formation by frost heaving such as Sigafoos (1951) has suggested for peat ridges on Seward Peninsula, Alaska. Wenner (1947), however, who seems to be the only person who as dug a cross-section through a ridge to the bottom of the bog, reports that the ridge and hollow retain their separate identities all the way to the bottom without the alternation of hummoc and hollow positions and functions that Osvald (1923) regards as fundamental to bog growth, and without superposition of the progressively more mesophytic species of sphagnum just discussed. As Osvald described it, in wet places at the base of a hummock grows S. cuspidatum which is replaced by S. papillosum. Le papillosum is overgrown by S. rubellum or S. fuscum. The hollow ween hummocks so formed is occupied by Sphagnum cuspida- tum. In time, another hummock forms in what was the hollow, Overtopping the first hummock. In this way an area is first hollow, then hummock, and its position alternates as the surface of the bog is raised by growth of peat. Such a process in action in the Strangmoor would suggest that ridges would become over- topped by growth of hummocks in the neighboring hollow. No excavations were made specifically to test such an idea, but those excavations that were made gave no indication of the possibility its action. To return from a discussion of the literature on the subject to the process of formation, bog ridges are developed on the surface 70 WILLIAM H, DRURY, JR. of a more or less swampy bog as the surface gradually becomes drier. As the bog water-table is lowered the uneven bog surface allows colonization of mesophytic mosses first on those localities that are slightly raised above the water-table. Hummocks appear and, in some cases involving Scirpus cespitosus, it has been suggested that these are bodily rearranged during spring floods. The evident expression of the hummock development is long lines of bog ridges at right angles to the slope. Collection of water where drainage is impeded by these bog ridges has been said to swamp the surface growth and thus form bog hollows. Auer’s explanations are v complete, but they do not adequately explain the initial establishment of the pattern and the fact that bog ridges resemble waves or ripple marks in form (Figures 14 and 15). They bend upslope near islands of forest and belly downslope in the open slopes between such relatively more consolidated places. In this they resemble the pattern of waves that are slowed up at the margins of a stream and are carried farther down where the water flows fastest in the center. Auer’s explanation does not account for the ridge 100 yards long and 3 yards wide. That such a ridge is formed by the aggregation and lateral expansion of hummocks seems to ask too much of coincidence. Ridges should fork and form clumps if there were not initial and additional forces. In other words, the first establishment of the pattern of bog ridges or Strangs is still not explained. Two alternative hypo- theses are offered for this. First, in a pliable medium such as a sedge meadow carpet in loose-growing sphagna, downslope sagging of the entire sodden surface will occur. I have seen waves formed by such sagging in the channel of a small stream choked with sphagnum. This sagging need not produce any convolutions or distortions of strata of peat. Movement is slight at best, as is shown by the minor surface expression, but has been observed to produce waves of slight elevation which were sufficient to allow the top to dry out. On these, less hydrophytic sphagna can colonize. This movement is probably not of sufficient magnitude to create the bog ridges by itself and it seems to occur only when the whole sedge carpet is completely thawed. It may or may not be of importance in the initiation of the ridge pattern, and clearly is not of the importance of the second mechanism suggested. ; CONTINUING is « pre pré above BOG FLATS AND PHYSIOGRAPHIC PROCESSES IN ALASKA 71 Second, where torrents of water (from sudden thawing of snow or a downpour of rain) run in a sheet in less consolidated mate- rials over a hard surface, riffles are formed and after the water has subsided debris collects on the festoon-like ridges created. This suggestion seems to return to the Sohju theory discarded by Auer, but it actually does not involve the collection of debris and refreezing of it that that theory requires. The pattern of festoons has been seen on the floor of a pine forest after an 11 inch rainfall, and in this case was expressed by step-like ridges of pine needles alternating with levels of the black organic soil below. Most extensively and clearly, however, the pattern has been seen where snow meltwater ran over frozen ground, as was the case in June 1954 on Bylot Island in the Canadian Arctic Islands. The features resulting from these soil processes werc studied in detail on Bylot Island to investigate whether the phenomenon observed there might have some application to the initial establishment of bog ridges. Figures 23 and 24 show that a sheet of water flowing across a surface of frozen ground creates steps and falls, whose steep places occur in festoon patterns. The same effect is well-developed in the beds of small streams in the same area and may be related in origin to the falls and levels of the valleys formed by valley glaciers as described by Matthes (1930). Figure 24 shows, in the foreground, crescentic riffles on which moss and sedges have grown and formed ridges. In the background similar ridges are large enough to dam a body of meltwater and are closely similar to bog ridges and bog hollows". In each pond held up by a peat ridge, the level of the water decreases stepwise downslope the way it does in terraced rice-paddies. In at least three ponds held in by large peat ridges, the pond below the ridge was $8 inches lower than that above. One of these cases is shown in Figure 24. 17) The formation of the larger ridges in this particular case, however, is and modified by the a rance in unconsolidated Pairs of at times as regular straight as if man- 5 peat ridges : £ readjustment cracks and peat ridges will be the subject of another paper. 72 WILLIAM H. DRURY, JR. In nearly all discussions of the formation of ridges and hum- mocks I have found, the chief emphasis has been on the per- sistance of a ridge or kernel of ice. To test this, the depth to frost below the surface of the soil and below the surface of the water was measured by probing from riffle to riffle across 40 such levels. Depth to frozen ground was tested daily during the period June 20—July 4, including that of violent thaw, June 24—29, 1954, and was found to be essentially regular. Below each step or steep part of the riffle, the frost was much closer to the surface (1—1 4 inches on June 28); crossing the pool toward the next riffle the depth increased regularly until it reached maximum on the upper edge of the next riffle (2 inches below the surface of the soil on June 28). The surface of frozen soil was concave upward from the upper edge of the pond to the lower and resembled the impression made by pressing a clam shell into wet sand. The depth to the frost table was similar to that of the water; where the water was deepest, the depth to frozen ground was greatest below the surface of the soil. The close correlation was that where melt water was deepest, the depth to thaw was greatest, and where the water was shallowest and moved fastest the depth to frozen ground was least. The contrast in depth to frost from above the riffle to below the riffle was as much as 1 inch in a distance of 8 inches, and this combined with the pilling of mosses and sedges by the water current to act as a dam. Auer noticed similar phenomena in Strangmoor. He reported that the water is coldest just after emerging from a bog ridge and warmest just before flowing into the next one downslope. In the arctic deposits studied, once a peat ridge was formed the ridge was often wall-like, both upslope and downslope, but repeated tests indicated that the thaw advanced equally rapidly from the surface into the soil in all places. The depth to frozen ground was as great or greater under a ridge as under a hollow. Because of the greater depth of vegetation and turf in a ridge, there was a noticeable rise in the frost table, although at any point in the ridge the frost was usually farther below the turf surface than in the low places between ridges. In such places the frost acted as a real dam holding back the flow of water as Figure 24 shows. But the ice did not persist as a kernel in the hummocks, ES Pe ieee uae ie se Onn ee Peat Re SRN IR ane IS a Belg Nig OR Rg Pee te oc) ke i et ha RE ROPER MAP TN Salen crake Roan ar M yes Sar prea Viethen ee aM eet) ye a te BOG FLATS AND PHYSIOGRAPHIC PROCESSES IN ALASKA 73 and in the riffles the ice persisted just below the step rather than under the terracette bank itself. The total effect of the formation of riffles by differential thawing during melt-water flow across the surface was the creation of hollows where the water collected and ridges (riffles) which dried out relatively early. In plan the riffles extended farthest downstream where the current was strongest and as a result a fan shaped terrace was produced across the stream which closely resembles the festoon type of peat ridge in Alaskan bogs. The terracettes on Bylot Island were seen to vary from 1 to 4 yards across and from 4 to 35 yards long. In addition to the long, festoon ridges, short isolated hummocks or ridges 5—75 feet long were found, again running across the slope. Where correlations were made these coincided with places where microtopography concentrated the flow of melt water while snow still lay in banks over most of the ground. These forms are transitional between Palsen or ice-centered hummocks and Strangs or bog ridges, and show that the essentials of their formation as described by Troll (1944) are correct although some of the details of their initiation are probably different. The feature of main importance to the formation of bog ridges in subarctic regions is that water flowing across frozen ground forms riffles which appear in the form of festoons. The result is a series of steplike levels. At the margin of each level, later develop- ment of moss and sedge turf forms a ridge. Standing water in the ponded areas thaws deeper into the frozen ground during the thaw and exaggerates the ridge and hollow pattern started in the riffles. Similarly when melt-water flows across the Aapa bogs in spring, it flows across a frozen substrate and creates conditions similar to those responsible for the riffles and festoons on Bylot Island. The formation of the riffles not only produces the drier ridge, but also the ponded water in the bog hollow at the same time. All students have agreed that the bog hollows and ridges cannot exist One without the other. This is the first hypothesis which accounts for the origin of the two at the same time and subsequent growth together in a way would find support in Wenner’s excavations in Labrador. The establishment of the pattern is a physical phen- omenon resulting from the flood of water through or over loosely compacted, thawed vegetation resting on a frozen substrate. No lateral movement of vegetation or turf is involved but rather only 74 WILLIAM H. DRURY, JR. the differential effects of erosion and thawing by a meltwater torrent. There is further evidence that patterns of ridges are a response to physical forces not explained here and, in fact, not understood. Patterns which seem to be identical to the nets of Strangmoor have been seen on the surface of brash ice collecting on a slope within range of salt spray (Figure 17). In the case observed, water seeped up through a hole in the ice, thawing the ice already present and freezing itself while it ran across the surface. The alternative that a skim of ice had first formed and later collapsed as the water under it escaped was investigated. Our field studies indicated some breaking and sinking of the ice, but much more conspicuously modification by formation of ice over the ridge without fracturing. This indicates at least that physical forces independent of the vegetation are able to create the pattern, and indeed as suggested before, that the vegetation only capitalizes on and emphasizes microtopographic features which frost action has created. If we make the assumption that formation of riffles as melt- water flows over the frozen bog can actually form ridges of sufficient size to initiate a pattern of bog ridges on a large bog, we would have an adequate explanation for their form and structure. Such an explanation would also account for the increase of ridges and their closeness together at the margins and down- slope ends of sloping bogs in contrast to their wider spacing in the middle. This sort of distribution is found in both raised bogs (Hochmoor) and regional bogs (Aapamoor) (illustrations in this paper and in Troll, 1944). Such a system would explain the transverse small-scale pattern or ridges within tht larger pattern. Most important, it would not contradict but would be a source of initiation of the pattern subsequently acted on by the forces so completely handled by Auer. On the raised bog ridges formed in this way, more mesophytic sphagna can colonize and their growth will allow differential persistence of ice. Persistence of ice under hummocks as Auer described it will maintain the segment raised in this way while the rest is flooded by water as it thaws. The remnants of frozen hummocks are elongate across the slope and will collect the slush and flotsam of spring high waters. Spring floods erode and push the hummocks which develop on such relatively firmer substrates, 4 a 4 se BOG FLATS AND PHYSIOGRAPHIC PROCESSES IN ALASKA 75 ordering them further into lines across slope. Persistence of ice in the hummocks protects them from further movement while they are modified in outline by ice-push which elevates and elongates the bog ridges and lengthens the basin of the bog hollow. Dam action of these bog ridges leads to collection of water in the areas between, swamping the vegetation there and exaggerating the differences between bog ridges and bog hollows. The patterns are phenomena of the general process of frost action, but no completely adequate explanation of their develop- ment and especially first establishment is yet available. Obser- vations during the spring thaw in Alaskan or Scandinavian Aapa bogs can easily establish whether the forces suggested here are actually effective. If such is indicated, then detailed marking and mapping of changes in ridges can indicate whether changes are still going on, where they occur, and whether new ridges are forming on parts of the bog. REGULARLY DISTRIBUTED ISLANDS A second remarkable pattern on the surfaces of the bogs is that of clumps of trees regularly spaced as if planted (Figures 14 and 21). It is poorly expressed in the Upper Kuskokwim River Region but clearly expressed in the Cook Inlet area and along the Alaska Highway near the Yukon border. The clumps of trees may be isolated in a generally treeless bog (Figure 14) or be clumps of taller and healthier trees in a generally poor and thin bog forest (Figure 21). Often there are dead trees in the middle of the clumps. Sometimes such trees are split from the base, and in several cases tears and other disturbances of the moss carpet have been seen. The islands are slightly elevated above the general surface of the bog. Frozen ground has been found as late as August and September underneath them. The clumps are con- sidered to be expressions of factors operating in the climatic transition zone between ice-centered hummocks (Pounn — Auer, 1920; Baidsaraki — Pleske, 1928; Pingo — Porsild, 1938; Thufur and Palsen — Troll, 1944) and the sinuous moss ridges (Strangs). A partial explanation is as follows. When the surface of the quaking bogs has become well vegetated with sphagnum, sedges and brush, ice hummocks develop as Troll (1944, p. 633) indicates. Local persistence of frost elevates the moss, dries it, and allows more aggressive growth or colonization by other species. The WILLIAM H, DRURY, JR. plants act as a protecting insulation and the ice kernel grows, further elevating the hummock. Within the regions of forests such elevations would offer better drainage, aeration, deeper root penetration and more stable substrate for seedbeds for trees. Their position on the elevations favors these trees over those immediately around them allowing them to grow fast and tall. Subsequent frost action and heaving centered around the ice kernel have killed and split some of the trees and torn the moss carpet, but the layering of Black Spruce has maintained a clump, marking the site of the elevation. The ability of lower branches to produce roots when covered with humus, layering, is widespread among the conifers, and these sprouts will live on when the main stem has died. This ability, however, is conspicuously better developed in Black Spruce than in the other conifers of Interior Alaska. For this reason the development of sprout clumps is a phenomenon peculiar to Black Spruce forests in this area. I have seen frost-heaved mounds regularly but widely spaced on nearly flat muskeg surfaces along the Alaska Highway (near the Alaska-Canada border and northwest of Big Delta, Alaska), where the vegetation mat has been removed beside the road. These appeared very similar to the Baidsaraki illustrated in Pleske (1928), but were much more widely spaced. The regularity must reflect factors similar to those that create the regular sizes of tundra polygons (both those due to ice-wedges and those not), and suggests that in saturated silt-organic soil there are regular limits to the spacing of hummocks. Perhaps structural readjust- ment of the whole lowland area due to movement resulting from removal of the ice sheets is responsible. It is obvious that many other forces are concerned with formation of polygons and frost crack patterns and that forces not included in what has been suggested must be involved. A feature entirely unexplained is that the initial colonization of emergent sphagnum in open water is often in clumps of quite regular distribution (Figures 14 and 15). How these form and why they show a regularity of pattern is not understood, but when the whole wet area is overgrown, the initial colonies wili have had the opportunity to become compacted. The compacted sites will be suitable for the initial colonization of trees, and this in its turn will lead to the patterns described. DISCUSSION GEOLOGICAL IMPLICATIONS There are several geological implications of the processes modifying the surface of the floodplain. First, bogs are a serious impediment to human occupation; second, they are a source of confusion of stratigraphic sequences; third, they are a potential source of the mucks; and fourth, they are an instance of soi! instability which has limited or eliminated forests. 1) EFFECT ON MAN’S USE OF THE LOWLANDS Large areas of the alluvial lowlands of the subarctic regions of Alaska are covered with peat bogs. These alluvial lowlands are the most accessible and potentially fertile lands of the region, but bogs have engulfed large sections of them. Such places are entirely unsuitable for settlement. The bogs are so unstable that ey are a serious obstacle to travel. Even in winter, when loosely compacted vegetation and peat occupying the sodden bogs will freeze only to shallow depths because of poor heat conduction, Peat near to the surface may remain unfrozen locally and tractor trains crossing such bog areas have been lost by breaking through the ice and sinking into the mire below. Drainage undertaken on a large scale would be required to make bog areas useable, and it is not proven that drainage would be effective under the Present climatic conditions. 2) EFFECT ON STRATIGRAPHIC SEQUENCES At the advancing margin of a peat bog, where thawing of the banks undermines a forested surface, as much as 10 feet of sediment may be considerably altered by slumping. Also, water Percolating through the soligenous bogs can carry silt and pollen grains into the subsurface levels. Such activities, although at any One period they do not penetrate to great depths, can contort, Feverse and complicate stratigraphic sequences by secondary deposition both in the bog and in the deposit that is being engulfed. They exist as a source of confusion and error in pollen and stratigraphic sequences in any bog which has been subject to the conditions of perennially frozen ground outlined in this paper. ch distortions, because the stirring is limited to certain levels, 78 WILLIAM H. DRURY, JR. will still allow misleadingly reasonable vegetational sequences. As is the case with many other sources of error in the study of llen analysis, this is not so serious as to invalidate results from carefully studied bogs, but must be considered as a real danger in bogs which may have had water flowing through them while they were growing. It is probably not of real danger, however, in either a pot-hole bog (a type of topogenous bog) or a raised bog (Hochmoor or ombrogenous type). 3) THE PRODUCTION OF MUCK The incorporation of a large amount of silt of either alluvial or aeolian origin with finely divided organic material either by repeated bog cycles or deposition by percolating water produces a deposit which fits the general definition of muck (Eardley, 1937). Similar mucks are accumulating at present. Although most of the valley mucks of the interior of Alaska are depositional slope or alluvial deposits, the present phenomenon will explain the existence of mucks in valley centers away from slope movement. 4) ELIMINATION OF FORESTS BY SOIL PROCESSES As the cycles of bogs and forests are conceived here, frost- produced soil conditions are responsible for the elimination of the forests. Some authors (Denny, 1940) have indicated that forest vegetation is the limiting factor in the distribution of frost- produced patterns in the ground, and other authors that structured soils are evidence of treeless conditions. But Benninghoff (1953) has shown that in southern Alaska remarkable frost features arc being produced in the soil today under a forest cover. It may be that, at the southern limit of patterned ground and other expressions of frost action, strongly rooted vegetation can control the expression of frost-movement patterns. In such a case it is an academic question as to whether climate or vegetation is respons- ‘ble. For it is obvious that where frost action is less effective the influence of vegetation is more effective. In the Upper Kuskokwim River Region evidence shows that trees have been uprooted by frost action, that on north slopes where frost activity is widespread trees are scattered, dwarfed and twisted.. Where such slopes are sheltered by rock outcrops, or where slopes are steep enough to allow efficient drainage the trees are healthy. Se aragt star tare Teta nt at ae eee Soop at BOG FLATS AND PHYSIOGRAPHIC PROCESSES IN ALASKA F de All these observations indicate that solifluction phenomena con- trol the distribution of trees on slopes, and that the activity of bogs as a cryoplanation process is another phenomenon in which the physical activity of frost controls the distribution of trees. BOTANICAL IMPLICATIONS There are several botanical implications from these studies which apply to the study of Pleistocene climates and vegetation. First, conditions for plant growth in Interior Alaska at present support the suggestion that the Interior was an area of plant persistence. Second, the conditions of climate, vegetation, frozen ground, and the distribution of forests and glaciers allow speculations on the periglacial climates of the northeastern United States: that mixed coniferous and deciduous forests survived close to the ice in a wetter but not markedly colder climate and that widespread permafrost and tundra are not indicated. Third, the vegetation of the bogs studied cannot be understood as a series Of integrated entitics or associations with interspecific depen- dencies. The Climax as a concept is found to be useless and confusing in contrast to Gleason’s attitude that vegetation must be studied on the basis of the actions of individual plants and Species. 1) AREAS OF PERSISTENCE DURING THE PLEISTOCENE ICE ADVANCES Large areas of forested lowlands exist in a region of cold, wet climate and soil instability in the Upper Kuskokwim River Region. This is supporting evidence for the conclusions drawn from floristic studies that there were large areas of forests in the interior of Alaska during some or several of the ice advances oF the Pleistocene. The climatic conditions suggested for glacial climates of the Continental Ice Sheets by recent investigators (Willet 1954) are similar to those of Alaska today, and to the conditions Péwé has suggested to me for the formation of the Fairbanks mucks. Péwé suggests that the wind-blown silts were deposited on the hills during the colder periods and washed thence into the valleys and incorporated with organic material during Warmer periods. It is necessary to have periods when vegetational ‘Over was very limited for sweeping winds to stir up clouds of loess, and for formation of sand dunes, but a pool of forest Vegetation must have been available close to these areas. The 80 WILLIAM H. DRURY, JR. indications of fossils for the conditions during the more or less continuous formation of mucks require that forests were not far distant. There is no justification offered in the present vegetation for suggestions that the forest was eliminated during the period of loess deposit and dune formation because climatic changes that would have a negligible effect on floodplain or hillside forests would bring about increase in the area of braided stream beds, and drying of — terrace surfaces On the other hand, Hopkins as told me that mucks on Saha Peninsula ea gram tied in a wet climate and with a treeless vegetation similar to the present. In these various places, conditions during muck formation in the past do not seem to be fundamentally different from the conditions today in that particular area. lf the areas that are now in Black Spruce were wet and treeless during the ice advances, the well-drained sites on hillsides and river bottoms could still have sheltered forests which would havz been sources for recolonization during warmer periods. The rapid appearance of forest vegetation after the retreat of the ice and the minor advances in speciation of trees and other forest species of plants and animals in interior Alaska argue strongly for the presence of isolated patches of vegetation like this. Major climatic changes are not needed to explain the formation of mucks in valley centers as described in this — Bogs and bog activities are able to persist either in a period of loess deposit (when freeze-thaw cycles would probably play little part) or in mild periods (when thawing of the frozen lowland sediments would be widespread). ~ 2) SPECULATIONS ON THE NATURE OF PERIGLACIAL VEGETATION N THE EASTERN UNITED STATES Evidence of periglacial climates Frost-moved soils and rounded hilltops in the White Mountains, Adirondacks, Appalachians, and as far south as the Great Smoky Denny (1951) and pao bl (1954) have cokad cate fens in Pennsylvania. ae ee te ae aM Pore ee Are BOG FLATS AND PHYSIOGRAPHIC PROCESSES IN ALASKA 8i A variety of evidence from the Mississippi Valley area seems to indicate intense frost climates there which tie in with essentially proven conclusions of frost climates in Central Europe during the ice advances. Adequate evidence (Davis, 1946; Frey, 1951; Potzger and Tharp, 1943, 1954) indicates the presence of spruce and fir in North Carolina, Florida and Texas, establishing the southward migration of elements of the boreal forest now extending south on the uplands into Virginia, North Carolina, and Tennessee. The present studies in the Upper Kuskokwim River Region, however, indicate that it is dangerous to deduce from these evidences that frozen ground and a broad zone of tundra and south of these a broad zone of boreal forest stretched away from the ice front and that the climate was similar in those zones to the present climate of boreal forest and tundra regions (Deevey, 1949; Oosting and Billings, 1951). The arguments in favor of a tundra zone and a boreal forest zone are of three types: first, fossil evidence of vegetation such as pollen of boreal trees in bogs in the southern United States; second, soil structures assumed to be developed in treeless areas with permafrost; and third, the assumption that the ice sheets must have been correlated with much colder climates which extended to the south of the ice. Each of these will be discussed in some detail. First: assumed migrations of vegetation associations It is assumed that if spruce and fir got to Florida they must have formed closed forests in the Virginia, Kentucky area, and there must have been a zone of tundra between those forests and the ice (Oosting and Billings, 1951). The idea of zones of tundra and boreal forests south of the ice grow out of the Darwinian concept of the major retreats of the vegetation zones before the advancing ice, abetted by the Clementsian concept of the “Climax Formation” which suggests the tundra formation advancing south, iving before it the boreal forest formation. There is actually no evidence against a mingling of spruce, pine and deciduous trees in a mixed forest south of the ice, and Lucy Braun’s work strongly supports such a condition (1947, 1951), although she does not give her data this interpretation. The evidence that does exist from pollen spectra describes a variety 82 WILLIAM H. DRURY, JR. of mixtures of trees with which we have little or no familiarity in the modern landscape, even taking into consideration the differential Preservation of various types of pollen. This should not be surprising, however. We have to suggest climatic conditions during the ice age that differ from all those we know at present and unless we think of the vegetation as organismic, we should assume different combinations of plants would appear in response to different combinations of environmental conditions. Hultén (1938) and Raup (1946, 1947) have shown that in the North, migration into the unglaciated regions has been at various rates and times from various sources. Individual species have migrated as individual plants rather than as a coordinated structured community. Yet that plants do migrate as associations is assumed in the suggestion that zones of vegetation retreated south before the advancing ice. Furthermore, bird and plant subspecies indicate that minor advances in speciation, similar to those which took place in the isolated forests of interior Alaska, have taken place in the needle- leaved forest zone of the Appalachians. This requires isolation of the Appalachian populations and contradicts any suggestion that the boreal forests swamped the Appalachian ridges and iowlands during the Pleistocene. Since the concentration and mingling of the forests during the glacial advances, the retreat of the ice has uncovered a great land surface allowing species to separate out according to their individualities and especially according to speeds of migration. The great vegetation zones of the North are perhaps best thought of as temporarily exaggerated in their segregation, as though in a great natural experiment resembling paper chromatography. Second: assumed conditions for development of soil types and structures In this second argument it must be clarified whether vegetation can be surely identified from fossil soi 1) Can steppe and tundra be told apart? Frenzel and Troll (1952) suggest both tundra and steppe i ages in contrast to forest, as suitable for loess deposition. s points to the possibility that the loess of the Mississippi drainage was deposited under grasslands. At the same time this BOG FLATS AND PHYSIOGRAPHIC PROCESSES IN ALASKA 83 suggests the difficulty of differentiating by fossil deposits between prairie vegetation and tundra, for example, the tundra east of Great Slave Lake. Biidel (1949) uses the coincidence of 10.5° C. isotherm with tree-line and says that all colder than this is tundra and that other treeless vegetation is steppe. Over much of North America such an assumption is gratuitous; in fact, the use of any temperature or precipitation line beyond local coincidence is misleading as is shown, for example, in defining the position of the eastern edge of the grasslands in the mid-western states and the position of both the northern and southern border of the coniferous forest. If tree-line can be drawn to separate steppe from tundra what does this mean? Few, including botanists, appreciate the variety of what is called tundra. Is it the tundra of western Alaska which approaches a peat bog in structure and aspect, or is it an area of lichens below low heaths and dwarf birch, or is it tussocks of sedges, or is it a grassy sedgy prairie, or is it mats of Dryas- Salix? These are all tundra, but they indicate very different climatic conditions. We cannot define the climate of the tundra, nor can we define the climate of the grasslands; we cannot be sure we can tell the two types of vegetation apart. How then can we assume that ice age climatic changes (largely unknown) caused a specific type of retreat in the vegetation preserving the same associations that we find under present climates? Furthermore, although it may be largely true, I am not convinced that the usual assumption that loess is deposited in treeless areas rather than forest is justified. I have seen a heavy mantle of wind-blown silt collecting in tall Oe | forests near Kluane Lake in the Yukon Territory. 2) Structured soils form under forests and a periglacial climate may produce only restricted tundra; do structured soils indicate regional tundra? Benninghoff (1952) has shown the development of spectacular rost features in south coastal regions of Alaska under tall forest vegetation. This does not contradict the use of soil structures as indication of frost climates, but denies the generally assumed requirement of permafrost and treeless vegetation. Ir throws doubt 84 WILLIAM H. DRURY, JR. onto assumptions that such soils indicate a tundra climate (J. Biidel, 1949; H. T. U. Smith, 1949 and many others). The studies in this paper were made in lowland areas in a region where the uplands and many upper slopes are subjected to soil processes similar to those indicated in the uplands of New England and Pennsylvania. Tundra is found on the hilltops in the Upper Kuskokwim River Region and a mixed forest on the warm slopes and dry sites in the lowlands. This sort of vegetation distribution is much more reasonable, and better supported by the meager evidence available, than is a zone of tundra blanketing the uplands and lowlands south of the con- tinental ice in New England. Goodlett’s (1954) work in Potter County, Pennsylvania, has led him to suggest this same sort of distribution of vegetation close to the continental ice of the Wisconsin stage. He suggests unstable soils, structured soils, cryo- planation and down-slope movement of rubble on the uplands near (10 miles) to the Olean Moraines and presumes that the forests may have been eliminated due to soil movement responsible for the thick superficial deposits. But in the south of Potter County, at a distance from the Moraines measured in tens of miles, he and Denny find bedrock out-cropping through a thin mantle of debris and conspicuously less indication of the effects of frost climate. Goodlett suggests that stable areas of stream banks, floodplains and glacial outwash supported trees even near the moraine; and that a variety of trees occupied the southern art of Potter County. Furthermore, he suggests, as did Raup (1951), that the plants which persisted in such areas were those now living in the northeastern United States rather than those of the subarctic and arctic. Stout (1952) in his studies of the soils in the Harvard Forest finds indications ot frost movements, but the movements are of various forces and ages. Since the deposition of loam soils on top of the tills, the movement and congeliturbation have been restrict- ed. This has preserved a sharp contact between the loams and underlying till which solifluction would have distorted and congeliturbation obscured. Evidently then, although the hillslopes show cryoplanation forms, these probably developed while the ice was in the vicinity and soils were relatively stable at a later time during the deposition of wind-blown soils presumably related to glacial advance. spoke eee, ee Re Bate ae pee yt mee aes aay SEF TaN BOG FLATS AND PHYSIOGRAPHIC PROCESSES IN ALASKA 85 Third: assumption that glacial periods are cold periods Flint (1953) words this as follows: ‘““The temperature minima (equivalent to glacial maxima of stratigraphic terminology) ....” This quotation represents the agreement of many geologists that the most important feature of the periods of ice advance is intensity of cold, although beyond the ice limits glacial periods are referred to as pluvial periods. Landsberg (1949) says “The only fact that can be stated without reservation regarding the relation between the Pleistocene Glaciations and atmospheric conditions is that during various stages of this epoch surface temperatures in many regions | were considerably lower than they are at present”. There can be no doubt that there was cooling in the areas of the ice sheets, especially during the summers. Bryan and Cady (1934) pointed out that the important change, however, is towards an increase of precipitation in middle and high latitudes and that the reason there are not continental ice sheets today is that there is too little precipitation in such places, especially during the cold months. In this vein Flint (1943), using ideas first published by Upham (1895) and Leverett (1916), suggests that the continental ice sheets started in the highland areas of Baffin Island, Labrador and Quebec, then expanded southwestward into moisture-laden winds. Precipitation collected where the winds were chilled at the margins of the piedmont: glaciers. We are led to assume an increase in precipitation in the centers of expanse of the ice during an advance. It has further been established by radio-carbon dating that climatic changes during the Pleistocene were world-wide and synchronous (Flint and Deevey, 1951). It seems permissible, then, to suggest a climate at the margins of the continental ice sheets in which there was at least as much precipitation as there is in those regions today, unless the shift in storm tracks was so extreme that the ice fronts were in a rain shadow between two zones of increased pre- cipitation. There exist then the alternatives a) that the margins of the ice sheets were in a rain shadow between two zones of pre- cipitation and b) that there was as much precipitation at the ice margins as at the ice centers and in the areas subjected to pluvial conditions south of the ice. If we assume that the climates were cold and dry, we should have widespread development of wind-blown and structured soils 86 WILLIAM H, DRURY, JR. similar to those found in north-central Europe and central Siberia. Some of the greatest development of patterned ground, and structured and wind-blown soils occur in areas of strongly continental, cold and dry, climate (Troll, 1944; Muller, 1947). Well-formed frost features would be readily detected in such studies as Denny (1951) has made in Potter County, Pennsylvania. Frost features of considerable dimensions are found but they decrease precipitately in clarity and depth of effects from the ice border to the south. Loess is found in the Mississippi River Valley indicating a dry climate, although not one proven to be intensely cold, as I pointed out before. In New England, although loess has been deposited, these deposits are thin and scattered. Perhaps there was little source for wind-blown silts in the broad hills and narrow valleys of New England. Similar valleys in Alaska are deeply mantled with loess. Perhaps all the loess blew out to sea before glacial anticyclonic northwest winds or perhaps the thin loess mantle can be used as an argument that there was a closed vegetation over much of New England in periglacial climates. Actually, then, there is no foundation in observable evidence to show that the ice fronts were in a rain shadow and that a continental tundra climate existed south of them. Instead, it seems justified to assume the other alternative, namely, that there was as much precipitation at the margins of the ice sheets as at the centers of nivation and in the areas subjected to pluvial conditions south of the ice borders. If we assume a climate at the borders of the ice sheets cold enough to produce a zone of tundra and its associated permafrost with precipitation as great as or more than that of today, those areas are ideally situated for widespread formation of bogs, if not in the tundra zone, then in the forested regions bordering it. The species of sedges, heaths, and sphagnum involved in the formation of bogs in Alaska are all known at present to occur in the northeastern United States and in Eastern Canada north to tree-line, and there are broad alluvium-filled valleys available for the development of bogs. It does not seem justifiable to assume a broad zone of tundra without a zone of ni frozen ground, and these without a climate on their southern border eminently suited to the formation of peat and muck. But in comparison to the deposits of peat and muck in Alaska today, i see ne fe tsi en mmf ae i Rlepnhaene as i ee BOG FLATS AND PHYSIOGRAPHIC PROCESSES IN ALASKA 87 there are none. A possibility is that during the post-glacial climatic optimum, the peat and muck deposits disintegrated. If so, the action of disintegration was highly selective and avoided pot-hole and depression bogs in the same area. Evidence is for a wet and cool but not cold climate There seems to be little evidence to support an intensely cold climate, whether wet or dry, very far to the south of the ice. In addition, there is a body of evidence and thinking presented by astronomers and meteorologists interested in these same climatic changes which indicates dampness without intense cold. Bell (1953) and Willett (1953), though they retain Zeuner’s idea of cool summers at 65° north latitude, do not think of an ice age as necessarily a cold age. Willett avoids the term Flint (1953) has chosen, the “thermal maximum’, and calls it “the climatic optimum”. In fact Bell and Willett point out that a general earth cooling would reduce atmospheric moisture and thereby cludiness, air circulation and storminess, producing the opposite effect of glaciation! Willett prefers instead a heating, a selective variation and concentration of heat reception as a result of a steady moderately disturbed state of the sun. These conditions lead to increased moisture uptake at the equator and increased circulation, carrying moisture as clouds north to collect and produce increased cloudiness and storms at mid and high latitudes. These are refinements of Simpson’s (1937) first hypotheses of a gradual heating which produces two glaciations. Under these conditions the mid-latitude maximum cyclonic precipitation (storms) extends 15° farther south into the more or less dry an settled weather area of the subtropical highs and by the same effects the equatorial rains are restricted. The zonal westerlies and the jet stream of the upper atmosphere are intensified. Storm tracks are displaced south and the storms are more frequent, Producing the pluvial period. What is responsible for the coincidence of the southern borders the several ice sheets? The general discussion presented by Brooks (1949) suggests on climatic evidence that if the polar ice cap grew another degrec of latitude in radius it would advance as a result of intrinsic forces until it met major climatic resistance. He says the ice sheet would extend to about 40° North Latitude where it would meet 88 WILLIAM H. DRURY, JR. warm climates that would prevent the ice from advancing further south. Flint (1953) points our that the similarity of termini of the major ice sheets indicates a similar controlling deterrent to their further expansion (as is assumed in C. E. P. Brooks’ discussion). Warmth is the effective agent in the northeastern United States. Presumably, then, other than local effects from the ice, the general warmth that halts the ice advance would govern the climate at its border. We are left with a relatively cool (but not cold) and wer climate at the borders of the ice in the northeastern United States and maritime Canada which would allow a narrow zone of arctic conditions reflecting the glacial anticyclones. These conditions make it readily conceivable that forests similar to those of the deciduous forests of eastern North America survived rather near the ice front on well-drained, stable soils. Such survival does not require the absence of considerable areas of soil instability in the uplands and on poorly drained slopes. The climatic conditions as suggested in this way indicate treeless areas on the mountains of New England, the Adirondacks, Appalachians and south to the Great Smoky Mountains, with forests largely of Spruce and Fir on the cold slopes below them. e southern extension of these trees is probably correlated with cool summers as E. Dahl suggested for southern limits of alpine plants in Scandinavia (1951), among other factors. But the extension is not as a mantle of coniferous trees covering all the central states, rather as deciduous trees in patches and sheltered lowlands mixing into the coniferous forests south and east of the mountains. Mixed deciduous and coniferous trees in combinations we do not know today must have existed on the exposed coastal plain in the north, the piedmont and lowlands in the southeast, in the Mississippi embayment and shallow areas in the northern Gulf of Mexico. Many deciduous trees probably survived in warm valleys in the central and southern part of the present deciduous forest area, while the present coastal plain flora moved only a little south. The forests of that time were greatly changed from what we now know and species grew together in combinations not present in our forests. The climate was greatly different from any we know and to think in terms of present vegetation associations and zonations persisting intact then is unjustified. Ic leads to , Pe et eda Cig he eee tata hae i heat one ae a Pittc ere aaa Pee el See Sree ON het BOG FLATS AND PHYSIOGRAPHIC PROCESSES IN ALASKA 89 the conviction on the one hand that the forests existed intact in sits and on the other that if Spruce and Fir forests mantled the southern states the Deciduous Forests must have been in Mexico. In contrast these Alaska studies as well as recent pollen spectra indicate that a conglomeration of tree species existed all together there, which species are widely separated in different associations now. Similar to the familiar pattern of migrations of birds, some species evidently migrated a long way from their present northern limit, while those now limited to the southern states probably moved only a little farther south and the two populations existed together in forests very rich in species. e conclusions are supported by the fundamental ideas of Fernald’s (1925) suggestions of plant persistence in the Gaspé- Newfoundland area and the now submerged areas of the coastal plain. They agree with Braun (1951, p. 145) in parts of her contention of survival of a mixed Tertiary forest in the near neighborhood of the continental ice sheet margin. These sug- gestions indicate availability of the land surface for persistence of forests. Instead of persistence of forests as we know them, I would prefer to suggest that a mixed forest resulted. This was repeatedly modified by migration and mingling of forest species with the changing climates. Migrations of forests south during ice advances would add to the richness of these mixed forests in the central refugium from which segregation of forest types has taken place since then by migration of species. In fact the Lake Forest (Hemlock-White Pine, Canadian zone) and boreal forests can be considered as segregates from mixed forest of glacial time. These suggestions support the presence of mesophytic plan: species near the ice limits at the margins of the continental sheets. ey are not consistent with the thinking of many pollen analysts and students of Pleistocene geology. The concepts of many of these students have been largely inherited from European studies (Deevey, 1949). In Europe, the conclusion reached that there were arge areas of tundra is supported by floristic studies, by paleo- botanical examination of the Alleréd deposits, by the physio- graphy of northwestern Europe, and by the regional geology which did not permit a reservoir of temperate populations to exist a short space beyond the ice margins. Temperate populations were trapped in trans-Alpine localities, and many remain trapped there today. There is no such evidence from the northeastern 90 : WILLIAM H. DRURY, JR. United States. These conclusions in no way contradict the suggested occurrence of boreal forest trees far to the south of their present distribution. On the contrary, it is in keeping with the importance of local microclimates. Exposed areas far to the south would be expected to show bitter local climates. Pollen analysts have been searching for clear indication of a tundra zone in New England with negligible success. Even deposits that rest on Tazewell till in southern New England do not show the “tundra — forest tundra — tundra” that Deevey (1953) says should be there. In fact the best evidence points only to local and questionable tundra. As Deevey says, “forests appear to have lived much nearer the ice in North America, so that an abundance of forest tree pollen masked the evidence of tundra”. These speculations apply to the climatic conditions at the margin of the substages of the last (Wisconsin) ice, lowan or Tazewell in Potter County and probably Cary at the Harvard Forest. Serious misunderstanding has resulted in the past from not efining this. When Fernald referred to the unglaciated areas of Newfoundland as untouched by the ice, as if there were but onc glaciation, he never showed he knew better. Similarly the geologists who have proven his sweeping statement wrong sho realize that for the value of Fernald’s suggestion, Nebraskan, Kansan or Illinoisan ice covering Newfoundland is of little if any significance. But the farther removed the area from the immediate vicinity of the ice front the more probable it is that the generalizations of plant persistence will apply to the whole Pleistocene Epoch. Thus, in the Cumberland Plateau Region the suggestions of persistence probably apply to all of the advances. 3) THE RELATIONS OF THESE PHENOMENA TO THE CONCEPT OF CLIMAX IN VEGETATION The third of the botanical implications applies to the funda- mental theories of the way plants associate together. The dis- cussion of these will first examine the general concept, then the development and uniformities of plant associations, the historical position of the formulation of the concept of Climax and finally analyses specific to the vegetation studied in Alaska. A) General Examination of the concept The concept of climax has been widely discussed since it was first presented about the turn of the century (Hult, 1885 and BOG FLATS AND PHYSIOGRAPHIC PROCESSES IN ALASKA 91 Schimper, 1898, Cowles, 1901). Within the concept of climax is the idea of succession, a much older idea whose roots go back to Kerner (1863) and are in classical writings such as those of Theophrastus. In America the idea of succession goes as far back as Thoreau (1860) and Belknap (1792). These concepts were developed to their full logical implications and complexity by Clements (1916, 1936). It seems to be the unfortunate fate of such a widely accepted and popularized theory that its original conception is soon so modified that it is not recognizable. As with Darwinism and the 19th and 20th Century controversy over mechanism vs. vitalism, each individual has his own “real climatic climax concept”, and many of the later ones resemble the original only remotely (Tans- ley, 1935, 1939; Whittaker, 1953). The climax idea remains a convenient system for the description of vegetation, however, and as such has been used by recent students of vegetation in Alaska. For example, Hultén (1941) refers to the interior forest as a White Spruce and Birch forest and Stoeckeler (1949) refers to the climax White Spruce and Birch forest established on the high, dry floodplain approximately 175 years after deposition. But Stoeckeler also refers to climax Black Spruce muskeg on wet, fine-grained materials, and includes no discussion of a time relation between these two. In this he has accepted a modified climax concept which includes a physiographic climax that Clements would call a serclimax. To Clements the climax was the mature organism, the end Product of successional development, and many Continental plant sociologists treat vegetation as uniform and integrated as if an Organism. A. G. Tansley represents a departure from this scheme and has come to speak for a large part of the English plant ecologists. The climax to Tansley (1935, 1939) is any stable community, Organized, mature and homogeneous under the existing conditions no matter what controls those conditions. They may be edaphic, Physiographic, biotic, due to fire or mowing by man; so that succession is marked by a series of climaxes each relatively stable, differing in habitat, life form, and floristics. Tansley really uses the term climax as American ecologists use the term plant association and he uses plant association to refer to what most Americans refer to as a climax formation. pF WILLIAM H. DRURY, JR. In general usage at present, the climax refers to a stable vegetation association toward which all vegetation types are moving, one which will reproduce and maintain itself as long as climatic conditions remain constant. Other characteristics such as degree of productivity, mesophytism, Sepabeenees, range, and uniform, high life-form are inherent in this use. The concept of climatic climax seems to have been based on the idea of soil stability during long periods of erosion and 3 gratuitous assumption that climatic conditions remain constant, These, in turn, as Cowles (1901) pointed out depend on geomor- phological concepts presented by William M. Davis (1899 in 1909) concerning cycles of erosion, uplift, and deep soils of the peneplain. Soil stability in the sense recognized in the past in temperate climates is not usable in the north because of the widespread effects of frost action (Raup, 1951). In fact, the recent work in Potter County, Pennsylvania (Goodlett, 1954) and at the Harvard Forest indicates that it does not exist in temperate climates either. H. T. U. Smith (1949) presents the best summary of the effects of frost climates on soils, and Denny (1938) de- scribes a specific case in southern New York. My observations in the Upper Kuskokwim River Valley indicate that the White Spruce-Birch forest of the floodplain is succeeded by a Black Spruce forest in the course of entirely natural vegeta- tional and successional processes, and entirely natural processes (Versumpfung or paludification) swamp this forest in turn, and the climax is perhaps just as well considered to be the peat bogs. As has been pointed out, mapping of vegetation over an area of 8,900 square miles in the region shows thar about 60% of the land surface is forested with Black Spruce. If, in the changing sequence of vegetation associations, one type is more stable than another, it should occupy an increasingly larger area with the passage of time. According to this, the sedge meadows and high brush bogs are as certainly the climax as the White Spruce and Birch forests. These contradict any consistent use of the climatic climax as understood in the recent writings. In this way the behavior of fine-grained saturated sediments on freezing and thawing contributes to Ce eal instability, controlling the succession of forest types and bogs. Pergelation and Depergela- tion (Bryan, 1946), factors of the environment which are not BOG FLATS AND PHYSIOGRAPHIC PROCESSES IN ALASKA 93 purely edaphic, obliterate the logical climax forest and lead succession toward a less mesophytic, less productive vegetation of lower life-form classification. Instead of attaining any main- tained equilibrium, the forests are in a state of change extending over an indefinite period of time, without necessarily any change in climate. This may be what Watt (1947) was getting at, but it is not clear. He includes forces of integration (towards order) with those of disintegration (towards disorder) in the community at the same time. The processes remain constant while phases change and yet the whole community remains essentially the same. Elsewhere, specific studies in nearly all major vegetation types such as Richards (1952) in the tropical rain forest, Shreve (1942) in the American deserts, Gleason (1926) and Cain (1947) in the deciduous forests of temperate America, and Griggs (1934) in the Arctic, have led their authors to reject the use of the climax concept. Recently Malin (1948), an historian, by a review of the problems and results of authors applying the climax concept to the “type locality”, the grasslands, proves better than any argu- ment could how theoretical and abstract it is. The studies of vegetation made in the Upper Kuskokwim River Region support in detail the conclusions Gleason reached about plant associations (1917, 1926) and succession (1927). In these Papers Gleason points out that a plant association is the expression of overlapping and complementary physiological requirements of those species of plants which can successfully grow on the site, are not excluded from the site by growth of other plants, and Whose seed source is available. Succession is the result of constant migration of individual plants into the site, of which some can become established so that some previously present species becomes eliminated. He argues against homogeneity of associations either ocally or over distances. Tansley’s climax concept appears to lie between that of Clements and this of Gleason, but where can the line be drawn? Does Tansley’s mowing climax last the life of one farmer, or One method of farming? Where is it separated from any group of plants one sees in nature which is stable during the lifetime of the plants growing there? Where can this be separated from the longer periods of stability due to grazing of wild animals, and, as Gleason (1927) said, from the longer periods in physiographic 94 WILLIAM H. DRURY, JR. and climatic change, from those slowly changing as a result of migration of new species from distant dispersal centers (Gleason 1917), or from changes due to species evolution (Cooper, 1926)? Yet the majority of ecologists still subscribe, if not to the organismic concept, to ideas which presuppose the organic, that is, development. For example, ecologists still refer to forces in the environment which prevent the development of the climax with the consequent maintenance of a subclimax stage. any Americans carry it further and use the climax expanded to include animals in the biome. How can there be a biome, a spruce-moose association of continental extent if there is any reality to the complex history, soil age, origin and migrations of flora from refugia as clearly shown for the Lake Athabaska- Great Slave Lake Region by Raup (1946)? Some zoologists, as Pitelka (1941), have rejected the biome but these men did not gu far enough in their analysis and see that the absence of fit applies just as well to plant groups. Perhaps Pitelka’s fault is only mis- placed respect for plant ecologists. Tansley (1935) rejects the concept of the biotic community but does not carry this through to the vegetation. Perhaps one of the chief reasons Clements’ classification has persisted and probably will continue to persist with those who look for a system of description, is that it has been thought all the way through. Thus, once vegetation is fitted into Clements’ system, the classification is done. In the succession to the climax once started, the vegetation is swept along in the stream and once placed in the stream all is explained. Gleason’s system demands that the author interpret and come to understand what he sees. Description only points to the need for understanding the process- es at work. Papers by Hopkins and Sigafoos (1951), Stour (1952, Warren- Wilson (1952), Goodlett (1954) look into what it is in the soil and external forces which make the vegetation what we find, rather than suggesting that it is that way because it is at that certain point in the stream of “development”. B) Examination of the ideas of development and uniformity of plant associations The use of the term or idea of climax indicates by definition or by implication that there is development or increase of See eee eee Nd si hehe ge apes Veeecuepieere SS TT ta Pe ae ee es eee Nie Sie aS Fone na ee ieee ane Bee en OE San ae ater Te pet ke eee tT \ 3 aoe ‘ Beis les heres BOG FLATS AND PHYSIOGRAPHIC PROCESSES IN ALASKA 95 organization or integration in successional sequences, such as the zonations of the depositing slope of the floodplain or those of the shelving shore of a silt-bottomed lake. The idea of develop- ment is important. Development to Clements was the embryology of the superorganism, the progress along a unidirectional path of changes of vegetation, each stage of which prepared the way for the next. “Succession is the process of the reproduction of a formation, and this reproduction rocess can no more fail to terminate in the adult form in vegetation than it can in the case of the individual plant’. (Clements, 1916). European ecologists, in common with most American akg have discarded the idea of the association as an organism, but the idea of development is actively retained (Tansley, 1939, 1953), and with it the idea of relatively rapid development toward an equilibrium which is maintained over a long time. For Tansley (1935), development is the progressive maturing of vegetation, its integration or establishment of consistent internal relationships producing homogeneity. To him, development is a Particular type of succession and centers around changes in the stable habitat due to biotic factors acting under uniform condi- tions. In this, only organically controlled succession is develop- ment. Tansley’s description of the development of marshes, fens and bogs in England is convenient for examination. Succession and Miiicmen: should be found in wet areas, the hydrosere of developmental classifications as Tansley has suggested himself. For this reason I want to discuss development and uniformity in plant associations and specifically their application to wet areas. The best studies of bogs have been made in Europe, using the system of plant sociology which is based on Clements’ logical system. For this purpose the difference of points of view withia the European schools, the Scandinavian vs. British vs. Montpellier On the nature of an association, are omitted, because they all include as a basic assumption the idea of uniformity of associations and the idea of development of vegetation. Among them, the differences seem to lie in terminology for the associations. At first, bog-vegetation changes seem to support an idea of Succession, and this seems especially to be true of the hummock Srowth of raised bogs. But the final product and total growth 96 WILLIAM H. DRURY, JR. of the raised bog perpetuate all the phases of the succession and there is no effective change to a stable “climax vegetation” at all. Instead, the climax is the whole gamut of successional stages. That there is sequence or zonation of wet-dry requirements is obvious, and that certain species of sphagnum will colonize a site only when it has been made less wet through the growth of more hydrophytic species is a direct corollary. But this is nor development because the wet-site sphagna colonize low areas in the new, higher surface and supplant the dry-site sphagna just as they in their turn are supplanted (Osvald, 1923). In fen vegetation, as described by Godwin and Bharucha (1932), increasing dryness allows the invasion of woody plants into Phragmites reed-swamps to form carr. At Wicken Fen, Cambridgeshire, England, it has been suggested that the limiting factor is the height of water level in winter. This high water level is so close to the low-water level of summer that even Tansley (1939, p. 767) finds it hard to suggest that there is “development” involved. There is not enough change in the volume of organic materials on the site. The second concept inherent within the plant sociological use of the idea of the association is uniformity of the vegetation. This is at the heart of the European terminology in plant sociology. That a name such as Callunetum or Sphagnetum is applied to a vegetation association (community of Tansley’s usage) indicates uniformity. “The indicator species” as a term implies a collection of conditions which are repeatable. This seems to grow out of the early 20th century botanical concept of a species as a type specimen without consideration of it as a variable population. Both botanists and zoologists now conceive of species, subspecies or varieties as variable populations, not as a series of morpho- logical criteria represented by a type specimen. Aggregations of taxa into associations should include, then, an idea of combina- tions of variations. If, on this basis, a Sphagnetum is conceived of as all the variable conditions possible in a bog, it is a justified idea, but not a much better term than sphagnum bog. As soon as 2 more specific term is presented, such as Callunetum, the problem arises of resolving within it all the various combinations of species possible under the conspicuous presence of Calluna. The implicit idea (which is in the term taxon) of a repeatable expression of definite structure constantly presents itself. Vedi aside ati aa ui a kts ae ae MEAL Aare sea EN Metre ab Nee ea pe ses hue sea BOG FLATS AND PHYSIOGRAPHIC PROCESSES IN ALASKA 97 Gleason pointed out that this homogeneity does not occur in space. Why then can it be assumed to occur in time? If there had been homogeneity of vegetation associations in the past there would have been selective advantage for animals to have a species-defined habitat and the biome to be a reality. But Pitelka (1941) shows this to be a fallacy and Tansley (1935) rejects the biotic community. This in itself argues for the lack of homo- geneity of a community or association of plants in the past in the “state of nature” as well as in the present under man’s disturbing influence. C) Historical position of the climax in the study of vegetation Developmental stages of the organism being integrated toward the climax is justly credited to Clements. If succession, which Clements so emphasized without the deterministic goal of develop- ment to climax, is to be used, both it and climax are misleading. For instance, when Cowles first presented the idea of the dynamic basis of plant society or association classification, he did so carefully separating the factors of climate from those of physio- graphy. He indicates that physiographic factors can have more local influence than climate, and that retrogression is an important part of the changes of vegetation. Although the general trend is towards mesophytism, it is not universal. Cowles says: “The condition of equilibrium is never reached and when we say that there is an approach to the mesophytic forest, we speak only roughly and approximately. As a matter of fact, we have a variable approaching a variable, rather than a constant.” He further includes the idea of development, but bases it on physiographic changes from youth to maturity and old age of the cycle of erosion (1901, pages 78-81). In other words, the changes are correlated with changes in the physiographic or edaphic conditions, and the plant associations are the pawns of the geological changes. Cowles does not build his structure on the changes which plants create in their environment as if these changes prepare the way for the succeeding generations of plants in a determined direction of succession. He says that the order of change of topographic forms runs according to well defined laws (those of W. M. Davis, 1909); so plant succession must 98 WILLIAM H. DRURY, JR. be in orderly fashion adapted to the new conditions. It was Clements who first brought in the concept that out of certain historical causes definite effects can be forecast. The difference is one of innate process determining the final goal of vegetation, as opposed to vegetation changing according to the changes in its edaphic surroundings. What Clements brought was determinism, a philosophy deeply rooted in mechanistic scientific thought of the early 20th Century. Determinism has come under more and more criticism in all kinds of fields since then, and attacked at its philosophical heart by Whitehead (1925), long after Hume first objected. Once the determinism which Clements put in has been eliminated, there is no essential disagreement between the ideas presented by H. C. Cowles and H. A. Gleason, and they are consistent with those of Warming and Kerner. Clements should not be thought of, though, in this absolute comparison with what has been learned since his time. He represents a revolt against the widespread idea in America that certain areas of vegetation were abnormal, such as the grasslands. Malin points out how the prairies were considered as inadequate in many ways and the continuation of such thinking exists in the great American myth of the Great American Desert. Clements and men of his time were emphasizing that vegetation was natural to a certain type of climate. The trend of philosophic thought at the turn of the century was deterministic and regional in geo- graphy and his conviction of the naturalness of vegetation types led Clements into conviction of the correctness of deterministic thinking. At that time political and sociological thinking were also emphasizing stability and equilibrium after the Industrial Revolution and 19th Century wars and before World War I. He became the spokesman for that group because of his ability to explain his subject. Criticism of his contributions is empty. And his contributions are many, but like many other scientists his contributions were not the ones he thought most important. For the reasons presented here that the terms are misleading and do not contribute to an understanding of vegetation I have dropped completely the words Climax, successional, development, biome, sere and the words associated with them. It should be realized, however, that Gleason published his objections to Clementsian doctrine in 1917, as soon as Clements large book appeared (1916). In this contemporaneity remnant yet at er eh RS nN A er cea re SP earn ee apearr ga nae eames Nee SS PELTON Pg Oe all Pe EO eR Ee oe tee si oes oie! - oe as. . oe BOG FLATS AND PHYSIOGRAPHIC PROCESSES IN ALASKA 99 Gleason’s clear contributions cannot be regarded as an historical development growing out of additional knowledge, but rather as the result of greater insight and better understanding. D) Specific analyses Returning to the problem of the vegetation of the bogs in the Upper Kuskokwim Region: the cyclic processes in operation in the bog areas indicate and emphasize that the vegetation of boreal and arctic regions can be understood better and must be studied against a background of the history of land surfaces, of vegetational migrations, of the intrinsic variability of the small number of plant species, and of geomorphic processes such as cryoplanation and cryoturbation. These are not the only factors but each of them has an important effect on the nature of the substrate, the flora, and the condition of the surface at the time of plant establishment. Certain of these forces can be seen in action in the bogs. They are responsible both for the kaleidoscopic patterns of the vegetation and for the type of order that exists. The physiological variability of the species of plants, the small number of species, the youth of the arctic and subarctic vegetation (not flora), the disturbance of the substrate by soil frost, and the flora are discussed in the following paragraphs. First: Morphological variations in many of the genera of the north such as Salix, Betula and the grasses occur in the same way that random genetic on do. Certain combinations of characteristics appear without regard for areal segregation or Plant geographical ee Others occur as clines across the north, or as isolated particular combinations. At the same time other populations are very limited in variation. All of this has ed to naming of many varieties, hybrids and questionable species. ere is no reason to expect that physiological variation should appear in ways different from the morphological patterns. Repeated isolation and reassembling of populations correlated with climatic changes of the numerous ice advances seems to have Produced much heterozygosity in some of these populations and very little variability in others. Their behavior is very well explained by the genetic and evolutionary processes suggested by Mayr (1954) of variable gene backgrounds and effects of Populations subjected to temporary isolation. This leads to the 100 WILLIAM H, DRURY, JR. fact that morphologically defined species populations occupy a wide variety of microclimates and thus enter several plant associa- tions. Because of varied degree of success in competition in these several habitats, they appear in all degrees of conspicuousness. As one moves through an area of “bog” or “low tundra”, first One species is primary and then another; then a combination of these three, then these other three. There may be a total of only P 2. species involved and 20 species present at any and all sites, and in nearly all possible combinations. Each different combina- tion of environmental requirements allows certain ecotypes to occupy the area: for example, while Carex lugens is primary in a wet hollow in the tundra in one spot, it may be primary in sphagnum under forest on sandy alluvium in another. The plants are the same morphological species population but quite different ecotypes (physiological variations or races within the species population). Yet a slight change in either habitat may allow another species such as Eriophorum vaginatum to take over and nearly eliminate the Carex. At each site different requirements overlap in different ways. Each recombination of environmental factors, such as variations in moisture, temperature or exposure, allows some species to be included in the association and excludes others. In some places variations that are only slight will eliminate a certain species while in others what seem to us to be fundamental habitat variations seem to have no important effect. It seems that aly wits variation within the species is often greater than that between two species. Second: pee there are few species, those species which can survive have a variety of habitats available to them. Interspecific competition is not so marked, or, better, there are not as man different kinds of competitors. Instead of scicmeihie competi- tion, competition is between ecotypes, and the results do not appear in floristic lists. Yet the microclimates of the North are as variable as those of more southern areas. Such variations cross the limits of tolerance of certain species or physiological variants and _enter the tolerance limits of others. With only a few species av: recombinations of these species in all possible ways result, a certain ones becoming conspicuous here and others e so that the vegetation seems disorganized or weedy (Griggs, 1934). The number of species is not adequate to provide different species BOG FLATS AND PHYSIOGRAPHIC PROCESSES IN ALASKA 101 names to describe each habitat, but the recombinations represent real features in the vegetation and habitat. They are just dis- tasteful to those trying to arrange an orderly classification. If new species were added to the lists at each site a new association would be named and accepted. The general vegetation discussions of Antevs (1932) and Polunin (1934-5, 1948) show the problems of rationalizing these dif- ficulties in preparing a regional description of the vegetation similar to those prepared for temperate regions. Tansley indicates meeting with this problem in his discussion of Arctic-alpine areas and Blanket bogs (1939). The effect is probably the result of numbers simply. The number of species in the total flora, if plotted against latitude, increases following an “S-shaped” curve from high latitudes towards the tropics. On the other hand, the number of individuals of any one species decreases following an S-shaped curve from high latitudes towards the tropics. In the middle latitudes, where these lines are roughly parallel, plant associations assume a regularity of sorts, and have been attributed qualities of organismic reality. In the north endless variations occur among a few species. At the southern end the same problem exists but for a different reason. As Cottam, Curtis and Hale (1953, p. 755) put it: “In Tropical rain forests, with 60 to 90 species per hectare (Black, et al, 1950) the number cf quadrats or points needed for even reason- able accuracy becomes impossibly great.... This failure of the usual sampling methods to give accurate determinations of such a basic, simple characteristic as relative composition is no doubt responsible for the emphasis placed by the Tropical ecologists on physiognomic classification (Beard 1944, Richards 1952)”. On this basis the suggestion of Braun-Blanquet that the vegeta- tion of the North is simple and that of the Tropics highly complex and organized seems to make no sense. The difference is one of the number of species available, and when it has been said that the vegetation of the North is less well organized or simpler, what has been meant is probably that it is made up of fewer species. It certainly is not simple. Third: The short periods of time available for migrations of ant associations since the exposure of the land over large areas of the North have not provided opportunity for spectes to settle out into their “normal” or “equilibrium” relations with eac 102 WILLIAM H. DRURY, JR. other. In other words, with the intrinsic problems described in 1) and 2) which require variable associations in the description of plant cover, there is the additional complication that the plants have not yet settled down in their inter-relationships, if they ever do. Although the species found in the North may be old, it cannot be doubted that their combinations are new to the areas they now occupy. The same is certainly true for bog vegetation within the forested regions. Gleason (1926) has suggested that increased uni- formity of “mature” vegetation types or associations is a result of the elimination of species which cannot compete during long periods of migration. He uses migration to refer to persistent attempts at invasion by species from surrounding habitats. This may be called “development” of the plant association in terms of clarifying its species composition, but it is not the development towards a determined goal which Clements conceived and which those who use the term imply. Fourth: Repeated freezing and thawing of the ground on which these plants grow churn the soil materials, inhibit the develop- ment of soil profiles, and disturb the habitat, all within the life span of the species which occupy that spot (Raup, 1951; Hopkins and Sigafoos, 1951). All species must remain adapted to disturbed conditions and be capable of “pioneering”. Species comprising the vegetation cannot develop complex structural inter-relationships because species are not able to occupy the surface undisturbed. Soil movement may destroy an individual plant of one species; then, within limits, any of the species from the surrounding vegetation will colonize the spot. To the degree that each species must retain characteristics of pioneer vegetation, the vegetation is weedy as Griggs suggested and migratory as Crampton (1912) suggested. Instead of inter-plant associational organization, most regular patterns of the vegetation are based on structured soils or other frost features which act through soil conditions such as drainage and disturbance. This produces the societies low on the sociological scale of the European phytosociological system and is in part justified, but largely an illusion. Fifth: The flora of wet places in the Upper Kuskokwim River Region is of three quite different types: 1) that of silt-shored ox- bow lakes or sloughs; 2) that of peat bogs; and 3) that of a peat bog-invaded silt-shored slough where the two types exist in com- bination. Such a bog has the species from both major types and BOG FLATS AND PHYSIOGRAPHIC PROCESSES IN ALASKA 103 because of its disturbed conditions has many additional species. The history of disturbance can usually be read both in the flora and in the physiography, and, obviously, changes in flora are fundamental to the vegetation patterns and life form aspect of the bog, and to the associations of plant species. SUMMARY Soligenous bogs act as traps for alluvial or aeolian silts and, persisting during a period of similar aggradation, are also sufficient to explain muck deposits without repeated cycles of invasion of the forested surfaces. These two processes will form muck deposits in the valley centers where the valley slope processes cannot Se functional. Such valley-center mucks have not been satisfactorily accounted for in the past, and their existence has thrown some doubt on the processes suggested for valley-side muck development. Physiographic and physical forces active in the floodplain deposits control the distribution of bogs and the patterns of bog vegetation on the bog surfaces. 104 WILLIAM H. DRURY, JR. Recognition of the existence of the processes described here will contribute to the understanding of the history of the Pleistocene by application to the study of Quaternary deposits, and to the interpretation of the periglacial climatic conditions for which they furnish evidence. At present the formation of bogs is an important force in the control of the pattern of vegetation in the interior of Alaska, and is contributory evidence for the development of ideas of the relation of soil instability to vegeta- tion, and to slope processes and slope development. It is an extension of the study of cold climate periglacial geological processes from the slopes to the alluvial lowlands. The studies indicate both the persistence of bogs during periods of climatic change and the continuation of change in bog-to-forest relations in the lowlands under constant climatic conditions. The formation of peat bogs as described depends upon the luxuriant growth of sphagnum moss, which is much more abundant in forested than tundra regions. The mucks of Interior Alaska that were developed by cycles of bogs suggest a boreal (forest) rather than arctic (tundra) environment for their deposition. THE FLORA OF THE BOGS In listing the flora, a compromise must be made between the tendency to list only those plants that are a conspicuous part of the vegetation type and the tempration to list all species that possibly can be squeezed in. The list of flora of these bogs is complicated by the infiltration of the Alaska Range in the areas of glacial till and outwash from the Alaska Range. In the following list, those plants that are considered proper members of the bog flats are unqualified. Those that are part of the Alaska -Range Flora, are marked with (a). Those that grow in ponds or puddles or on the floor of the black spruce muskeg and thus are not really characteristic of the flats are mark The floras of J. P. Anderson (1943—1952) and E. Hultén (1941—1950) were used as a starting point in identifying the higher plants. Documenting specimens of the plants will be deposited in the herbaria of the United States National Museum the Harvard Herbarium, the National Museum of Canada aa the Riksmuseet at St Im. BOG FLATS AND PHYSIOGRAPHIC PROCESSES IN ALASKA 105 LICHENS I have not oo to list the entire lichen flora of the bogs. All of ae lichens grow in the muskegs on the drier Sphagnum fuscum hammocks or living or ar trees. Species identifications of my collections were made - Lucy Raup. STICTACEAE * Sticta pulmonaria (L.) Hoffm. on dead trees PELTIGERACEAE * Peltigera — (L.) Willd. on hummocks and stumps * Peltigera canina (L.) Willd. on hummocks and stumps . hen arcticum (L.) Torss. on moss hummocks CLADONIACEAE * Cladonia alpestris (L.) Rabenh. between or on hummocks * Cladonia rangiferina ae hig between and on hummocks * Cladonia sylvatica (L.) H do. * Cladonia sylvatica (L.), f. "sets Oecd. do. * Cladonia deformis (L.) H on Sphagnum hummocks and on stumps * Cladonia a. ie Schaer. between and on hummocks * Cladonia uncialis (L. do. * Cladonia elongata (ac) on do. (more common under — C. gracilis (L.) W White Spruce) * Cladonia lepidota Nyl., f. cerasphora on Dicranum and on stumps * Cladonia pyxidata (L.) Fr. : on dead trees : — — (Sommerf.) Vain., between and on hummocks f. De i (Bor.) Th. Fr. . Clidonsa pares (L.) Willd. do. PARMELIACEAE * Parmelia physodes (L.) Ach. on dead spruce and birch branches * Parmelia saxatilis (L.) Ach. o. . eer sece islandica (L.) Ach., var. Delisei between and on hummocks ( . Ce islamdica (L.) Ach., var. platyna do. Ach. r. * Cetraria cucullata (Bell.) Ach. do. USNEACEAE a on dead branches do * Usnea longissima Ach. Alectoria jubata (L.) Ach. 106 WILLIAM H. DRURY, JR. Mosses This list, again. does not claim to be complete. However, as_ most of the $ er | re ed. The species of Liverworts and Mosses except for Sphignum were identified by Dr. Wm. C. Steere, Jr. A field reference herbarium was made, using specimens he identified, and all field identifications were referred to those standards POLYTRICHACEAE Polytrichum commune Hedw. in muskegs or tractor-trails * Polytrichum juniperinum Willd. do. DITRICHACEAE * Ceratodon purpureus (Hedw.) Brid. tussocks and logs, grass slough Distichium capillaceum (Hedw.) Bry. Eu. drained till pool DICRANACEAE * Dicranum Bergeri Bland. muskegs * Dicranum elongatum Schleich. do. * Dicranum fuscescens Turn. do. * Dicranum scoparium Hedw. muskeg Oncophorus Wahlenbergi Brid. drained till pool. rotting wood POTTIACEAE f Didymodon rufus Lor. drained acta” Probably present because of the Tortella fragilis (Hook. & Wils.) Limpr. do. ' mineral soil Trichostemum cuspidatissimum do., primary moss; primary Card. & Ther. also in many Morice Sciepus a SPLACHNACEAE Splachnum luteum Hedw. on moose droppings anywhere AULACOMNIACEAE @ Aulacomnium acuminatum (Lindb. & Arn.) muskeg Paris Aulacomnium palustre (Web. & Mohr.) — slough and recently Schwaegr. * Aulacomnium turgidum (Wahlenb.) wakes sinc grass slough Schwaegr BARTRAMIACEAE «@ Catoscopium nigritum Brid. drained till pool ES Ee ee et ee ee ee ee BOG FLATS AND PHYSIOGRAPHIC PROCESSES IN ALASKA 107 MEESIACEAE * Paludella squarrosa (Hedw.) Brid. grass slough MNIACEAE * Mnium affine Bland. stumps HYPNACEAE * Calliergon cordifolium (Hedw. ) Kindb. grass slough in forest do * Calliergon stramineum (Brid.) Kind do, a ing aga (Hedw.) primary, M)rics-Serrpus areas * Clin denies (Hedw.) edge of grass slough, forest Web. Drep: pens pelea (Hedw.) Warnst. bog pool invaded by flats, and M —— areas ae sue au exannulatus (Gumb.)_. pool in Wa Gacucuie badius (Hartm.) R old floodplain slough * Drepanocladus fluitans (Hedw.) ui muskeg in tufts, and floodplain aa ee revolvens (C. Mull.) primary, nag -Seirpus, Carex ruginvala are ’ ie ee uncinatus (Hedw.) Warnst. dry muskegs * Hylocomium splendens (Hedw.) Bry. Eu. muskegs Hypnum Bambergeri Schimp. Myrica-Scirpus areas, primary Hypnum callichroum Brid. invaded floodplain pool Hypnum Lindbergii Mitt. grass slough in fans Hypnum subi sq. mu Pleurozium Schreberi (Wéilld.) Mitten muskeg Ptilium Crista-castrensis (Hedw.) de Not. muskeg oe triquetrus (Hedw.) muskeg *ta Rhytidium rugosum (Hedw.) Kindb. white-black spruce forest Scorpidium scorpioides (Hedw.) Limpr. primary in Carex raginata, Scirpus-Myrica areas Tomenthypnum nitens (Schreb.) Loeske invaded flood-plain sloughs LESKEACEAE *@ Thuidium abietinum (Brid.) Bry. Eu. — muskeg SPHAGNACEAE _ I identified the mosses of this family using LeR. —— ge (1913). They are the —— joes in the dev Ans of the bogs. A e found within the bog areas, but certain ones are characteristic of margi se area such as ain piste aarad or Fold floodplain sloughs. These are are + tev sca ma tce species are 108 Cal "3 * * * - *« eon 4 gellanicum Brid. Sphagnum it L: Sphagnum papillosum Lindb. Spha Sphagnum ee Hornsch. agn Sphagnum Dusenii C. Jens. Sphagnum Lindbergii Schimp. Sphagnum usum Warnst. Sphagnum pelehrum Monge Warnst. Sphagnum recu Sphagnum ri nei a st Sphagnum capillaceum (Weiss) Schrank Sphagnum fuscum (Schimp.) H. Klinggr Sphagnum plumulosum Roll Sphagnum quinquefarium ( — Wanrnst. Wilso Sphagnum rubellum Russow Sphagnum robustum (Russow) Roll HEPATICAE MARCHANTIACEAE ecu polymorpha Con halum conicum ( (i) Dumort. Preissia iatiats (Scop.) N PTILIDIACEAE Ptilidium ciliare (L.) Nees CEPHALOZIACEAE Cladopodiella sp. HARPANTHACEAE Mylia anomala (Hook.) S. F. Gray WILLIAM H. DRURY, JR. emergent, sedge meadows, brush do. do. do. 0. grassy sloughs, ox-bows invaded ox-bows, beaded drainage wet holes in muskeg invaded ox-bows, beaded drainage emergent, sedge meadows, brush submerged or emergent, sedge dows 0. grassy sloughs, ox-bows emergent, sedge meadows, brush muskegs brush areas and muskegs do. brush areas brush areas and muskegs do. muskeg wet holes mus brush and muskegs exposed silt under slumps do. Sphagnum fuscum brush and muskegs JUNGERMANNIACEAE Gymnocolea inflata (Huds.) Dumort. wet Carex tufts in muskeg BOG FLATS AND PHYSIOGRAPHIC PROCESSES IN ALASKA 109 SCAPANIACEAE * Scapania sp. VASCULAR PLANTS EQUISETACEAE Equisetum sylvaticum L. Equisetum fluviatile L. Equisetum fluviatile L., f. Linnaeanum Déll) Broun (@ Equisetum palustre ae @ Equisetum scirpoides Equisetum arvense L., var. apete (Bong.) Ledeb. LYCOPODIACEAE * Lycopodium annotinum L., var. alpestre Hartm. PINACEAE Larix peta (Du Roi) K. Koch, ken u * Picea mariana (Mill.) Britton, Stern & Pogg. SPARGANIACEAE Sparganium angustifolium Michx. *(@ Sparganium hyperboreum Laest. Sparganium minimum (Hartm.) Fr. ZOSTERACEAE *(a ee a Balbis, vi baa ) Og var. Nuttallii sont & "Schlecht.) Fernald Potam *( Potamogeton pestulicies 1. var. Richardsonii Benn. * Potamogeton vaginatus Turcz. muskegs wettest bogs, especially old ox-bow sloughs and beaded drainages do. pools or Scirpus areas brush and muskegs muskegs muskeg wettest muskegs and bog ridges rare in dry muskegs or at high altit muskegs and bog ridges pond in m pond in a paren pool in pond in till, spruce forest pool in bog pool in bog bog pond on till in spruce forest clear streams 110 * * * * # & LZ D> JUNCAGINACEAE Triglochin maritimum L. Triglochin palustre Scheuchzeria palustris a var. americana Fernald GRAMINEAE Alopecurus aequalis Sobol., var. natans Lemay Fernald PRESTO arundina ) Beal Agrostis alaskana Hultén Agrostis scabra Willd. variation includes ultén var. scabra (Presl) Hitche ' var. robusta (Vasey) St Ss \ champsia caespitosa (L.) Beauv., Poa sec R. ag ales is (Nash) Batchelder Cheatin pulchella (Nash) K. Schum. Festuca altaica Trin. Q < a CYPERACEAE Eriophorum angustifolium Honckeny (includes var. majus Schulz Eriophorum brachyantherum Trautv. Eriophorum gracile Koch Eriophorum Chamissonis C. A. Meyer, var. albidum Fernald / var. aquatile (Norman) Fernald \ (includes var. spissum Fernald) Scirpus hudsonianus (Michx.) Fernald var. callosus Bigel. ime s saan A {Le & & ieee euaniak (Light) Link, var. Fernaldiana Svenson Eleocharis acicularis (L.) R. & S., Svenson f. rigidula (Reichb.) Rhynchospora alba (L.) Vaht WILLIAM H, DRURY, JR. bog hollows Myrica area bog hollows pond in muskeg muskeg brush do. edges of bogs, ox-bow sloughs variation includes both varieties brook margins muskeg silt of brook through musk silt bottomed ponds, ox-bows do. do. brush, slope general, wetter parts of bogs ull hollows in bogs wet muske; general, wetter parts of bogs variation includes both varieties bog ridges or muskegs wet brush wettest parts of bogs, oftenest in bog hollows, bare silt till ponds wet slope (this ts a considerable range extension British wet Scirpus-Mjrica area (This is nsion from BOG FLATS AND PHYSIOGRAPHIC PROCESSES IN ALASKA 111 (@ gowns simpliciuscula (Wahlenb.) wet slope ackenzie @ on ex gynocrates Wormskj. wet slope on till Carex leptalea Wahlenb. invaded ox-bow @ Carex microglochin Wahlenb. wet slope on till Carex pauciflora Lightf. ridges Carex chordorrhiza Ehrh. fundamental in wettest parts, esp. invad ds *@ Carex bipartita Bellardi till hollow *(@, Carex neurochlaena Holm pools in burned = *@ Carex canescens L. lake in ti arex canescens L., invaded ox-bows _ var. subloliacea Laestad Carex a (Pers.) Poir. old ox-bow variation includes var. piucrostctne (Tuckerm.) Kik. ae in muskeg * Carex disperma Dew ol in muskeg Carex tenuiflora Wahlenb. i mmocks or ridges in bogs Carex loliacea L. pool in fores * Carex Garberi Fernald, pool in forest var. bifaria Fernald *- Carex lugens T. Holm muskeg Carex aquatilis Wahlenb. usually in old ox-bows (@ Carex > ean hp aque. Smith wet slope and strangm Carex limos. fu nda amental in ahh ies and meadow Carex magellanica Lam peas in cael ponds (variation includes C " paupercula Michx. and C. irrigua Wahlenb.) Carex livida (W eet) Willd. bog hollows in standing water @, Carex atrofusca wet slope and till hollows — (@ Carex capillaris L. wet slope Carex lasiocarpa Ehrh., margins ro streams in bogs var. americana Fernald Carex rostrata Stokes wettest arcas variation includes var. utriculata (Boott) Bailey (¢ Carex rostrata membranacea delta of small stream out of flats . onto a silt Carex paludivagans Drury zones of bog between C. rostrata and C. rotundata Carex rotundata Wahlenb. fundamental in sedge meadows Carex physocarpa Presl wet margins of old ox-bows ‘« Carex membranacea Hook. pools in forest ARACEAE Calla palustris L. wettest parts of bog in Sphagnum papéillosum Scirpus areas, ridges g pon sedge meadow, brush wet slopes, till hollow muskeg 112 WILLIAM H, DRURY, JR. JUNCACEAE * Juncus castaneus Smith * Juncus filiformis L. muskes Juncus stygius L., var. americanus Buchenau (Juncus triglumis L. * Luzula rufescens Fisch. LILIACEAE % % % Tofieldia glutinosa (Michx.) Pers. Tofieldia pusilla (Michx.) Pers Fritillaria camschatcensis (L.) Ker-Gawl. IRIDACEAE Iris setosa Pallas ORCHIDACEAE Habenaria hyperborea (L.) R. Br. Spiranthes Romanzoffiana ham Schlecht Malaxis paludosa (L.) Swartz SALICACEAE Salix reticulata L. Salix - L., var. acutifolia Schneider Salix nil abe Sali x a Sarg Salix cteaieles Pallas Salix pulchra Cham. MYRICACEAE Myrica Gale L.. var. tomentosa C. DC. CORYLACEAE Betula nana L., var. sibirica Ledeb. Betula glandulosa Michx. sed aa (Ait.) Pur (includes ssp. sinuata aah Hultén) Alnus tenuifolia Nutt. Scirpus-Myrica areas wet slope, low ground in forest wet slope old ox-bow slough Scirpus-Myrica areas muskeg wet M)rica brush areas, a range extension from southeastern Alaska wet slope muskegs wet slope at forest edge muske ox-bow sloughs, grassy areas, wet brush areas and grassy ox-bow sloughs fundamental in flooded areas, with sparse or no Sphagnum fundamental in Sphagnum-brush areas old ox-bow sloughs m old ox-bows and forest edges do, a ee ee ee Tae me ee ns ee ea ae BOG FLATS AND PHYSIOGRAPHIC SANTALACEAE * Geocaulon lividum (Richards.) Fernald POLYGONACEAE Rumex arcticus Tra Polygonum ae. (Small Wight Polygonum viviparum * * CARYOPHYLLACEAE Stellaria longipes Goldie NYMPHAEACEAE Nuphar polysepalum Engelm. ymphaea tetragona Georgi RANUNCULACEAE * Caltha palustris var. arctica oy Br) Huth var. asarifolia oe H Ranunculus Gm nag *® * *.* Turcz. var. Richardsonii (Gray) Boivin CRUCIFERAE * Cardamine pratensis iL. var. pal Wimm. & Grab. DROSERACEAE Drosera anglica Huds. Drosera rotundifolia L. SAXIFRAGACEAE PROCESSES IN ALASKA muskegs brush areas and muskegs muskeg m brush areas ponds and lakes ponds and lakes on old surfaces wet holes in muskeg variation includes both varieties ponds, common variation includes both varieties muskegs alder slough shore of muskeg lake wet slope brush on edge of forest Calamagrostis grassy slough Scorpidium-Drep. revolvens areas Sphagnum fuscum areas et slope Si eed oles of ane ox-bow sloughs or forest 114 ROSACEAE Spiraea peat Schneider * Rubus arcticus Rubu Potentilla palustris (L.) a (@, Dryas integrifolia M @ Sanguisorba ue LEGUMINOSAE *@ Oxytropis foliolosa Hook. GERANIACEAE * Geranium erianthum DC. EMPETRACEAE Empetrum nigrum var. preter t (Lge.) Sor. VIOLACEAE * Viola epipsila Ledeb var. repens (Ta) W. Bckr. ONAGRACEAE @ goog palustre L. . lapp Epilobior palustre L., var. palustre HIPPURIDACEAE Hippuris vulgaris L. UMBELLIFERAE Cicuta mackenzieana Raup CORNACEAE Cornus suecica L. ERICACEAE Ledum ae L., var. decumbens Ait. Ledum g cum Oeceder Andromeda aaa i Chamaedaphne calyculata (L.) Moench, var. angustifolia (Ait.) Rehder WILLIAM H. DRURY, JR. brush areas and muskeg muskeg Sphagnum fuscum areas rare in the valley bottom ox-bow sloughs and invaded areas till hollows Betula-Myrica areas forest hollows on till brush margins Sphagnum fuscum areas muskeg wet muskeg on till, wet slope variation includes both varieties old, invaded ox-bow pools and lakes invaded ox-bows brush areas “ brush and muskegs egs a meadows, brush, muskegs, - fundamental in low brush areas drier brush and muskegs BOG FLATS AND PHYSIOGRAPHIC PROCESSES IN ALASKA 115 Arctostaphylos rubra (Rehder & Wilson) do. Fernald Vaccinium Oxycoccos L. Sphagnum fuscum areas Vaccinium uliginosum L., general in lowland regions ee ee (variation includes both varieties var. alpinum Bigel. \ Vaccinium Vitis-Idaea L., Sphagnum fuscum areas var. minus PRIMULACEAE (@ Primula egaliksensis Wormskj. wet slope «, Dodecatheon ihe E% Cham. & Schlecht. wet slopes Trientalis europ Sphagnum fuscum ot var. arctica preter Ledeb. S. papillosum areas GENTIANACEAE ( @ Swertia perennis L. wet slope, an extension from the S coast : Menyanthes trifoliata L. general and fundamental in wettest parts of bogs : POLEMONIACEAE | ; z 4 Polemonium acutiflorum Willd. wet brush SCROPHULARIACEAE Pedicularis labradorica Wirsing muskegs Pedicularis sudetica Willd. Scirpus-Myrica bog hollows @ Pedicularis verticillata L. wet brush areas LENTIBULARIACEAE Pinguicula villosa L. Sphagnum fuscum areas Pinguicula vulgaris L — — E . papillosum areas : Utricularia intermedia Hayne rer water of bog hollows : (Myrica areas : Utricularia vulgaris L., deep water of pools var. americana Gray Utricularia minor L. pools or streams through bogs Utricularia ochroleuca Hartm. pure or lakes RUBIACEAE : Galium —e ard sedgy invaded ox-bows Galium trifidum do. ‘ Galium tinctorium do. ; var. subbiflorum Tea Fernald 116 WILLIAM H. DRURY, JR. = ADOXACEAE Adoxa Moschatellina L. muskeg blow-down, silt soil VALERIANACEAE Valeriana capitata Pallas wet slope Valeriana capitata Pallas, muskegs var. bracteosa (Britt.) Hultén COMPOSITAE Solidago multiradiata Ait. muskegs or wet slopes, especially @ Aster Ss Rydb. Re EP ox-bow te wet 5 = -3 = yn ee me a ee Pe wet muskeg hollows do. wet muskegs, especially @ congestus (R. Br.) a wet muskegs var. ee ratus (Ledeb.) Fernald leg ne Satan : rae var. palustris (L.) Fernald \ variation includes both varieties The following species are regarded as characteristic of areas with shallow, shelving, silt shores, i.e. lakes and ponds: Calamagrostis canadensis, var. scabra Calliergon cordifoliu G — var. robusta variation pa Calliergon giganteum Carex physocar Drepanocladus aduneus Salix arbutifolia Sulix Bebbiais Drepanocladus bets $ Sphagnum squarrosum Betula puget Sphagn obtu. Eguisetum fluviati N. conte sth 86 Potumogeton epibydrus, var. Nuttallii Ranunculus Gnielini ximus — Potentilla palustris famogeton granmiineus, v4¥. Wa: g Potamogeton watans Hippuris vulgavis The following species are regarded as characteristic of peat bogs: Cumpjlium stellatum Scirpus cespitosus Drepanocladus revolvens ex limo. Scorpidium scorpioides ivi Hypaum Bambergeri Carex rotunda Spha papillosum aludosa 5 negtam magellanicum Myrica G. um balticu Betula nana, vat. : ene - Drosera anglica Sp Sphagnum Dusenii SS er mr ey a BOG FLATS AND PHYSIOGRAPHIC PROCESSES IN ALASKA 117 Triglochin maritimum rami Polifolia Triglochin palustre cinium uliginosum Scheuchzeria palustris, var. americana ponte trifoliata Eriophorum Chamissonis Pedicularis sudetica Eriophorum angustifolium Utricularia intermedia e following species are regarded as characteristic of the areas under the combination of influences of peat bogs invading silt bottomed ponds: Aulacomnium palustre Carex tenuiflora Sphagnum teres Carex canescens, vat. subloliacea Sphragnum Angstrémii Carex magellanica Sphagnum recurvun Carex es Sphagnum riparium Cicuta mackenzia Sphagnum robustum a ‘aly. ata, Sphagnum fimbriatum var. angus Sphagnum plumulosum Galium pa Carex chordorrhiza Gaulium Brandegei GLOSSARY These isanueaeen are intended only to clarify the words as used in this pape Aapamoor or Aapasuo. German and Finnish terms for the bogs of ne area, chiefly treeless, that occur near treeline and are charac- tically patterned mt peat ridges. Alluvial Lowland. Valley bottoms, of low relief, whose — are stream-borne bier such as gravels, sands and s Association. Plant species which grow together and are coat to e to eliminate each other from the site. As used here, no uniform internal structure is intended such as society or com- munity connotes. Bar Lake. A pond or lake on the floodplain cut off from the river deposition of a silt or sand bar as the stream migrates. Base aris A surface of such low slope that streams flowing across cannot carry a load and further lower the surface by erosion. Beaded Deanna A stream that runs between successive enlarge- ments such as lakes, po marshes or bogs. Bog. A vegetation-filled, water-soaked depression characterized by growth of mosses, chiefly sphagnum. Bog Hollow. The wet, sedgy area between bog ridges. Where best developed they have at their center a mud-bottomed area with Ps gs and Scorpidinm mosses and emergent sedges. See flarke in German and Rimpi in Finnish. 118 WILLIAM H. DRURY, JR. Bog Rid: ve. Ridge of peat moss supporting brush or trees and superimposed on a matrix usually of sedge meadow. The ridges are marrow and elongated italy across slope and may form into net patterns. See Strange in German and Ponnut in Finnish. Braunmoor. German word for one of the bog types in the Orr er classification. This is a bog of salcabebai influence, characte- rized by a mud bottom and certain species of Prebeaiade and Scorpidium and no sphagnum. Bruchmoor. German word for one of the bogs in Cajander’s classi- fication. This is an acid bog where there is peat deposition, charac- terized by tall heath and birch shrubs and scattered trees, usually Pinus silvestris. Depositional Slope. A ccnstructional hillside or slope formed by deposition of coalescing alluvial fans or frost-moved materials from above. They usually are gentle slopes, intermediate between the lowlands and the bedrock hills. Fen. A term used chiefly oe Europeans for a grassy marsh. In Eng- land the term is usuall rved for reed marshes such as Typha, Scirpus or Phragmites. ta ‘Scandinavian Abies it applies to any grassy marsh in contrast to a peaty or h bo Flats. In Alaska, wet peaty alluvial sade ‘ok alternating areas of quaking bogs and Black Spruce forest; for example, the Yukon Flats, Tanana Flats, Nixon Fork Flats, Susitna Flats. This is a A Flarke. The German word for the wet, bog hollow between bog ridges. Ideally in a well developed Strangmoor, these at their cen- ter have a mud-bottomed area with Drepanocladus and Scorpidium mosses and emergent sedges. The term is applied to any wer =e area between bog dies (called Strdénge in German). See Hollow and Rimpi. Graded Slope. Slopes adjusted to the action of the erosion agent. Equivalent to slopes of mature topography in the Davis sense. Streams draining the graded slopes flow in smooth beds without falls or sudden changes of gradient. The minor streams have achie- ved adjustment to the master stream and the master stream to its local Hochmoor. German word for a raised bog, one which has ‘grown lary action in the peat. There is usually one or several very acid ponds, Kole, near the center and around the rim is a sedgy chan- nel where water collects and flows away, the Lagg. BOG FLATS AND PHYSIOGRAPHIC PROCESSES IN ALASKA 119 Indicator Species. Used in one of the European systems of classi- fication; a species of plant restricted to certain physiological re- quirements. An indicator species is limited to and characteristic a certain vegetation association and, for that reason, is used to define that association. Kulju. Term equivalent to bog hollow, rimpi or Flarke. The wet, shallow water or sedgy hollow between bog ridges. This is referred to in Auer’s paper without reference to the language. Lagg. Swedish word for the channel around a raised bog, which car- ries away run-off. Loess. Term taken into English geological usage from German which refers to wind-blown mineral deposits of silt sizes, pulverite. As originally defined the deposits will stand with vertical walls and contain land snails and have other limiting features, but the term in general use has become extended to include the more general meaning of wind-blown silt. Mire. A term used in Scandinavian writings as equivalen: to_ bog. It is taken, presumably, from their word Myr. In English writings this term refers to muddy morasses. Moat. Term introduced in this paper for an area of water or coarse sedges in standing water at the margin of a bog which is actively advancing by thawing (Figure 22). Where the soligenous bogs responsible for thawing merge into bogs with more Hochmoor structure, the moat merges structurally and functionally into the Lagg. Muck. gE used by placer miners in Alaska to refer to any fine- grained overburden over the gold-bearing gravels. In this papet the term is restricted to a generally accepted geological usage as a mixture of fine-grained mineral sediments and finely-divided organic material, usually frozen. ; Muskeg. A swampy area usually moss-floored, characterized chiefly an organic soil. Muskeg most often refers to a Black Spruce forest with a thick mat of mosses (Hypnaceae and sphagnum) underlain by peat, bur can be used loosely to refer to a willow- grown sedgy low place, and in loosest terms is any wet lowland such as a slough or bog. The term is originally American Indian cutting off of a stream meander by the migration of the stream. aking Bog. A carpet of bog vegetation that is floating and sinks and quivers when walked on. Palse. German word for an ice-centered hummock. 120 WILLIAM H. DRURY, JR. Paludification. Von Post's term for the formation and expansion by the rise of water table correlated with changes in drainage of growth of moisture-holding mosses. See swamping and Versumpfung. Permafrost or Socnile Frozen Ground. Any ground remaining frozen over several years, in contrast to Annual Frost which thaws each summer. Rape anyor Term referring loosely to events or areas near or under influence of conditions of climate and geology controlled by an ee heet Pingo. A mound formed by the growth of an ice lens in the soil. As the ice grows it heaves up and eventually bursts open the surface. These are larger than a Palse and may be several hundred yards long. Bin Species. The species of plants which characterize an asso- ciation of plants by their SS and by their bene limited to that type of vegetation. A subjective term used in cribing vegetation. Pounu. The Finnish word for peat-, moss- or bog-ridge. These are ridges of sphagnum moss and other vegetation such as brush or trees superim on a matrix o} ges (us These ridges are narrow and elongated across-slope, string-like. They are the Strange of German. Raised Bog. Hochmoor. A bog that grows into a convex form, usually has several ponds in the center and is surrounded by a Reisermoor. Gumus word for a heath-birch shrub-covered bog in Cajander’s classification. Rimpi. Finnish word for the wet, “tna hollow between bog ridges (Pounn). See Flarke and Bog Hollow Secondary Species. Species which occur singly or ace comand in a plant association and are not used to chara Solifluction. Movement of soil, usually restricted to movement due in which freezing and thawing play a central part. Strang. German word for the string-like ridge of vegetation such as brush or trees growing on a mossy bog ridge on a bog surface. The ridges are narro narrow and usually ed across slope; they may be in nets. These alternate on bog hollows which are sedge meadows or standing water with emergent sedges. See Bog Ridge, Pounn. Strangmoor. A bog on whose surface are patterns, festoons or nets, ee Sy eRe: aes pee 9 Sa Ns the Pathe SS eae Ve am ES DLO M. Oye pele tipraeee he aie odie Ney area gee de Rie cee Rt we Lip SN Reet Be en Nee ce Mt Ee, Seve eee gems A, Tair ho ELE AS Err Mets Td MEE pre Oe AT PL Na i oe ee we SERS gee eS awe ree We eT eae Seams oh 7 a She She ge re cat Ee geht Pea wae RT AY ae ion oy. See es Rin, mn 8 BR SACs ceria Fg a a SS oe ET (Rael BOG FLATS AND PHYSIOGRAPHIC PROCESSES IN ALASKA 121 of vegetation growing on mossy ridges between which are pat- ches of standing water or sedge meadows _ Structured scenes Soils. Soils with regular patterns resulting rting of unconsolidated materials into accumulations of coarse and ‘of fine particles or consisting of frost heaved patches or cracks. Sometmes the soil features are visible to the human eye and sometimes only expressed by the growth of plants, from the German companying or as a result of the growth of moisture conserving mosses. = paludification and Versumpfung. Thaw Slump. The process by which thawing and decrease in volume of Pi silt forms a depression into which soil and vegetation fall. Topogenous. Adjectival term meaning that the source of water for a is the water ning in a place where water has collected in a pre- existing depress Versumpfung. ences word for the process of expansion of bog vegetation at the expense of forest, correlated with accumulation of water in a lowland accompanying or as a resule of the growth of moisture conserving mosses. See Paludification and Swamping. Weissmoor. The German word for a bog with a conspicuous deve- lopment of sphagnum and sedges or low brush in Cajander's classi- fication. Yazoo Stream. Stream of deep water at the edge of a bog. The stream is formed by drainage and by the breaking of the carpet of vege- tation when spring high water floats the bog mat. These are par- tially equivalent to the Lagg of Scandinavian Hochmoors. The standing water or coarse sedges in these marginal streams resemble the moat of thawing bogs but are of a physiographic origin. LITERATURE CITED AnpERSON, E. 1949. — Hybridization. (Iris, pages 1—48). Johns AnpEeRSON, FE. 1953. Ph cS Hybridization. Biological Reviews 28: 280—307 ANpERSON. J. P. 1943—1952. Flora of Alaska and adjacent parts of Camada. Parts 1—9. Iowa ee College Journal of Science. Vols. ere tae 24 and om Axpersson. G. and H. HEssELMAN. 1907. Vegetation oct — (in Cajander (1913)). Meddcl. fr. Statens golaensintiosayg : 7O— sg DREWS, or LeR. 1913. Order Sphagnales. North American Flora 15(1). New York Bot. Garden 122 WILLIAM H. DRURY, JR. mien E. 1932. Alpine zone of Mt. Washington Range. Publ. by the uthor. eta Maine the ER, V. 20. Die die Entstehung der Strange auf den Torfmooren. Acta i i Fennica 12: 23—145. Bearp, J. S. 1944. Climax vegetation in tropical America. Ecology 25: 127—158. BELKNAP, J. se The history of New Hampshire, Vol. III. Boston, 480 pages. Bee, “BO Solar variation as an explanation of climatic change. In Climatic Chane H. Shapley, Ed. Sesilitia Univ. Press, Cambridge, Mass. 3— i pa S2 in. et Interaction of vegetation and soil frost phenomena. Arctic 5(1): 34—4 Berc, L. S. 1950. Natural aah of the U.S.S.R. Macmillan Co.. New York. 6 pages. Brack, G. A., T. DospzHansky and C. Para. 1950. Some attempts to estimate species divetsity and population of trees in ef Seren forests. Bot. Gaz 413—425. BLACK, : . 1950. Permafrost. Ann. Report Smithsonian Inst. 1950: 273—301. BLACK, and W. L. BarRKSDALE. 1949. Oriented lakes of northern Alaska. oe ie 57: 105—118. Bra =. L. 1947. Development of the deciduous forests of eastern North UN, hinntics. Ecol. Monog. 17(2): 212—219. Braun, E. L. 1951. Plant ee ea in relation to the glacial boundary. The Ohio Journ. Sci. 51(3): ; mag Poses i J. 1932. Se Sociology. Translated by G. D. Fuller and H. S. rad. McGraw-Hill, New York. 439 pages Brooks, A. = 1911. The Mt. McKinley region, Aiscke: U.S.G.S8. Prof. Paper 70, 234 pages “sC.- EF: 1949. Climate Through the Ages. McGraw-Hill, New York. Brown, c . 1938. The flora of Pleistocene deposits in the western Florida parishes, West Feliciana Parish, and East Baton Rouge Parish, Louisiana. Contributions to the Pleistocene history of the anise parishes of Louisiana. Louisiana Dept. Conservation, Geol. Bull. 4: 59— Brown. J. S. 1926. The Nixon Fork Country. U.S.G.S. Bull. ee: "97130. frost action, with suggestions on nomenclature. Am. Journ. 244: 622—642 Bryan, K . C. Capy. 1934. The Pleistocene climate of Berm Journ. Sci. (ser. 5} 27: 241—264. 1949. Die Riumliche und Zeitliche Gliederung des Eiszeitklimas. a Die ger apeier 36: 105—112. 133—139 Carn, tistics of iat inca: wed Sects. ie thie 5. 1947. Characte ps Ecol. Monog. 17(2): is ae CaJANDER, A. ~ on Studien tiber die sone Finnlands. Acta Forestalia Fennica 2: 3— : Ciements, F. E Tk Plant Succession. Carnegie Institute of Washington Publ. 242. ts, F. E. 1936. Nature and structure of the climax. Jour. Ecol. 24: 252—284. BOG FLATS AND PHYSIOGRAPHIC PROCESSES IN ALASKA 123 Cooper, W. S. 1926. The fundamentals of vegetational change. Ecology 7: ane a Corram, G., J. T. Curtis and B. W. Hate. 1953. Characteristics of randomly based psaein, Ecology 34(4): 741—757. Cow es, H. 901. The physiographic ecology of Chicago and ag a study of the origin and development of plant societies. Bot. Gaz. 31: 73—10 1 18 CowLes. : Ve 1911. The causes of vegetation cycles. Bot. Gaz. 51: 161-—183. Crampton, C. B. 1912. The geological relations of stable and migratory . pg nat Scot. Bot. Rev. 1912(1): 1— 1951. On the relation between summer temperature and the Cistibution of alpine vegetation in the lowlands of Fennoscandia. Oikos 3(1): 22—52. Darwin, < 1859. The Origin of Species. London Davis, J. H. 1946. The peat deposits of Flock: Geol. Bull. 30. State of Florida, Dept. Conservation, Florida Geol. Surv. Tallahassee, Fla. Davis, W. M. 1909. Geographical Essays. Ginn. and Co. Derver, E. S., Jr. 1949. Biogeography of the Pleistocene. Bull. Geol. Soc. Amer. ashi 1315—1416. Deevey, E. S., Jr. 1953. Pa wes and Climate. In Climatic Change. H. Shapley, or Harvard Univ. Press, Cambridge. pp. 273—318. Denny, C. S$. 1938. Glacial pa of the Black Rock Forest. The Black Rock Forest Bulletin 8: 70 pps. Cornwall-on-the-Hudson, New Yo rk. Denny. C. S. 1940. Stone-rings on New Hampshire mountains. Amer. Journ. Sci. 238: 432—438. NYC. 8. 1951, ima frost action near the pig of the Wisconsin drift in aN aggre The Ohio Journ. Sci. 51(3): 116—125. Du Rierz. 1936. state: Tce: and nomena of vegetation units 1930—1935. Svensk. Bot. Tidskr. 30(3): 580—58' EARDLEY. Unconsolidated sediments and pone features of the lower bas valley. Geol. Soc. Amer. Bull. 49: 303— FeLeMan. J. 1906. Anteckningar under Min Viacke i sheet 3: 228. S ey) my Feanacp. M. L. 1925. Posie of plants . unglaciated areas of Boreal America. Mem. Amer. Acad. of Arts and Sci. 15(3): 239—342. Fut, R. F. 1943. Growth of Nort “ American ice sheet during the Wisconsin age. Geol. Soc. Amer. Bull. 54: 325—362. Fur, R. F. and E. S. Deevey. 1951. Radiocarbon dating of late-Pleistocene go Eurasiens wiihrend des ied ice Eiszeitalter Gegenwart. 2: 154—167. Frey, D. G. 1951. Pollen — ion in the sediments of Singletary Lake, North Carolina. Ecology 32: — 533. Gams. H. 1918. ese bg der Vegetationsforschung. Vierteljahrsschr. d. Naturf. Ges. in Zurich. 63. Gueason, H. 1917. The structure and development of the plant association. Bull. Torr. Bot. Club. 44: 463—481 124 WILLIAM H. DRURY, JR. Greason, Fl AS 1926:" The petindsneery concept of the — association. Contr. N. Ys Bot. Garden No. 279 1. Tor lub. 53: 7—26. Pa H. 1927. Further views on the succession concept. Eeolosy 8: 326. elena H. and F. R. Buarucua. 1932. Studies in the ccology of Wicken Fen. II. The ‘i water table and its ta of plant communities. Journ. Gooptett. J. C. 1954. Vegetation adjacent to the border of the Wisconsin drift in Potter County, Pennsylvania. Harvard Forest Bull. No. 25. Gnices. R. F. 1934, The problem of Arctic vegetation. Journ. Wash. Acad. Hyetr, H. and R. Hutt. 1885. Vegetationen och F loran i en del af Kemi Lappmark och Norra apron Medelanden Soc. pro Fauna at Flora dass 12: 42—43. (in Cajander 1913). Hopkins. D. M. 1949. Thaw lakes and thaw sinks in the Imuruk Lake area, sisi a gace ae wee Journ. Geol. 57: 119—131. Hopkins. f. and . SiGaFoos. 1951. Frost action and bape patterns on Sewuat ei Alaska. Geol. Surv. Bull. 974-C: 51—100. Hopp, J. T. 1951. Local — summary with comparative data, 1950, McGrath. Alaska. U. S$. Weather Bureau. Anchorage, Alaska. Hutt, ‘RK. “1885. Sts vegetation Ett bidrag till vaxformationernas utoecklingshistorie. Med. Soc. Faun. Flor. Fenn. 12: 161. Referred to in Clements. 1936. Hutren, E. i. Flora of Alaska and Yukon. Lunds Univ. Arssk. N.F. Avd. 2, Vol. 37(1)—46 Karz ‘N. }. wee Sphagnum logs of Central Russia: Phytosociology, Ecology and Succession. Journ. Ecol. 14(2): 177—202. KerRNeR VON Maritaun, A. 1863. pe Pflanzenleben der Donaulinder. Wagner, Innsbruck Univ. 2 Lanpsserc, H. 1949. Climatology oe sa ‘Pleistocene. Bull. Geol. Soc. Amer. 69(9): 1437—1442. 1916. power’ fa ge of Minnesota and adjacent districts (distinct). Geol. Soc. Amer. Bull. 27: 68—69. a hoe 1910. The a. gold-placer district, Alaska, with accounts of the Cental Kuskokwim Valley and the Ruby Creek and Gold Hill placers. 15.G.s. 10 Maun, J. e a The cage an ap North cesses acing ena to its amass Esp. Chapter 7, pp. 62—S1; Chapter pp. 120—155. Lawrence, nS F. E. 1930. Geologic history of the Yosemite Valley. U. S. Geol. Mart Surv., ioe Mayr, E. 1954. Change of gi enetic environment and evolution. In Evolution as a Process. Allen and Unwin, London. pp. 157—180. Metron. F. A. 1936. An empirical classification of floodplain streams. Geogr. Rev. 3514) 593—609. and G. HARRINGTON. 1924. The Ruby-Kuskokwim area, Waka eee Bull. 754, 129 pa Merrie. J- 1936. Mine = one Ruby-Kuskokwim region, Alaska. USGS. Bull. 864e: 115—255. PET ee Se BOG FLATS AND PHYSIOGRAPHIC PROCESSES IN ALASKA 125 von Mippenporr. A. 1864. - Ubersicht der Nord- und Ost-Sibiriens, Sibirische Reise 4(1). St. Petersburg. futterR. S. W. 1947. Permafrost or permanently frozen =“ and related engincctier ee J. W. Edwards, Ann Arbor. Norruin, J. P. Berattlese 1 anledning af en n till ge verkstalld naturalhistorisk resa. Not. ur Sillsk. pro Fauna et Flora Fennica 13: 256—257. (in Cajander, 1913). Ouson, J. S. 1952. Historical changes in Lake Michigan dune — Paper ping to the Ecological Society of America at AIBS Meeti Cornell, Sept. 8, 1952. Otson, J. S. 1952. Some vegetation-soil relations of Lake Michigan dunes Paper presented to the Ecological Society of America at AIBS Meetings, t _ ioe) ~I ke Cornell, Sep 1952. Oostinc. H. J. and W. D. Bituincs. 1951. A comparison of god spruce-fir forest in the northern and southern Aypalachiaa system. Ecology 31: 84—103, esp. page Osvacp, H. 1923. Die Vegetation des Hochmoores Komosse. Svenska Viaxtsoc. Sallsk. Handl. I. Akad. Abhandl. Pearsatt, W. H. 1938. The soil complex in relation to plant communities. Ill. Moorlands aa Bogs. Journ. Ecol. 26: 298—313. PEARSALI 950. Mountains and Moorlands. The New Naturalist Ng “callin, Sogia 312 pages. PéweE, 1948. Terrain and permafrost of the Galena air base. Galena. aa : '$.G-S, Permafrost Progress Report No. 7 (Restricted). Péwt. T. L. 1950. Pag of the upland silt in the Fairbanks area, Alaska. . 61: 1942 u j PeEwk, Ti. 195%. "Multiple nies in Alaska. U. S. Geol. Survey Circular 13 pag Pirecka. F. 1941. Distribution = pai in relation to major biotic communities. The Amer. Midland Natural. —137. PLeske. T. 1928. Birds of the Ro. tundra. Mem. Boston Soc. Nat. Hist. Potunix. N. 1934—35. The vegetation of Akpatok Island. Journ. Ecol. 22: 337—395; 23: 161—209. bee N. 1947. Botany of the Canadian waeroigg Arctic. Part HI. Thalle- phyta and ee ease sees Mus. Bull. No. 97, Biol. Ser. No. P weie PM. of the Canadian Eastern Anti Part IIT. Veacintica and ecology. me pra Cn = 104. Bio. von Posr, L. 1937. The geographical survey os insh bogs. The Irish tle: 227. Porzcer, J. E. and B. C. Tuarp. 1943. Pollen record of Canadian spruce and fi e 98: 4. Porzcer. J. E. and B. C. Tuarp. 1954. Pollen study of two bogs in Texas. —466 Raprortu. N. W. 1952. Suguested classification of muskeg for the engineer. Engin. — Nov. 1952, 12 page: © igs cee Lapin i since eeigptig Dea Cajander 1913). H. 1946. Phytogeographic core in Athabasca-Great Slave iia I. ‘Fowets Journ. Arnold Arb. 27: >. 126 WILLIAM H. DRURY, JR. Rave. H. M. 1947. The botany of southwestern Mackenzie. Sargentia VI: Pp. Raup, H. M. 1951. Vegetation and cryoplanation. The Ohio Journ. Sci. Ricuarps, T. W. 1952. ey tropical rain forest. Cambridge Univ. Press 50 p. Scuimper. A. F. 1898. ‘ohicea ese auf Physiologischer Grundlage. esp. - pages 173—176. Jena. Translated into English and published by the Clarendon SHaLer, N. S. 1890. General account of the freshwater sige of the United States. age gg Dir. U.S.G.S. 10: 225—339 (esp. 283—285). Sureve, F. 1942. e desert vegetation of North America. Bot. Rev. 8: 1 24 Stcaroos, R. S. 1951. Soil instability in tundra vegetation. The Ohio Journ. 8 SicaFoos, R. S. 1952. Frost ge as a —_ physical factor in tundra plant communities. Ecology 33(4): Smpson, G. 1937. Ice ages. Proc. = oe 30: 125—142. Syjors, H. 1948. piensa I Bergslagen. Acta Phytogeogr. Suecica 21: English Summary pages 277—299. SkENE, MacG. 1915. The He of Sphagnum and its relation to chalk and mineral salts. Ann. Bot. 29: Saarn, H. T. U. 1949. Physical effects of Pleistocene climatic changes in non-glaciated areas: eolian phenomena, frost action, and stream terracing. Amer. 60(9): 516. ~~ : - 1939. fees! geology of Alaska. U.S.G.S. Prof. Paper 192, SPURR, es . 1898. el reconnaissance in southwestern Alaska. U.S.G.S. 20th nn. Report. Part VII: 31—264. Steere, W. C.. Jr. ip Musci. in Polunin. N. Botany of the Canadian Eastern Arctic, Part II. Can. Nat. Mus. Bull. 97: 377—382. ES E. G._ 1949. Identification and evaluation of Alaskan vegetation — with reference = soil moisture and permafrost ¢ — Dept. of the Army. Corps of Engineers, St. Paul Dist., Printed ose Fai ES. y- rT, 2 eee distribution and soils in the Harvard Forest. ‘Sarard Forest Bull. No. 24. Taser, S. 1929. Frost ioe Journ. Geol. 37: 428—446. Taper, S. 1943. Perennially — —_ in Alaska: its origin and history. Bull. Geol. Soc. Amer. 54: 1433 Tanstey, A. G. 1935. The use and se of vegetational concepts and terms. Ecology 16: 284—307. Tanstey, A. G. 1939. The sage Islands and their vegetation. Cambridge Univ. Press, LES XXXVili 30 pages. Tanttu, A. a cc ae Biilten und Strange der Moore. Acta oak Fee 4: 14—23. (in Cajander 1913). Tuoreav, H. D. 1860. Succession of forest trees. Mass. Board of Agriculture, Report VIII. ee SP nr 3 Sa emcainimaene manic aerate ips rmmanaiben: i BOG FLATS AND PHYSIOGRAPHIC PROCESSES IN ALASKA i27 Trout, C. 1944. Strukturboden, Solifluktion und Frostklimate der Erde. Geol. Rundschau 34: 545—695. (esp. 639—640 and Figs. 58 and 59). Tucx, R. 1940. Origin of the muck deposits at Fairbanks, Alaska. Geol. Soc. Amer. 51: 1295—1310 ime J. B. 1910. Changes in ene in yelaniganege Canada since the glacial period. In .,Die hire Mig des Klimas = den Maximum der letzten Fiszeit.” 11th Inter. Geol. Co ongress, Seah , 1910: 389—391. Upoam, W. 1895. ee of glacial Babaciaee and invasion. Geol. Soc. Amer. Bull. 6: 343—352 Waxvace, R. E. 1948. Coed -in lakes in the Nabesna, hi ere and Tanana River vallies, eastern Alaska. Journ. Geol. 56: 171—18 WarMinc, E. 1909 (1895). ( persen en ae “of plants. Oxford Univ. Press, London. xi — WarREN-Witson, J. 1952. bei: ac yee ee with soil movement on Jan Mayen Island. Journ. Ecol. 40(2): Watson, W. mbes Sphagna, their eh ae Pest and associates. Ann. Bot. 32: 535—55 Wart, A. S. Ar. "een and process in the plant community. Journ. Ecol. 35(1 and 2): 22, Weaver, J. E. ay F. E. CLementrs. 1938. Plant Ecology. McGraw-Hill, — 601 pages Wennern, C. 1947. Pollen diagrams from Labrador. Geografiska Annaler WurtEHEAD, A. N. 1925. Science and the Modern World. Macmillan, 2ES. Wiutterr, H. C. 1953. Atmospheric and oceanic lace as factors glacial-interglacial changes in eet In Climatic Change. H. Shapley, Ed. Harvard Univ. Press. pp. 51 INDEX Advanced linear phase, Phase II 16 Advanced linear phase, Vegetation 32 8 Area of study, ea on Aspen-lined 53 Associations of shi: Descriptive method 38 Auer, V. 6, 60—72 Bar lake 17 Base level in bogs 28 Bed rock 8 Black spruce clumps 6213 Black spruce forest 14, 33 26 iB cyc Bog effects on exploitation of lowlands 77 Bog — 38 ollow 46—52, 58—59, 60 Bog origin pe actin 71—74 Bog origin 1 Bog origin, Colonization of water 19 Bog o origin, Swamping 20 Bog origin, Thawing and slump 20 14, bag 60 Bog, Patterned surfaces Bog ridge Boe ridge, Origin and dele 7i—74 Scandinavian classifications = Fg size in Alaska Bog vegetation 36—38, - 13, a Bog vegetation, Scandina 54—61 Bogs, Descrip 5, 13 Bogs, Distribution Circumboreal 5 upper tr Se a , Vegetation = rane Surfaces 52, 60—61 Bylot Is 66, 71—74 Cajander, A. K a; Clements, F. E. 6, 9 — of the upper et tice on hd iver regi lnox concept Climax poi Historical position i Climax fores 32 Clumps of vegetation regularly : tuted dis 62, 75 Coalese hase, Phase II 17 Coslesen phase, Vegetation ae Cowles, H. C. 92, 97 Descri of vegetation 30—54 oubea method used for plant ns 3 ciatio Exposures in river bank 13 Exposures in river bank, Under g 25 Flats, Described and defined 2 looded bo 14, 45 Floodplain Seagg 16, 17, 30—34 Forest climina 7 orest types, ee of area 10—11 Forestation o| 5, 38 leason. H. A. 93 Graded slopes i in bogs 28 Growing 10 High brush 14, 45, 60 Ice sheet. oo and expansion 85 Indicator 5; 55 Knick int 29 Linear phase. Phase I 16 Linear phase, ae: Low brush , Aa, 99 Margins of centers 47—52 McGrath 8 Migration of plants. Post Pleistocene 81, 101, 102 Mineral soil in bogs 13, 24 Moat i, 44, 58 Muck 6, 23—-28, 78 Ombrogenous bog Oxbow 53 Patterned bog surfaces Patterned bog surfaces, Vegetation 45—52 ——. = nergee Comparison th Scandinavia Peat’ R Ridge 45—52, 60—6l1 Peat Ridge, On lake shore pa Persistence re vegetation during Pleistocene 79—80, 88—90 Periglacial climates 80, 84, 86 130 Permafrost table 16—18 Péwé, Classification of floodplain 16 Physical forces controlling 2 _ vegetation patterns 71—74 Pot holes 7 Quaking bog 1214 Quarternary alluvium 9 Scalloped phase, Phase I'V 17 Scalloped phase, Vegetation 34 eae ate studies of bog vegetation . 54—61 Sedge meadow 14, 44, 47, 58—59 Shallow water and emergent aquatics 14, 44, 47, rg A Sjér's studies on bogs 5 Sloping bogs 17—18, a Sohju_ theo: Soil instability 1s. 78.Abe ligenous mires 23 Soligenous 56 Speciation in ref 82 no 99, Sphagnum, Meaning er presence 62, 3 Sphagnum, Succession 36 Sphagnum, arance in the floodplain forest 33 Sphagnum. In the swamping of the floodplain INDEX Sphagnum, Superposition — 46, 60 ang moor 46, 50. 60—62 seine sequences, Effects of nd soligenous waters 77—78 Structured soils. Conditions for development 82—84, 102 Successional sequence, In bogs 6,.35; cycles a 36, 58—60, 68 Successional sequence, = floodplain O—35, pl. 12 Swamping of low le 2 Tansley, A. G. 91—94 eas of continental ice poe Topogenous bogs 56 Transformation of bog basin cui Tree line Triangular lakes Tundra zone south of continental ice shee s80—81, 90 Vepssine description 10—11, 35—-38 bie tis description, of wet 43—52, 116—117 Ph sea POE Ete Uniformity iP centers . 58—60 ue spruce forest 32—33 White spruce — mixed forest 9, 32—33 Neer eee ee eee ee Re Stas Gila ck A saa akg agbhdac Mew epe a AN pa Sect EE