Shrub Bog Research Articles

Seasonal patterns and controls on net ecosystem CO2 exchange in a boreal peatland complex

We measured seasonal patterns of net ecosystem exchange (NEE) of CO2 in a diverse peatland complex underlain by discontinuous permafrost in northern Manitoba, Canada, as part of the Boreal Ecosystems Atmosphere Study (BOREAS). Study sites spanned the full range of peatland trophic and moisture gradients found in boreal environments from bog (pH 3.9) to rich fen (pH 7.2). During midseason (July‐August, 1996), highest rates of NEE and respiration followed the trophic sequence of bog (5.4 to −3.9 μmol CO2 m−2 s−1) < poor fen (6.3 to −6.5 μmol CO2 m−2 s−1) < intermediate fen (10.5 to −7.8 μmol CO2 m−2 s−1) < rich fen (14.9 to −8.7 μmol CO2m−2 s−1). The sequence changed during spring (May‐June) and fall (September‐October) when ericaceous shrub (e.g., Chamaedaphne calyculata) bogs and sedge (Carex spp.) communities in poor to intermediate fens had higher maximum CO2 fixation rates than deciduous shrub‐dominated (Salix spp. and Betula spp.) rich fens. Timing of snowmelt and differential rates of peat surface thaw in microtopographic hummocks and hollows controlled the onset of carbon uptake in spring. Maximum photosynthesis and respiration were closely correlated throughout the growing season with a ratio of approximately 1/3 ecosystem respiration to maximum carbon uptake at all sites across the trophic gradient. Soil temperatures above the water table and timing of surface thaw and freeze‐up in the spring and fall were more important to net CO2 exchange than deep soil warming. This close coupling of maximum CO2 uptake and respiration to easily measurable variables, such as trophic status, peat temperature, and water table, will improve models of wetland carbon exchange. Although trophic status, aboveground net primary productivity, and surface temperatures were more important than water level in predicting respiration on a daily basis, the mean position of the water table was a good predictor (r2 = 0.63) of mean respiration rates across the range of plant community and moisture gradients. Q10 values ranged from 3.0 to 4.1 from bog to rich fen, but when normalized by above ground vascular plant biomass, the Q10 for all sites was 3.3.

Oxygen-18 Composition of Sphagnum, and Microenvironmental Water Relations

The 180 isotope content of Jive species of Sphagnum from three peatlands in Ontario was analyzed. Sphagnum fuscum and S. capillifolium had significantly more enriched 6'80 values than S. fallax, S. magellanicum, and S. papillosum. The enriched values in the former species indicate that water involved in photosynthesis in hummock species is much more affected by evapotranspiration than in adjacent species situated in depressions. These data suggest that microenvironmental water relations regulate evapotranspiration in Sphagnum species. Environmental isotope analysis (180, 2H) is another technique that may shed light on understanding Sphagnum ecophysiology and water relations along microenvironmental gradients in northern peatlands. The vertical distribution of Sphagnum species along microtopographic gradients has long been noted, indeed, ever since the earliest ecological studies of northern bogs and poor fens (Andrus et al. 1983; Clymo & Hayward 1982; Sj6rs 1948; Vitt et al. 1975). The genus Sphagnum represents a unique group of hydrophytes especially adapted to periodic drought conditions because of large seasonal changes in water supplies reaching peatland surfaces. Ecological factors such as cation exchange capacity, desiccation resistance, water holding capacity, degree of drying, and photosynthetic ability have been found to be important for Sphagnum microenvironmental distribution patterns (Andrus 1986; Clymo & Hayward 1982; Lindholm 1990). Environmental isotopes--oxygen18 (180) and deuterium (2H)-that form part of the water molecules and organic compounds in plant organic matter, have been widely used in ecology and geochemistry (Fritz & Fontes 1980; Rundel et al. 1989). One specific application is in the reconstruction of paleoclimate based on the isotopic composition of fossil wood and peat cellulose (Dupont & Brenninkmeijer 1984; Edwards et al. 1985; Yapp & Epstein 1982). Paleoclimatic inferences are based on the relationship between the 180 and 2H composition of plant cellulose and the isotopic composition of the plant water source and relative humidity (Burk & Stuiver 1981; DeNiro & Epstein 1979; Epstein et al. 1977). We report some results of a project aimed at assessing the relationship between the isotopic composition of bog plant species and environmental factors. The understanding of this relationship is essential if stable isotopes in fossil peat plant cellulose are to be used in paleoclimate studies. This paper proposes that isotopic analysis of Sphagnum mosses may be an important aid in understanding the ecophysiological requirements and microenvironmental distribution of sphagna. MATERIAL AND METHODS Field methods.--Mosses for this study were collected from three peatlands in Ontario. Oliver's Bog is a floating bog and fen complex (sensu Tarnocai et al. 1988) developed in a closed kettle basin 4 km southeast of Cambridge in southwestern Ontario (43019'45N; 80017'15W). An open bog and poor fen mat has a soft bottom closest to the open water area in the centre and is firmly anchored in the treed bog area around the edge of the basin. The mat portion, where mosses were collected for this study, is characterized by oligotrophic lawn and carpet vegetation with some areas periodically in contact with waters of the underlying lake system, and having small ombrotrophic hummocks. Emo Bog is a large ombrotrophic raised bog complex, about 680 ha in size, situated 8 km northeast of Emo in northwestern Ontario, near the Minnesota-Ontario border (48042'N; 93047'W). It is a forested-low shrub bog with Picea mariana, numerous ericaceous shrubs, and welldeveloped hummock microtopography. Rainy River Bog is 40 km west of Emo Bog and 6 km north of the Town of Rainy River (48047'30N; 94034'W). This ombrotrophic bog is developed in a small basin about 10 ha in size with a vegetation cover and microtopography similar to that at Emo Bog. Laboratory procedures. -Representative Sphagnum mosses were collected at each study site for isotope analyses in June or July 1987. Samples that included several moss stems were selected from adjacent paired microsites-hummock top and a position near the bottom of the side of the hummock. Sphagnum nomenclature follows Anderson (1990). Voucher specimens are deposited in the Wetlands Ecology Laboratory at the University of Waterloo. Cellulose was extracted from the living plants 0007-2745/92/445-448$0.55/0 This content downloaded from 157.55.39.54 on Fri, 01 Jul 2016 05:38:50 UTC All use subject to http://about.jstor.org/terms 446 THE BRYOLOGIST [VOL. 95 TABLE 1. Isotopic composition of Sphagnum from three peatlands in Ontario. The sample of S. magellanicum from Emo bog may contain some Polytrichum strictum. Collection sites: OB = Oliver's Bog, a quaking bog, southwestern Ontario; RRB = Rainy River Bog, a raised basin bog, northwestern Ontario; EB = Emo Bog, a raised basin bog, northwestern Ontario; SMOW = standard mean ocean water.

Phosphorus Availability in Acid Organic Soils of the Lower North Carolina Coastal Plain

A natural landscape gradient of P—deficient pocosin (evergreen shrub bog) and bay forest ecosystems was used to test three hypotheses concerning P availability and P reabsorption efficiency: (1) that variations in vegetation height and aboveground biomass among even—aged pocosin stands are related to variations in soil P availability; (2) that microbial processes control P availability in these soils; and (3) that the proportion of P reabsorbed from senescing foliage in native plants is inversely related to soil P availability. The natural gradient from short pocosin to bay forest was found to represent a gradient of increasing P availability, as indicated by increases in: (1) total P in surface (0°15 cm) soils, from 19.6 to 118.3 kg/ha; (2) seasonal and annual PO43— supply to in situ anion—exchange resins; and (3) P concentrations in native plant foliage. Increases in P availability were accompanied by a significant decline in the N:P ratios of both plant leaves and surface soils. Microbial processes strongly influenced the supply of available P. Soil microorganisms immobilized up to 90% of added 32PO43— during laboratory incubations, but when microbial uptake was inhibited, as much as 90% of added 32PO43— remained in the available pool. A decline in the ratio of microbial: total P in surface soils, from 56.3 to 37.4%, and an increase in the ratio of net: gross P mineralization (estimated by isotope dilution), suggested that as P availability increased, microbial control decreased. Native plants reabsorbed both P and N efficiently from senescing foliage, reaching maxima of 86.6 and 80.5%, respectively; natural variations in soil P availability did not affect the proportion of either nutrient reabsorbed. In general, however, plants tended to reabsorb greater proportions of foliar P than N. Natural variations in P availability, the factors controlling P availability in these soils, and the potential for differential nutrient reabsorption by native plants have important implications for the cycling and accumulation of P and N in pocosin and bay forest soils. Landscape—Level processes may have contributed to the observed variations in vegetation and P availability along this natural gradient.

Pocosins: An ecological perspective

Pocosins and associated wetlands (PAAWS) of the North Carolina Plain are among the least studied ecosystems in the U.S. The purpose of this overview is to (1) classify pocosins and give their geographical location, (2) provide information on their geological origin, and give an ecological analysis of these ecosystems, and (3) analyze development trends. Recent surveys have shown that pocosins occur on the southeastern coastal plain from Virginia to north Florida and once covered more than one million hectares in North Carolina. A broad definition of pocosins (sensu lato) would include all shrub and forested bogs, as well as Atlantic white cedar stands and some loblolly pine stands on flooded soils on the Coastal Plain. A stricter definition (sensu stricto) of pocosin for delineation purposes would only include the classic shrub-scrub (short pocosin) and pond-pine-dominated tall pocosin. Common synonyms for pocosin, including bay, bayland, bayhead, xeric shrub bog, and evergreen shrub bog, further confuse what is and is not classified as a pocosin. The advanced identification and delincation of pocosins from upland and other associated wetland types will require detailed field analyses following the stricter definition of pocosins. Over 51% of the forested palustrine wetlands in North Carolina have been disturbed, and approximately 33% of pocosins have been destroyed. The 1991 U.S. Fish and Wildlife Service data follow the general trends for pocosin loss reported by Richardson in 1981. The high rate of wetland modification, especially for pocosin-type palustrine wetlands, suggests that stricter wetland laws are needed if we are to follow the concept of “no net loss of wetlands.”

The effect of nitrogen and phosphorus addition on the growth, reproduction, and nutrient dynamics of two ericaceous shrubs

Aboveground growth, reproduction, and foliar nitrogen and phosphorus contents of two ericaceous shrub species were compared over two seasons in (a) an undisturbed shrub bog (pocosin), and (b) a factorial fertilization design in which three levels each of nitrogen and phosphorus were added in all possible combinations. One species, Zenobia pulverulenta, is deciduous whereas the other species, Lyonia lucida, is evergreen. In the nutrient-poor undisturbed pocosin the two species exhibited similar foliar nitrogen and phosphorus concentrations and aboveground growth rates. Neither species flowered. In response to nutrient-addition Zenobia increased growth rates more than Lyonia. Foliar phosphorus concentrations of both species increased in response to enhanced phosphorus availability. in the first season neither species flowered in any treatment. In the second season Zenobia flowered only in the fertilized plots, with the most flowering in the high phosphorus treatments. I conclude that, by virtue of high growth rates and efficient use of nutrients and despite differences in leaf phenology and morphology, both Lyonia and Zenobia are successful in a competitive community under conditions of extremely low phosphorus availability. However, unlike Lyonia, Zenobia can take advantage of temporarily increased nutrient availability, which occurs following fire in the pocosin, to increase growth and reproduction.

Bogs as Beaver Habitat in North-Central Minnesota

In a 100-km2 area surveyed for beaver activity, one-third of the active colonies were in bogs and two-thirds were in lakes and rivers. From 1979 to 1981, the percent of colonies in bogs increased from 29% to 36%. Of 481 bogs in the study area, 200 (42%) had current or previous history of beaver activity (colony sites and work areas). At 101 bogs (21 %), dam construction in seepage zones created flowages. Most impoundments were built at minerotrophic sites. Moats were used at 99 bogs (21 %), where beavers preferred sedge-moss cover and avoided tall shrub and wooded cover. The results show that bogs are suitable habitat for beavers. INTRODUCTION The suitability of lake and stream habitat for beavers (Castor canadensis Kuhl.) has been well-studied (Retzer et al., 1956; Northcott, 1964, Tufts, 1967). Although beavers frequently colonize lakes and streams that are adjacent to bog areas, few researchers have mentioned that beavers are capable of colonizing bogs that lack open water. Northcott (1964) reported a single instance of a colony in a closed bog, but reference by others to similar phenomena has been anecdotal (Schwintzer and Williams, 1974; Buell and Buell, 1975; Worley, 1981). Marshall and Miguelle (1978) in their review of wildlife on Minnesota peatlands mentioned only that beavers colonize bogs that have existing open water. Although there are few accounts of beavers colonizing bogs, colonies in bogs are locally common in N-central Minnesota, and in some areas they are more prevalent than in lakes or streams. Beavers are well-adapted to aquatic habitats, but their ability to create open water in closed bogs is an unusual situation that may involve specialized behavior. The purpose of this paper is to document the extent of bog use by beavers in a section of Cass Co., Minnesota, and to determine what biological and physical features of bogs are associated with use by beavers. METHODS AND PROCEDURE The 100-km2 study area, located E of Backus in Cass Co., included most of Deerfield Township (T139N R31W) and a few adjacent sections of Hiram, Bull Moose, Powers and Birch Lake townships. The topography is characterized by a rugged terminal moraine belt with ca. 250 small lakes and 500 bogs in depressions between hills. The upland forest is predominantly paper birch (Betula papyrifera) and trembling aspen (Populus trernuloides). I used aerial photographs to detect the 512 bogs larger than 0.12 ha in the study area. Bog areas adjacent to lakes and streams were not included unless the connection was minimal and the bog occupied a separate basin. In the field, 481 bogs were evaluated for the presence of beaver activity and classified according to predominant cover type as sedge-moss, tall shrub or wooded bog. Sedge-moss bogs were characterized by floating mats of sedges (Carex) and/or Spagnum moss and ericaceous shrubs covering most of the basin. Tall shrub bogs were dominated by alder (Alnus rugosa) (the brushy fen of Boelter and Verry, 1977), or by Betula pumila, Larix laricina and Picea mariana -the early seral stages of bog forest on ombrotrophic sites. Wooded bogs were dominated by L. laricina, P mariana and occasionally by trees from later stages. Most bogs in the study area were less than 3 ha and had either monotypic vegetation or predominant cover types that were easily recognizable. Aerial and ground surveys were used to detect both colony sites and work areas at 'Present address: Department of Botany, Louisiana State University, Baton Rouge 70803.

Sphagnum Vegetation of the Low Shrub Bogs of Northern New Jersey and Adjacent New York

KARLIN, E. F. (Sch. of Theor. and Applied Science, Ramapo State College of NJ, Mahwah, NJ 07430) AND R. E. ANDRUS (Dept. Biological Sciences, SUNY at Binghamtom, Binghamton, NY 13901). Sphagnum vegetation of the low shrub bogs of northern New Jersey and adjacent New York. Bull. Torrey Bot. Club 113:281-287. 1986.-Twenty-five species of Sphagnum were found in fourteen low shrub bogs located in northern NewJersey and adjacent New York. Sphagnum bartlettianum, S. cuspidatum, S. fallax, S. magellanicum, S. recurvum and S. rubellum were the most prominent species. Three new state records were established: Sphagnum bartlettianum in New York and S. flexuosum and S. fuscum in New Jersey. Range extensions within New York State were also found for four other species (Sphagnum flavicomans, S. majus, S. platyphyllum and S. riparium). The S. majus and S. riparium collections represent the southernmost sites reported for North America.

Impact of right-of-way construction on vegetation in the Red Lake Peatland, northern Minnesota

During the winters of 1978–79 and 1979–80, a 500-kV electrical transmission right-of-way (r-o-w) was constructed across the Red Lake Peatland in northwestern Minnesota, the largest contiguous peatland in the lower 48 states of the USA. Immediately before, and for two years following construction, vascular vegetation was monitored within the r-o-w and in undisturbed control areas. Monitoring was carried out in five vegetation types: a thicket swamp, a low shrub bog, a graminoid fen, a treed bog, and a treed fen. Evaluation of construction impacts was based on vegetation structure, irrespective of species composition, and on community composition (species data for low shrubs and herbs). Construction eliminated trees from the r-o-w. Vegetation structure, excluding trees, was markedly altered in the two bog types and the treed fen type in the first postconstruction growing season. By the second season, measurable recovery to control levels had begun. The sample plots were placed into a previously developed vegetation classification system for the Red Lake Peatland, on the basis of herbs and low shrubs. There was a shift in composition in the low shrub bog and in the treed vegetation types following construction. Results of both methods of data analysis were consistent. Major vegetation changes following construction occurred in the low shrub bog and treed types. The open, low-stature fen types showed almost no changes related to construction. Even in the affected types, all vegetation strata except trees were returning to their preconstruction status by the second growing season following construction.

Sedimentation and sediment accretion in Michigan coastal wetlands (U.S.A.)

Sediment transport and accretion rates were studied for three different types of coastal wetlands of Lake Michigan, U.S.A. A forested intradunal wetland (Cranberry Bog No. 1) near Stevensville, Michigan, displayed sediments in the form of detritus, and exported them. Very little sedimentary material entered by streamflow. These features of bottom-land wetlands were confirmed by radioisotope studies on cores, which showed low accretion rates of 210Pb. The mass fluxes were found to range from 0.01 to 0.2 g cm −2 yr. −1. The equivalent linear growths are 0.5–9.0 mm yr. −1. The suspended material from this wetland entered a filling shrub—bog—lake complex known as Cranberry Bog No. 2. A considerable fraction of the incoming sediments were retained in this intradunal basin wetland. This was confirmed via radioisotope studies, with cores showing higher accretion rates than in the bottom-land wetland. The river-delta wetland at Pentwater, Michigan, had little sediment trapping ability. Large amounts of water and suspended materials entered this riverine wetland, but no annual average removal was found. Radioisotope studies of cores at selected locations within this system showed little or no accretion in stream channels, and normal accumulation of organic material in non-channel areas. No net annual effect on water quality of the streams as they discharged to Lake Michigan was found. Storm events caused no notable import or export of suspended material. Presumably the sand substrate in the vicinity of all three wetlands caused runoff events to be mediated by subsurface flow through this porous soil. These three sand-based coastal wetlands appear to have little net effect on suspended solids. The 210Pb method proved more valuable than 137Cs because of the small depth of occurrence of the latter isotope. Interpretation of isotope profiles required information on acid-fraction-soluble profiles. Results are expressed in terms of mass units, since compaction of cores distorts the linear accretion rate. Delayed transfers of radionuclides appear to have occurred, perhaps as a result of integration of atmospheric inputs by the watersheds. The total activities of 210Pb and 137Cs are well correlated, which indicates particle trapping efficiency as the primary factor for affecting accumulation of both isotopes.

Water Analysis of Habitats of Sphagnum in the Coastal Plain of North and South Carolina

The habitats of approximately 20 species of Sphagnum in the Coastal Plain of North and South Carolina are briefly described. Water samples from a range of specific habitats showed generally low pH values (3.5-5.6) and low spe- cific conductivity (18 to 118 pmho). Data relating species presence to height above the local water table confirm the widespread observation that Sphagnum species tend to be narrowly distributed along a gradient from strictly aquatic to variously less aquatic species. Coastal Plain habitats, such as pocosins and savannahs, differ sufficiently from more northern bogs and fens to continue the use of separate names for them. There are approximately 20 species of Sphagnum in the Coastal Plain of North and South Carolina. Including varieties, the total number of taxa is about 25. Although most of these taxa were illustrated and discussed briefly by Blomquist (1938) following the taxonomy of Andrews (1913), essentially no information is available about the ecological relationships of Coastal Plain sphagna. Extensive observations and collecting in the field, often aided by Sphagnum specialists from other regions, left the distinct impression that the Coastal Plain species of Sphagnum have definite habitat preferences. Following the general habitat classification of Wells (1928, 1942), a preliminary description of these habitats and their associated species, based on field notes and presence lists, was given by Lane (1979), and is summarized briefly below. Sphagnum does not occur commonly in such habitats as pine flatwoods, oak-hickory- pine forest, wire-grass sandhills, or in any seaside habitats, including maritime forests and salt marshes. Sphagnum is generally common in swamps, especially cypress (Taxodium distichum (L.) Rich. and T. ascendens Brognm.) and gum Nyssa sylvatica Marsh., N. biflora Walt., and N. aquatica L.) swamps along creeks or lake shores that may include Atlantic white cedar (Chamaecyparis thyoides (L.) BSP) as well. In these swamps two varieties of S. macrophyllum often occur floating in deep pools. Sphagnum forms a sig- nificant portion of wet evergreen shrub bogs, or pocosins, and grass-sedge bogs, also known as savannahs. Pocosins, which can be quite variable as to both the species of dominant shrubs and individual site conditions, often contain members of the Sphagnum section of the genus, such as S. perichaetiale, S. imbricatum, S. henryense and S. ma- gellanicum, as well as other species. Drainage ditches through pocosins or along roadsides often provide disturbed habitats for these species. The habitat with the most unusual assemblage of species is the savannah. Savannahs often result from repeated fires that eliminate the shrub layer of pocosins (Christensen 1977) leaving scattered trees (Pinus serotina Michx. or P. palustris Mill.) over a diverse mixture of herbs and wire grass

Blade Tissue Organization of Foliage Leaves of Some Carolina Shrub-Bog Species as Compared with Their Appalachian Mountain Affinities

1. Leaf blades of twenty-seven woody species of North Carolina have been studied. Leaves of fourteen species were collected from coastal shrub bogs; two, from coastal swamp forests; and eleven, from mesic forest habitats of the Appalachian mountain slopes. The selection of species was based upon a list prepared by B. W. Wells, showing how the shrubbog species have arisen: by geographical isolation from the Appalachians; by northern migration from the tropics; or by southern extension of boreal forms. Thus, the mountain species chosen are those that have Carolina coastal affinities. Conclusions are based upon comparative descriptive and statistical data from the group analyses as well as from data of the eight individual species pairs. 2. The leaves from both the shrub-bog and the mountain mesic forest collections are elliptical or modified elliptical, but the shrub-bog leaves are more perfectly elliptical. 3. The shrub-bog leaves are thick and, for the most part, coriaceous, evergreen, and heavily cuticularized, while those of the mesic forest are thinner, deciduous, and less cuticularized. 4. Greater volumes of palisade mesophyll account for most of the greater blade thickness of shrub-bog leaves as compared with the mesic mountain forms. The actual dermal volume for the two groups is about the same; both the palisade and the spongy mesophyll volumes are greater for the shrub-bog leaves, but the palisade volume is porportionally greater. 5. The shrub-bog leaf blades are smaller than are those of the mesic forest. Correlated with the smaller blade areas of the shrub-bog leaves are smaller surface areas of individual epidermal cells, smaller stomates, and higher stomatal frequencies. 6. Although the surface areas of epidermal cells are smaller for shrub-bog leaves, the cross-sectional area of the palisade cells for most of these leaves is larger; thus, the epidermal-palisade-cell ratio is smaller-that is, fewer palisade cells are contiguous to an epidermal cell. 7. In general, the shrub-bog leaves show less contrast between palisade and spongy mesophyll than do the mountain mesic forest leaves. For many of the bog leaves the vertically oriented palisade intergrades with a predominantly vertically oriented spongy mesophyll. 8. The radial walls of the epidermal cells are less sinuous in the shrub-bog leaves, especially in the leaves with thick cuticles. 9. The shrub-bog leaves have denser vein patterns than do the mesic leaves. 10. In the shrub-bog leaves the denser vein patterns are associated with greater palisade volumes and vertically oriented spongy mesophyll. 11. Five minor vein categories are described. Although all five patterns occur in leaves of both major groups studied, the dendriform, imbedded pattern is more frequent among shrub-bog leaves; veins carrying bundle sheath extensions and anastomosing around polygonal units of chlorenchyma occur more frequently among the mountain mesic leaves. 12. Those shrub-bog leaves which have anastomosing veins with bundle sheath extensions are the ones most heavily cuticularized. 13. The dendroid, imbedded vein patterns in the shrub-bog leaves are related to all-palisade mesophylls or to spongy mesophylls with predominantly vertically elongate and contiguous cells or parts of cells. 14. Leaves of both series have uniseriate dermal layers. 15. Crystal idioblasts and sclereids are associated with particular genera rather than with habitat. 16. Bundle sheath extensions are associated with anastomosing vein patterns rather than with dendriform vein patterns. 17. Hypotheses are proposed concerning the significance of certain blade characteristics of shrub-bog and mesic forest leaves, as to whether they are static, inherent characteristics or plastic responses to the immediate environment. The hypotheses are ones of induction based upon the occurrence of the blade characteristics within species groups or within individual species of paired affinities. Conclusions must be based upon experimental evidence.

WATER RELATIONS OF POCOSIN OR BOG SHRUBS

The pocosins, or broad-leaved evergreen shrub bogs, constitute one of the most distinctive plant communities of the southeastern coastal plain. They occupy flat, poorly-drained areas where the soil is frequently water-logged for long periods of time, but where surface water seldom stands to a depth of more than a few inches, and then only for comparatively short periods. The water relations of the plants growing in pocosins present an interesting problem to the plant physiologist because structurally the leaves of many of the species resemble those of plants growing in a dry habitat rather than those of plants of wet places. The shrubs have simple, more or less elliptical leaves which are rather thick, leathery, and heavily cutinized. The leaf structure is of a type generally believed to have a low rate of transpiration, but there is never a lack of soil water except during severe summer droughts. Though the problem of the water relations of bog plants has long been a subject for discussion, all of the earlier workers confined their investigations to northern bogs. Schimper (25) was among the first to notice the -discrepancy between leaf structure and habitat. He advanced the theory that bogs are physiologically dry, the absorption of water being hindered by the presence of certain humic acids and soluble salts released during decomposition of the organic matter in the soil. He believed that the structure of the bog sclerophylls materially aided in preventing excessive water loss from plants whose root systems were unable to absorb much water because of the physiological dryness of the habitat. This theory of physiological drought was more or less generally accepted. Most of the earlier workers merely speculated as to the cause of the alleged physiological drought, however, and failed to present much real experimental data on the actual water relations of the plants. A number of causes of physiological drought have been suggested. Low soil temperature was regarded by Schimper (25), Warming (30), and Transeau (27) as a factor in northern bogs. Davis (13) and Burns (3) ascribed the xerophytic nature of bogs to the water-retaining power of the peat and to periodic droughts. Transeau (27), Clements (6), and Bergman (2) regarded the lack of oxygen in bog soils as the controlling, or at least the primary, factor in physiological dryness. The importance of toxic compounds was first suggested by Livingston (17, 18) and has been advocated by Dachnowski (9,10,11) and Rigg (23). These so-called toxic compounds are supposedly the result of incomplete decomposition of organic substances in a soil deficient in oxygen. The hypothesis of physiological drought, originally developed in connection with bogs of cool climates, has been extended to explain the vegetative