Palaeoecology and Development of Peatlands in Indiana.
ABSTRACT.--The progressive development of ten Holocene peatlands in northeast Indiana was determined by analysis of macroscopic subfossils recovered from sediment cores. All of the peatlands began as extremely mineral-rich lakes or ponds after retreat of Wisconsin-age glacial ice. The oldest basal date was 13,170 radiocarbon [C.sup.14] y BP. The subfossil assemblage characterizing the lake stages included Chara sp., Ceratophyllum demersum, Najas flexilis and Potamogeton spp. The transition from lake to peatland was marked by a marsh flora dominated by Nuphar sp. Nymphaea sp., and, especially, Brasenia schreberi. Evidence for the development of a fen began early in the hydrosere (in the limnic phase) and colonization of the open water by a floating fen mat brought closure to the emergent marsh. Calcicolous mosses of the family Amblystegiaceae, including Calliergon stramineum, Colliergon trifarium, Campylium polygamum, Drepanocladus aduncus and Scorpidium scorpioides, dominated the fens. During the later stage s of the fen phase an association of Calliergon trifarium and Meesia triquetra dominated the subfossil assemblage. Depending on the morphometry and geology of the respective basins, some of the fens became Sphagnum-dominated bogs. Subfossil assemblages of the bog phase included remains of ericaceous shrubs as well as sedges. Regardless of their developmental pathways, all of the peatlands exhibit an apparent trend towards senescence into lowland forests dominated by Acer rubrum.
Although conditions favoring the genesis of peatlands occur throughout the world (Lucas, 1982), the composition and structure of the successive stands of vegetation that colonize the deposits will ultimately be determined by regional climate, composition of the predominating soil parent-material and biogeographic history (Worley, 1981). In humid oceanic climates with moderate temperatures the later stages of peatland development may result in the build-up of peat into a convex mound that is mostly isolated from the influence of groundwater (ombrotrophic) (Crum, 1988). In peatlands of the southern Great Lakes region, however, where seasonal temperature variations are extreme and short droughts are frequent, decomposition hinders the formation of raised peatlands. Here groundwater nearly always influences the structure and composition of the vegetation (minerotrophic). Complexes of hummocks and hollows of Sphagnum (Vitt et al., 1975; Crum, 1988) seldom develop and, as the peat becomes more dense, trees coloniz e and form a closed canopy (Crow, 1969; Swinehart and Starks, 1994).
In addition to the effects on peat accumulation, climate affects the composition of peatland vegetation. Whereas a Sphagnum-dominated peatland in northern Michigan might be characterized by Carex oligosperma, Eriophorum virginicum, Chamaedaphne calyculata and Picea mariana, a geologically similar Sphagnum-dominated peatland in central Indiana might be dominated by Dulichium arundinaceum, Toxicodendron vernix and Acer rubrum (Markle, 1916; Cain, 1928). In the southern Great Lakes region peatlands seem to be most favorable to the plants typical of marshes, swamps and sloughs because these plants are better adapted to the present climatic and edaphic conditions of the region (Transeau, 1905).
Although peatland vegetation in different latitudes varies in response to climate (Worley, 1981; Crum, 1988), there seems to be a worldwide similarity in peatland development; from ponds to brown moss (Amblystegiaceae)-dominated fens to Sphagnum-dominated bogs (Kuhry et at., 1993). Allogenic factors, such as climate, indirectly influence these processes in a variety of ways. For example, regional water-level changes cause paludification (Miller and Futyma, 1986) and affect lake morphometry, limnology and, consequently, ontogeny. However, the transition from open water to fen and from fen to bog can also be satisfactorily attributed to autogenic internal processes such as lake-fill, clogging of drainage by fine organic detritus and acidification by means of cation exchange in mosses (Glime et al., 1982; Wetzel, 1983; Crum, 1988).
Many investigators have examined the development of peatlands using macroscopic subfossils as indicators of the past structure, composition and succession of local and extra-local vegetation. Most of the North American studies have been conducted in Canada (Zoltai and Johnson, 1985; Kubiw et al., 1989; Nicholson and Vitt, 1990; Kuhry et al., 1993), the northeastern United States (Miller, 1987; Hu and Davis, 1995) and the northern Great Lakes region (Rosendahl, 1948; Miller and Futyma, 1987). While many accounts of macrofossils from lake and wetland deposits in the southern Great Lakes region (southern Michigan, Ohio, Illinois and Indiana) have been presented (Baker, 1918; 1920; Potzger, 1936; Friesner and Potzger, 1946; La Rocque, 1952; Reynolds, 1959; La Rocque and Forsyth, 1957; Zimmerman 1960; Oltz and Kapp, 1963; Whitehead et al., 1982; Jackson et al., 1986; Kapp, 1986; Lepper et al., 1991; Swinehart and Starks, 1994; Swinehart, 1995a, b), no detailed macrofossil studies of peatlands have been conducted there. Because these peatlands occur considerably south of the boreal forest, they represent important palaeoecological resources for comparing differences in the development of peatland vegetation in relation to differences in regional climate. Moreover, the fact that the peatlands in the southern Great Lakes region occur in a less-than-optimal regional climate for peatland development may cause internal and microclimatic factors that favor peatlands to be more easily discerned.
The present study reconstructs the biotic succession of peatlands in the southern Great Lakes region based on macroscopic subfossils recovered from northern Indiana peatlands. The objectives of the study were to: (1) determine the macroscopic subfossil components and developmental stages of ten Indiana peatlands and (2) compare the subfossil assemblages from Indiana peatlands to those of other regions in North America.
The ten peatlands studied occur in northern Indiana on glacial drift of Wisconsin age. Evidence for drifts from at least two previous glaciations in Indiana, the "Kansan" (often a collective term for pre-Illinoian drifts) and Illinoian, has been reported (Melhorn, 1997).
Most of these, with the exception of portions of the Illinoian in southern Indiana, have been covered or obliterated by Wisconsin-age drifts. The combined drifts in northern Indiana range from 15 m to 150 m thick (Wayne, 1956). The deepest deposits occur throughout the interlobate area of the former Saginaw and Erie Lobes in northeast Indiana, averaging approximately 90 m deep. The drift becomes shallower west and south, averaging 45 m deep. The bedrock underlying this unconsolidated material is composed chiefly of Silurian and Devonian shales. The influence of lime-rich glacial diamicton imparts a strongly alkaline reaction on the groundwater, lakes and wetlands of the region.
Northeast and north-central Indiana is characterized by hundreds of lakes. A distinct lake district stretches from the Huron River Valley of Michigan (Transeau, 1905) through the northeast corner of Indiana to a point nearly 160 km southwest. These lake basins include kettle-holes, irregularities in moraines and channel-lakes created in the interlobate region of the former Saginaw and Erie lobes of Wisconsin age (Scott, 1916). The lakes occurring west of the interlobate area (north-central Indiana) are a result of the stagnation and deterioration of the Saginaw Lobe ca. 15,000 y BP. While examples of several types of lake-basins have been identified in Indiana by Scott (1916) and others, the peat-filled basins in the present study, on the basis of their morphometry, would be best classified as kettles (see Swinehart 1997).
Peatlands in this study were selected on the basis of their characteristic vegetation and their restriction to a confined basin (rather than riparian conditions). In the present paper bogs are classified as Sphagnum-dominated peatlands regardless of the source of their water and nutrients. Fens are non-Sphagnum-dominated peatlands. Qualifiers such as "mineral-rich," "mineral-poor," "tall-shrub" and "leatherleaf" are included to elucidate specifics about nutrient status and vegetation structure.
Representative fens, Sphagnum lawns, leatherleaf bogs, tall-shrub bogs (carrs) and forested peatlands were selected for sediment coring. Before coring, the peatlands were probed systematically with metal rods to determine the deepest areas. Probing was conducted at 25-in intervals on basins less than 10 ha and at 50-in intervals on basins over 10 ha. The dense woody flora prevented systematic probing of Blueberry, Little Arethusa and Ropchan Memorial bogs. In these peatlands probing was conducted along a transect traversing the long axis of the basin.
Coring was conducted at the deepest probe location in each peatland. A modified Hiller corer with a chamber diameter of 5 cm was used to collect sediments. Cores were sectioned into 25-cm lengths and placed into plastic bags. A standard 20-ml volume was taken from each interval and gently rinsed through a 0.4-mm mesh sieve. The material remaining in the sieve was placed in a petridish and examined under a dissecting microscope for identification and quantification of macroscopic subfossils. Subfossils are here defined as: (1) remains of once living organisms that are composed, at least in part, of their original organic constituents or (2) inorganic material of biotic origin that has not yet been lithofied (i.e., mollusc shells, Chara tests).
Fruits, leaves, moss thalli (exclusive of Sphagnum), algal oogonia, shells, trichopteran casings, exoskeletons and bones were counted directly and expressed as the actual number of subfossil remains per 20 ml of sediment. Voluminous particulate remains of Sphagnum mosses (mostly leaves), leaves of Ceratophyllum and stem thalli of Chara were expressed as volume percentages based on estimated cover of material placed in a petri-dish (Kuhry et at., 1993). Poorly defined plant tissues were not included in the analysis and Sphagnum remains were not identified to species. Questionable subfossils that seemed referable to a known taxon are preceded by the Latin abbreviations "cf" (conferre).
The remaining sediment from each sampling interval was also rinsed through a 0.4-mm mesh sieve. The material was then placed into a white enamel pan and examined for large infrequent subfossils such as bones, large seeds and leaves that might not have been represented in the 20-ml subsamples. Any subfossils found exclusively in the bulk material were recorded as "present."
Representative subfossils from each peatland were placed in vials with a 60% solution of ethanol. Voucher specimens were placed in the private herbarium/museum of the first author at Hillsdale College and are available for examination. Illustrations and descriptions of the subfossil taxa are presented by Swinehart (1997).
Analysis of the cores was conducted by the first author. Subfossil spectra and resulting diagrams were created with Tilia(C) and Tilia*Graph(c) computer software (Grimm, 1992). Developmental stages of the peatlands were based on assemblages of subfossils that denoted seral stage and/or structural characteristics of the respective ecosystems and were determined by visual inspection of the subfossil diagrams.
Selection of material at different depths for radiocarbon ([C.sup.14]) dating was based on the developmental stages represented in the subfossil diagrams. Whenever possible, aquatic and wetland bryophytes were excluded from the radiocarbon samples. The samples were treated using the Acid-Base-Acid method and were analyzed by accelerator mass spectrometry at the Purdue Rare Isotope Measurement (PRIME) Laboratory. All results are reported in years before present (BP) and have been corrected to Delta [C.sup.13] of -25PDB.
To characterize the predominant extant vegetation the peatlands were sampled systematically. A baseline was established along the long axis of each peatland. Transects spaced at 25 m or 50 m intervals (depending on the size of the peatland) and stretching across the predominant vegetation of each peatland (i.e., exclusive of lagg) were established perpendicular to the baseline. Nested quadrats were placed at 25 m or 50 m intervals along each transect. In tall-shrub bogs and some forested peatlands terrain hindered systematic spacing of sampling plots. In such cases a single line transect was established through the predominant vegetation in the respective peatland.
Vegetation was stratified on the basis of height or stem diameter at breast height (DBH): The tree layer included living stems [greater than]4 cm DBH, the shrub layer included living stems [leq] 4 cm DBH and over 1 m in height, the herbaceous layer was vegetation (including woody plants) less than 1 m in height, the ground layer was restricted to bryophytes and the aquatic layer comprised submerged or floating plants (e.g., Potamogeton (pondweeds) and Lemna spp. (duckweeds)). The tree layer was sampled with 400 [m.sup.2] quadrats, the shrub layer with 100 [m.sup.2] quadrats and all other layers with 1 [m.sup.2] quadrats. Total percent frequency, total percent cover (density and basal area for trees) and importance values were calculated for each species in each peatland (Swinehart, 1997).
Svoboda Fen (Fig. 1) Noble County, Indiana, T35N., R11E., Sec. 8.--This 25-ha peatland has an elevation of 290 m and is situated in loam till of Trafalgar formation (Gray, 1989). It is an open fen with a treacherous floating mat that in many places is separated from the bottom sediments by as much as 6 m of water. An open pond remains in the center of the peatland and is drained by an outlet on the northwest side of the basin.
The 1400-cm core consists mainly of hemic and sapric organic sediment. Sediments immediately above the till are composed of silt and clay. The earliest identifiable organic remains were dated at 13,170 radiocarbon ([C.sup.14]) y BP. Although the composition of the subfossil assemblage is largely uniform throughout the profile, two stages are recognized due to the significant increase in subfossil frequency beginning at about 450 cm. The presence of submergent aquatics, Najas flexilis and Ceratophyllum demersum, indicates the existence of an extremely mineral-rich lake throughout the history of the basin. A larval case of the trichopteran, Micrasema, a species common in lotic situations, suggests the possibility of moving water in or near the basin early in the development of the peatland.
The frequency of Drepanocladus aduncus, Najas flexilis and Ceratophyllum demersum increases in Stage II. A floating mat may have encroached over the core locality, perhaps reducing decomposition by decreasing the surface area and agitation of the water. A false bottom (see Welch, 1952), supporting Najas, Ceratophyllum and Nuphar advena, probably preceded or accompanied the arrival of the mat.
The overall composition of the current bryoflora is different than that suggested by the uppermost subfossil-bearing sediments (225 cm). The floating mat is dominated by Calliergonella cuspidata, Sphagnum fimbriatum, Campylium polygamum, Fissidens adianthoides and Drepanocladus fluitans (in decreasing order of abundance). Dominant vascular plants include Typha spp., Scirpus acutus, Thelypterus palustris, Toxicodendron vernix, Cornus stolonifera and Salix serrissima. The ecotone between the open mat and peripheral forest (dominated by Acer rubrum) is occupied by a ring of Larix laricina. Interstitial waters of the mat have a mean pH of 6.5 and a mean conductivity of 371 [mu]MHOS.
Kiser Lake Fen (Fig 2), Kosciusko County, Indiana, T33N., R7E., Sec. 13.--This fen has an elevation of 260 m and occurs in pitted outwash (Gray, 1989). Peat has accumulated well over 1 m above the surface of the associated lake waters. This is attributed to the groundwater being under pressure, as metal rods inserted through the sediments resulted in upwelling of water to the surface. The 1000-cm profile was composed of organic sediments and marl underlain by sand and gravel. Stage I contained subfossils of molluscs, aquatic macrophytes (Chara and Najas flexilis) and fish (family Centrarchidae), indicating a eutrophic alkaline lake.
An increase in Najas achenes in Stage II corresponds with an increase in the numbers of molluscs. Only Gyraulus parvus persisted to the extinction of open water in Stage III. Its tolerance of weed-choked protected waters (Dexter, 1950; Harmen and Berg, 1971) allowed it to survive where other species fared poorly.
The beginning of Stage III is marked by the complete disappearance of aquatic macrophytes and the appearance of Carex and the brown moss Amblystegium riparium, indicating transition from marsh to fen. This transition occurred at approximately 4360 radiocarbon ([14.sup.C]) y BP. The middle and latter parts of stage III lacked mosses and fruits and were composed chiefly of highly humified vegetative structures of sedges and woody plants.
The peatland is currently dominated by Amblystegium riparium, Eurynchium hians, Carex spp., Potentilla fruticosa, Cornus stolonifera and Cornus foemina. Trees in the peatland are extremely sparse. Interstitial waters have a pH of 6.9 and a conductivity of 606 [mu]MHOS.
Binkley Fen (Fig. 3), Steuben County, Indiana, T38N., R13E., Sec 31.--This 31- has fen has an elevation of 296 m and occurs in morainal deposits of Trafalgar formation (Gray, 1989). No open water remains, but interstitial waters are drained by a minor outlet.
The 1000-cm core is composed of hemic and sapric peat. The till at the base of the core is overlain by 150 cm of silt and clay. Stage I comprises most of the profile and represents a wind-sheltered lake environment indicated by subfossils of Najas flexilis, Chara and Potamogeton. Evidence of a marginal fen mat is given by occasional sparse remains of Calliergon stramineum, Campylium stellatum and Drepanocladus sp. The increase of Brasenia schreberi in the latter part of Stage I suggests the development of a marsh as sedimentation decreased water depth. Najas flexilies gradually decreases in abundance until it becomes rare in Stage II.
Stage II is marked by the increase of subfossils of Campylium polygamum, Drepanocladus sp., Scirpus and, eventually, Vaccinium macrocarpon. Najas fiexilis and Brasenia schreberi decrease significantly but persist to the uppermost level, indicating that closure of open water over the core site was relatively recent. Radiocarbon ([C.sup.14]) dating placed the transition at 173 [pm] 80 y BP.
The peatland is currently dominated by Campylium stellatum, Calliergonelta cuspidata, Scirpus acutus, Eleocharis spp. and Osmunda regalis. Larix laricina, Toxicodendron vernix and Acer rubrum are occasional within the peatland. Several localized areas are becoming Sphagnum-dominated. Interstitial waters have a pH of 6.5 and a conductivity of 238 MHOS.
Dutch Street Bog (Fig. 4), Noble County, Indiana, T35N, R10E., Sec. 11.--This 16 ha peatland has an elevation of 290 m and occurs in undifferentiated outwash (Gray, 1989). Inlets and outlets are lacking. The center of the elongate basin retains standing water, choked with the water-lilies Nuphar and Nymphaea.
The 1400-cm profile is composed of fibric and hemic organic sediments. The lower 3.75 m is dense silty clay. Stage I is characterized by subfossils of the submergent aquatic plants Ceratophyllum demersum and Najas flexilis. The presence of a pioneering marginal fen-mat is evidenced by fragments of Drepanocladus aduncus, Meesia triquetra and Calliergon trifarium.
Stage II is marked by the decline and eventual disappearance of submergent plant remains. Subfossil seeds of Brasenia schreberi and Nuphar advena indicate that open water developed into a shallow marsh. A sharp increase in fragments of Drepanocladus aduncus indicates the encroachment of the fen mat to the vicinity of the coring site.
In Stage III Drepanocladus declines sharply and is eventually replaced by Meesia triquetra and Calliergon trifarium (approximately 4110 radiocarbon ([C.sup.14] y BP). Leaves of Vaccinium macrocarpon become abundant, suggesting the presence of a low-shrub stratum. The end of Stage III is marked by the abrupt disappearance of calcicolous mosses.
Remains of Sphagnum dominate the peat in Stage IV. The presence of subfossils of Sarracenia purpurea, Chamaedaphne calyculata and Vaccinium corymbosum indicate an acidic leatherleaf bog much like that of the extant community. The transition from brown moss-dominated fen to Sphagnum bog occurred at approximately 1140 radiocarbon ([C.sup.14]) y BP.
The peatland is currently dominated by Sphagnum fallax, S. recurvum, S. magellanicum, Chamaedaphne calyculata, Scirpus cyperinus, Rubus hispidus, Vaccinium corymbosum, Aronia melanocarpa and Cephalanthus occidentalis. Interstitial waters of the peat have a mean pH of 3.7 and a mean conductivity of 79 [mu]MHOS.
Burket Bog (Fig. 5), Kosciusko County, Indiana, T32N., R5E., Sec. 20.--This 25-ha peatland has an elevation of 250 m and occurs in mixed drift resulting from the collapse of sub-ice tunnels and open ice-walled channels (Gray, 1989). It has no open water and lacks inlets and outlets.
The 1500-cm profile is composed of fibric and hemic sediments over silt and clay. The earliest sediments bearing macroscopic subfossils are characterized by sparse remains of submergent aquatics, namely Potamogeton sp., Najas flexilis and Ceratophyllum demersum. Evidence of a well-developed and closely proximal fen mat is given by the abundance of Drepanocladus aduncus. A sharp
decline in D. aduncus and an associated humification of the sediments in the 850-900 cm level suggests hydrologic changes that caused increased decomposition of the sediments and perhaps an actual decline in the dominant bryofiora. A similar decline occurs early in Stage II.
The area around the coring site in Stage II developed into a shallow marsh, evidenced by seeds of Brasenia schreberi and Nuphar advena, with a fen mat containing Dulichium arundinaceum, Menyanthes trifoliata, Scirpus validus/acutus and Vaccinium macrocarpon becoming ever closer to the coring site. The end of Stage II is marked by a sharp decline of Drepanoclad us.
Sphagnum abruptly dominates the subfossil bryoflora in Stage III. The transition from brown moss fen to Sphagnum bog occurred at approximately 4010 radiocarbon (14C) y BP. The Sphagnum mat was occupied by Scirpus validus, Vaccinium macrocarpon, Andromeda glaucophylla, Carex sp. and Chamaedaphne calyculata. Occasional seeds of Nuphar advena are attributed to individual plants growing within the mat in wet areas, a common occurrence in modern peatlands in Indiana. A brief decline in Sphagnum and corresponding increase in Drepanocladus aduncus between the depths of 350 and 400 cm indicates local disturbance.
The peatland is currently dominated by Sphagnum papillosum, S. cuspidatum, S. bartlettianum, Chamaedaphne calyculata, Carex oligosperma and Woodwardia virginica. Shrubs of Acer rubrum, Toxicodendron vernix, Aronia melanocarpa, and Cephalanthus occidentalis are occasional on the open mat. The interstitial waters of the peat have a mean pH of 3.9 and a mean conductivity of 44 uMHOS.
Yost Bog (Fig. 6), Lagrange County, Indiana, T38N., R8E., Sec. 32.--This 8 ha peatland has an elevation of 253 m and occurs in pitted outwash deposits. It is one of only two Sphagnum bogs in the present study that exhibits a central pond of open water. The pond is choked with water-lilies (Nuphar and Nymphaea) and the basin has no inlets or outlets.
The 1200-cm profile consists of fibric, hemic and sapric organic deposits. A thin layer of silt and clay separates the organic sediments from the sand and gravel base. The earliest organic remains were dated at 11,560 [pm] 120 radiocarbon (14C) y BP. Stage I contains subfossil remains of submergent aquatic plants (Potamogeton sp., Najas fiexilis and Chara sp.), indicating the presence of a lake or pond. Drepanocladus aduncus, indicating the presence of a marginal fen early in the development of the pond, reaches a peak at the 875-900 cm level and then disappears entirely at 775 cm. Most of the submergent aquatic plants had disappeared by 8060 radiocarbon (14G) y BP, the end of stage one.
The beginning of Stage II is marked by the disappearance of submergent aquatic plants and the beginning of a stratum almost completely devoid of moss subfossils. The scarcity of these groups suggests desiccation of the peatiand. The plant community at the time may have been dominated by a relatively dry sedge meadow.
Inundation of the peatland in Stage III, to the point where standing water was agaln present, is indicated by seeds of Brasenia schreberi and Nymphaea tuberosa. A fen mat occupied by Scirpus subterminalis, Dulichium arundinaceum and Fuirena pumila was also present. Sphagnum subfossils were rare. These mosses probably occurred only as occasional hummocks on the fen mat. However, Sphagnum increased substantially at the depth of 225 cm, marking the beginning of Stage 1V at approximately 1133 radiocarbon (14G) y BP.
The Sphagnum-dominated substrate of Stage IV was occupied by Menyanthes trifoliata, Chamaedaphne calyculata, Scirpus sp. and Andromeda glaucophylla. As build-up of peat produced dry hummocks, the xerophyte Polytri chum strictum became abundant.
The peatland is currently dominated by Sphagnum recurvum, S. magellanicum, Woodwardia virginica, Chamaedaphne calyculata, Toxicodendron vernix, Acer rubrum and Vaccinium corymbosum. Interstitial waters of the peat have a mean pH of 3.2 and a mean conductivity of 96 p.MHOS.
Little Chapman Bog (Fig 7), Kosciusko County, Indiana, T33N., R6E., Sec. 35.--This 9-ha peatland has an elevation of 252 m and occurs in undifferentiated outwash (Gray, 1989). It is an old embayment of Little Chapman Lake, but lacks lotic inlets and outlets. The 1500-cm profile was composed of fibric and hemic organic sediments over a thin layer of sand followed by 4 m of silt and clay. The open eutrophic lake conditions of Stage I are characterized by aquatic macrophytes (Najas flexilis and Cerataphyllum demersum) and fish scales (family Centrarchidae). Evidence of a marginal pioneering mat is given by scarce remains of Calliergon stramineum, Drepanocladus aduncus, Meesia triquetra, Dulichium arundinaceum and Scirpus validus/acutus. Bones of the turtle Sternotherus odoratus were recovered from the 600-625 cm level. The transition between Stages I and II is characterized by the disappearance of submergent aquatic macrophytes and the appearance of floating-leaved marsh plants, namely Brasenia schreberi and Nu phar advena.
In Stage II open water is overtaken by a fen mat dominated by Drepanocladus aduncus. This occurred at approximately 1850 radiocarbon ([C.sup.14]) y BP. Meesia triquetra, Calliergon trifarium, Carex sp., Vaccinium macrocarpon and Menyanthes trifoliata were also present on the mat. The abundance of Calliergon trifarium increased significantly toward the close of Stage II, possibly indicating a dryer less alkaline substrate. While Calliergon trifarium is not listed in the bryoflora of Indiana (Welch, 1957), a radiocarbon ([C.sup.14]) date of 45 [pm] 80 y BP taken at a depth of 25-50 cm indicates that the species existed in the State in recent history.
Stage III (0-25 cm) marks the complete disappearance of brown mosses and the dominance of Sphagnum. The narrow transition zone suggests a rapid conversion from brown moss-dominated fen to Sphagnum-dominated bog.
The open peatland is currently dominated by Sphagnum fimbriatum, Aulacomnium palustre, Sphagnum palustre, Vaccinium macrocarpon, Thelypterus palustris, Osmunda regalis and Toxicodendron vernix. Interstitial waters of the peat have a mean pH of 6.0 and a mean conductivity of 95 [mu]MHOS.
Blueberry Bog (Fig. 8), Elkhart County, Indiana, T38N., R7E., Sec. 24.--This 10-ha peatland has an elevation of 263 m and occurs in pitted outwash deposits (Gray, 1989). It has no inlets or outlets. The nature of the surface vegetation prevented systematic probing of the basin, therefore a line-transect along the long axis of the peatland was established for probing. Consequently, the profile obtained by coring may not be the deepest in the basin. The 275-cm profile was composed mainly of fibric and hemic deposits. The organic material occurred directly over the bottom sand and was not separated by silt or clay. Because this stratigraphy is somewhat unusual, it might be concluded that the core site was at a marginal, paludified location where silt and clay did not accumulate appreciably. Organic material at the bottom of the basin was dated at 2190 radiocarbon ([C.sup.14]) y BP.
Stage I was characterized by an abundance of Najas flexilis achenes. Achenes of Potamogeton cf. P pusillus were also present. Both Najas and Potamogeton indicate a pond or lake environment. Remains of the conifers Abies balsamea, Picea sp. and Larix laricina suggest that forest was close to the core site. The presence of a marginal fen is indicated by abundant fragments of Drepanocladus aduncus. Two fires, one in Stage I and another in Stage II, are indicated by the presence of charred plant remains. The response of the flora to fire (or to dry conditions that may have caused the fires) was an increase in Carex and a decrease in the brown moss flora in both cases.
Stage II is marked by a sharp decline of Najas and the presence of floating-leaved aquatics such as Brasenia schrelieri and Nymphaea tuberosa, indicating a shallow marsh condition. The prominence of Drepanocladus aduncus indicates the continued presence of a marginal fen. Calliergon trifarium becomes abundant on the mat except for a decline at the 150-200 cm depth (attributed to fire).
Similar to the situation in Yost Bog, subfossils are nearly absent in a portion of the profile comprising Stage Ill. Again, it is presumed, on the basis of the lack of aquatic plant remains as well as a lack of woody debris, that the ecosystem at this time was a relatively dry peatforming sedge-meadow. Later, in Stage IV, Sphagnum subfossils become abundant, along with an increase in woody plant remains. Stage V is characterized by humified peat that essentially lacks identifiable subfossils. This decomposition is attributed to "grounding" of the peat beyond the water table (and subsequent oxidation) resulting from increasing bulk density.
The peatland is currently covered with a nearly impenetrable community of shrubs, including Vaccinium corymbosum, Aronia melanocarpa, Cephalanthus occidentalis and Flex verticillata. The understory is dominated by Sphagnum recurvum, S. cuspidatum, Hypericum virginicum and a member of the Poaceae. Chamaedaphne calyculata is occasional between the tall-shrub thickets. Interstitial waters of the peat have a mean pH of 3.5 and a mean conductivity of 68 [mu]MHOS.
Little Arethusa Bog (Fig. 9), Kosciusko County, Indiana, T31N., R7E., Sec. 7.--This 15-ha peatland has an elevation of 264 m and occurs in mixed till and stratified drift in chaotic form (Gray, 1989). It has no known natural inlets or outlets. However, the hydrology appears to have been altered by anthropogenic drainage ditches.
The 1600-cm profile is composed mainly of fibric and sapric organic sediments separated from the underlying till by approximately 2 m of silt and clay. The post-glacial condition was a lake (Stage I) with submergent aquatic plants, chiefly Ceratophyllum demersum. The presence of fish communities is evidenced by several scales referable to the family Centrarchidae (J. A. Holman, pers. comm.). At some point between the depths of 500 and 1000 cm, a prominent marginal fen mat was initiated, marking the beginning of Stage II. The fen phase (Stage II) was characterized by abundant remains of the moss Drepanocladus aduncus at approximately 2670 radiocarbon ([C.sup.14]) y BP. Less abundant calcicolous fen mosses that occurred on the mat were Calliergon trifarium and Meesia triquetra. Larix laricina, represented by leaf fragments, was likely a common arboreal pioneer on the mat. These species appeared in the middle of Stage II and increased significantly toward the beginning of Stage III. Whereas the fen mat was in c lose proximity, the area immediately associated with the coring site in Stage II was occupied by an emergent marsh. This is indicated by relatively abundant seeds of Brasenia schreberi and Nuphar advena. They disappeared entirely at the close of Stage II, suggesting elimination of open water due to the encroachment of the fen.
Stage III is characterized by a decrease in Drepanocladus aduancus and a corresponding increase in Calliergon trifarium and Meesia triquetra. Low shrubs such as Vaccinium macrocarpon and Andromeda glaucophylla appeared. Larix leaves became exceedingly more abundant, and fruits of Betula alleghaniensis appeared, suggesting the development of a forest canopy. Drepanocladus and Calliergon disappear from the subfossil record and Sphagnum increases significantly, apparently colonizing at about the same time as the forest. Meesia triquetra remained until the uppermost layers (59 [pm] 80 [C.sup.14] y BP), but is presently not found in the extant flora of the peatland. It is not known from Indiana's historic bryoflora (Swinehart, 1995b).
Currently, the entire peatland is forested with Acer rubrum, Ulmus americana and Acer saccharinum. Depauperate stands of Larix laricina are still common in the forest interior. The understory is dominated by Sphagnum affine, Thuidium delicatulum, Pallavicinia lyellii, Impatiens capensis, Symplocarpus foetidus, Osmunda cinnamomea, Vaccinium corymbosum, Toxicodendron vernix and Ilex verticillata. Interstitial waters of the peat have a mean pH of 6.2 and a mean conductivity of 196 [mu]HMOS.
Ropchan Memorial Bog (Fig 10), Steuben County, Indiana, T38N., R12E., Secs. 22 & 23.--This 14-ha peatland has an elevation of 293 m and occurs in mixed till and stratified drift in chaotic form (Gray, 1989). It has no inlets or outlets. The 900-cm profile was composed of fibric and hemic peat underlain by silt and clay. A significant portion of the profile was not recovered due to its flocculent nature. The earliest organic deposits (Stage I) are characterized by submergent aquatic plants (Chara sp. and Najas flexilis). Sparse remains of Scirpus validus/acutus, Calliergon stramineum, C. trifarium and Meesia triquetra suggest the establishment of a marginal fen mat. Somewhere between the depths of 300 cm and 575 cm remains of submergent aquatic plants disappeared entirely and fen mosses became more abundant.
Stage II (ca. 2835 [pm] 80 y BP) is marked by abundant remains of the calcicolous mosses Meesia triquetra and Calliergon trifarium. Calliergon stramineum was present in lesser amounts. The fen mat was occupied by Menyanthes trifoliata, Scripus validus-type, Andromeda glaucophylla and Vaccinium macrocarpon. Occasional patches of Sphagnum are evidenced by its appearance in the subfossil record late in Stage II.
Stage III is characterized by the dominance of Sphagnum. Calcicolous mosses disappear entirely. Low shrubs such as Andromeda glaucophylla and Vaccinium macrocarpon eventually disappear as well, suggesting increasingly less inundated conditions due to peat build-up. By Stage III the peat had become grounded and was subsequently able to support forest vegetation both in terms of stability and edaphic conditions. The development of a pioneer forest of Larix laricina is indicated by subfossil leaves in the most recent sediments.
Currently, the entire peatland is forested with Acer rubrum, Larix laricina and Betula allegheniensis. The understory is dominated by Sphagnum fallax, S. affine, Aulacomnium palustre, Pallavicinia lyellii, Osmunda cinnamomea, Majanthemum canadense, Acer rubrum, Ilex verticillata and Vaccinium corymbosum. Interstitial waters of the peat have a mean pH of 4.4 and a mean conductivity of 84 [mu]MHOS.
All of the peatlands follow the same sequence of development from lake or pond to brown moss-dominated fen. Subsequent to the establishment of a fen, some of the peatlands developed into Sphagnum-dominated bogs, others became forested. With the exception of Blueberry Bog, none of the peatlands showed subfossil evidence of disturbance by fire as seen in some northern peatlands (Kuhry et al., 1993). Each of the major phases of peatland development, in terms of local and regional variation, will be treated separately.
THE LIMNIC PHASE
All of the peatlands began as mineral-rich lakes or ponds. The resulting sediments were highly humified and rich in silt. Four genera were characteristic. Chara is most common in the earliest subfossil-bearing sediments and likely pioneered newly created, relatively oligotrophic, extremely alkaline lakes. It is a common component of aquatic subfossil assemblages throughout northern North America, especially in the earliest sediments (Rosendahl, 1948; Daily, 1961; Oltz and Kapp, 1963; Miller, 1973; Birks, 1976; Whitehead et al., 1982; Anderson et al., 1986; Jackson et al., 1986; Kapp, 1986; Kubiw et al., 1989; Nicholson and Vitt, 1990; Kuhry et at., 1993).
Potamogeton subfossils, while represented only by occasional achenes, are widespread in limnic sediments in Indiana. The genus was likely an important component of the aquatic flora as suggested by modern analogues in the Midwest. Potamogeton subfossils have been reported in many other late-Pleistocene deposits in North America (Rosendahl, 1948; Argus and Davis, 1962; Watts and Winter 1966; Whitehead et al., 1982; Jackson et al., 1986; Kapp, 1986; Kuhry et al., 1993; Hu and Davis, 1995; Swinehart, 1995a).
By far, the most widespread and frequent macroscopic subfossil of the aquatic sediments was Najas flexilis. Curiously, while common in many peatland profiles in the midwestern and eastern United States (Rosendahl, 1948; Swinehart and Starks, 1994; Hu and Davis, 1995; Swinehart, 1995a; 1996; Watts and Winter, 1966), Najas was not found in aquatic sediments of Canadian peatlands studied by Nicholson and Vitt (1989) and Kuhry et al. (1993). The species is presently widespread in Canada (Gleason and Cronquist, 1991).
What is particularly characteristic of the aquatic phase of peatlands in the southern Great Lakes region is the prominence of subfossils of Ceratophyllum demersum. This species was found in 6 of the 10 peatlands studied and was often very abundant as leaf fragments. Muenscher (1944) states that [the achenes] are often found in muck soils in the bottoms of former ponds and lakes." The only other subfossil records for Ceratophyllum that could be located were from Rosendahl (1948) from an interglacial deposit in Minnesota and from Birks (1976) from ca. 9000-12,000 y. BP. The species is absent from most published accounts of Holocene peatland profiles.
Of the aquatic flora represented in the subfossil assemblages, Ceratophyllum may be the best indicator of the conditions during the limnic phase. First, it is usually characteristic of quiet eutrophic waters-mainly small lakes and ponds and protected bays of larger lakes (Muenscher, 1944; Gleason and Cronquist, 1991). Secondly, it is restricted to temperate climates and presently is found only as far north as southern Canada (Gleason and Cronquist, 1991). The presence of Ceratophyllum, then, would suggest a quiet pond or lake in a temperate climate not unlike extant lakes and ponds in the Great Lakes region. While many of the aquatic macrophytes could indicate shallow water, and therefore (due to the great depth of the deposit) suggest subsequent rises in the regional water table, such a conclusion is not warranted here. First, while these species may be most common in shallow water, all can be found in relatively deep water. Second, it is uncertain as to whether the remains were produced in situ, or transpo rted from shallow lake margins. And last, there are few modern analogues in the same seral stage with the same trophic character with which to make comparison.
Many, perhaps most, of the lakes were deep enough and trophically rich enough to support fish populations. The probability of finding fish scales in a relatively narrow core is unknown and depends on the frequency and distribution of the scales. The recovery of scales was limited to small clusters within a given depth-interval rather than random occurrences throughout the profile. The scales could represent the previous vicinity of a carcass.
All of the fish scales recovered belong to the family Centrarchidae (J.A. Holman, pers. comm.). Because the present natural distribution of the sunfish family in North America does not extend far north of the Great Lakes Region (Page and Burr, 1991), and members of the family are generally found in temperate or subtropical climates, it might be concluded that their presence indicates temperate conditions. Additionally, bones of the musk turtle (Sternotherus odoratus), recovered from a depth of 600-625 cm in Little Chapman Bog (3680 y BP), indicate climatic conditions at least as warm as those of the present (Holman, 1990). Records of fishes and turtles from other Holocene peatland cores are not known, although excavations of other wetland types from both the Holocene and earlier periods have been published (Baker et al., 1986; Holman, 1990; Holman and Richards, 1993; and others).
THE SHALLOW-WATER MARSH PHASE
Transition from lake to fen was marked by a marsh with standing water in nearly every profile investigated. The establishment of the marsh stratum is attributed primarily to the reduction in water depth as a result of sedimentation. Three species distinguished the marsh phase from the previously limnic and subsequent fen phases: Brasenia schreberi, Nuphar advena and Nymphaea tuberosa. While all of these species are typical of the littoral zones of modern lakes, the latter two species are usually much more abundant than the former. If the frequency of subfossil seeds is a reliable indicator of relative abundance, Brasenia schreberi appears to have been the most prolific aquatic vascular plant in the littoral zones of the basins examined. As many as 50 seeds were recovered in a 300 ml sample. Brasenia was also found in peatland profiles obtained by other investigators (Watts and Winter, 1966; Birks, 1976; Futyma and Miller, 1986; Miller and Futyma, 1987; Hu and Davis, 1995). It is also a common component of pr esent-day bog lakes (see Winkler, 1988). It may indicate that the waters of the basin were not only protected from wind but neutral to slightly acidic in pH, as modern specimens are often most common in these conditions (Muenscher, 1944; Voss, 1972; Gleason and Cronquist, 1991; Swink and Wilhelm, 1994).
The end of the marsh stage seems to be more of a function of horizontal mat growth than vertical peat accumulation. In many of the peatlands, a significant portion of the profile associated with the marsh/fen interface was missing due to its flocculent nature or complete lack of sediment (open water). This is attributed to the encroachment of the fen mat over the open water of the marsh. Additional evidence for this hypothesis is given by the presence of fen mosses at the base of the pocket of "false bottom" as well as at the top, suggesting the settlement of detritus from the underside of the fen mat. Since a "terrestrial" substrate is created long before the basin fills entirely, a state of "hydrological equilibrium" is reached, where any biomass produced on the mat is pushed down into the water-table and kept inundated and anoxic. This condition prevails until enough mat- and detritus-peat accumulates and "grounds" the mat. Once this grounding occurs, biomass can accumulate beyond the limits of groundwat er and subsequent oxidation moderates the ratio of productivity and decomposition (Kratz and DeWitt, 1986).
THE FEN PHASE
Subjossil bryophytes of the pioneering mat.--A marginal mat of floating alkaliphilic vegetation occurs early in the development of Indiana peatlands, suggesting the presence of certain morphometric and/or hydrologic qualities that favor mat and peat formation in these basins. Bryophytes of pioneering fen mats included Calliergon stramineum, Calliergon trifarium, Campylium polygamum, Drepanocladus aduncus, Meesia triquetra and Scorpidium scorpioides. Drepanocladus aduncus was, by far, the most abundant subfossil in the early and middle portions of the fen phase in seven of the ten peatlands studied. In many cases, it was found in pure strata to the exclusion of all other bryophytes. Only rarely were associates such as Campylium polygamum, Calliergon spp. and Scorpidium scorpioides recovered. If preservation was not particularly differential among the bryophytes, it can be concluded that Drepanocladus aduncus was the dominant peat-forming moss during the fen phase in most Indiana peatlands.
The conspicuous presence of Drepanocladus aduncus has been documented in other Holocene wetland deposits in the Great Lakes region (Miller, 1973; Miller and Futyma, 1987) and other members of this genus (including D. aduncus) commonly dominate fen strata throughout North America (Rosendahl, 1948; Kubiw, 1989; Kuhry et al., 1993). Domination by D. aduncus in the fen phase of Indiana peatlands suggests extremely mineral-rich conditions with a pH near neutral or slightly below. J. Janssens reports the mean pH for habitats containing D. aduncus as 6.06 (data presented by Kuhry et al., 1993).
Subfossils of Drepanocladus aduncus varied according to position in the profile. The earliest subfossils were large and elongate, similar to variety kneiffii (Janssens, 1983a). Fragments recovered from the upper portion of the profiles were often smaller and less etiolated, much like var. aduncus (Janssens, 1983a). A mixture of the two forms was found in the middle portions of the stratum. This morphologic difference is attributed not to taxonomic distinction, but to ecophenic responses to wet (perhaps submergent) and relatively dry conditions, respectively.
Bryophyte succession on the aging mat.--In six of the peatlands investigated Drepanocladus became associated with, or in many cases was replaced by, Calliergon trifarium in the later stages of the fen phase. The establishment and eventual dominance of C. trifarium was almost always accompanied by Meesia triquetra, the same association described for Tamarack Bog, Noble County, Indiana (Swinehart, 1995b). Wisconsinan subfossils of C. trifarium in North America have been reported from Alaska (Janssens, 1983b), Indiana (Swinehart, 1995b), Michigan (Miller and Futyma, 1987), Minnesota (Birks, 1976; Janssens, 1983c; Janssens and Glaser, 1986), Ohio (Wynne, 1945), Wisconsin (Cheney, 1931; Wilson, 1938) and Canada (Lowdon et at., 1971; Janssens, 1983b). Wisconsin-age subfossils of M. triquetra in North America have been reported from Alaska (Janssens, 1983b), Indiana (Swinehart, 1995b), Iowa (Montagnes, 1990), Minnesota (Birks, 1976; Janssens, 1983c; Janssens and Glaser, 1986), New York (Miller, 1973), Ohio (Montage s, 1990; Miller, 1992) and Canada (Persson and Sjors, 1960; Kuc, 1973; Janssens, 1983b; Kubiw et at., 1989; Kuhry et at., 1993). Nowhere, except in the Indiana peatlands, was the Calliergon trifarium-Meesia triquetra association evident.
The increase in Calliergon trifarium and Meesia triquetra before and during the transition from fen to bog in Indiana is well defined and consistent and suggests some change in the hydrology and/or chemistry of the fen mat. While specific values for pH preference and height above water-table are apparently not available for Calliergon trifarium, quantitative studies have been conducted for Meesia triquetra. Montagnes (1990) reports that M. triquetra occurs in habitats with a pH ranging from 4.7-8.1 (n = 61). This is a broader range than that given for Drepanoctadus aduncus which is 6.6-7.5 (n = 9) (Janssens, 1983a). The values for conductivity for the two species show a similar pattern, where M. triquetra tolerates a broader range (12-1035 [uScm.sup.-1]) (Montagnes, 1990) than D. aduncus (150-575 [uScm.sup.-1]) (Janssens, 1983a).
Because the Calliergon trifarium--Meesia triquetra assemblage occurs immediately before the transition to Sphagnum-dominated peat (indicative of increasingly acidic conditions), and because studies of the habitat of M. triquetra show a greater pH tolerance of acidity, it is concluded that this association was initiated as a result of increasing acidity of the fen mat. Acidification of the fen mat is attributed to increased isolation from the influence of ground water (resulting from thickening of the mat) and cation exchange by the brown mosses (Glime et at., 1982). Additionally, a decrease in D. aduncus with a corresponding increase in M. triquetra and C. trifarium probably indicates dryer conditions. Although the mean height above the water table for M. triquetra determined by Janssens (data presented by Kuhry et at., 1993) was lower than that of D. aduncus (6 cm and 10 cm, respectively), Glime et at. (1982) found Drepanoctadus only in the wet sedge and cattail-dominated zones of a southern Michigan fen.
Modern analogues for Drepanodadus-dominated fens.--What is most striking and significant about the subfossil moss assemblages is the lack of modern analogues in Indiana. While living Drepanoctadus aduncus is still found in Indiana (Welch, 1957), none of the peatlands studied currently harbored this species and the genus was never represented to the extent that it was the major peat-former in the current bryoflora of the three fens. The fens are currently dominated by one or more of the brown mosses Cattiergonetta cuspidata, Ambtystegium riparium, Campytium steltatum, C. potygamum and Fissidens adianthoides.
At least one other Indiana fen shows a similar decline in Drepanocladus Cabin Creek Raised Bog, in Randolph County. Although a layer of Drepanocladus peat nearly 1 m thick was found in profiles of the peat (Friesner and Potzger, 1946), no specimens of this genus were recorded in the modern flora (Welch, 1962). The fen is currently dominated by Campylium stellatum, Fissidens adianthoides and Cratoneuron filicinum, among other alkaliphilic mosses (Welch, 1962). Similar assemblages can be found in Cedar Bog, Champaign County, Ohio (McQueen, 1975).
Lawrence Lake Fen, Barry County Michigan, is the only known published account from the southern Great Lakes region where Drepanocladus is still a major peat-former (Clime et al., 1982). Abundant growth of the genus, however, was restricted to the very wet area of the peatland, where Scirpus acutus, other sedges and cattails dominated (Glime et al., 1982). This situation is very similar to that of Svoboda Fen, but Drepanocladus was relatively infrequent there, Drepanocladus aduncus may be more common as a submergent species in Indiana, as it was found in luxuriant masses at depths up to one meter in a Noble County pond. Given inferred abundance of Drepanocladus in the subfossil record and apparent rarity in the current bryoflora, it can be concluded that: (1). most or all of the peatlands in Indiana have developed beyond the sere that is most favorable to Drepanodadus, (2) Drepanocladus was growing submerged and was not sampled in those fens which retain deep ponds or (3) Drepanocladus has declined significan tly in the peatlands of Indiana and the southern Great Lakes region as a result of major climatic and/or anthropogenic changes.
Modern analogues for the Calliergon trifarium-Meesia triquetra assemblage.--Extant populations of Calliergon trifarium and Meesia triquetra have not been reported from Indiana. Both species are considered boreal, and in the northern Great Lakes region they are considered to be infrequent relics of previously cooler climates (H. A. Crum, pers. comm.). While extant populations of both species are common in peatlands of the far north (and often occur together in the same peatland), a dominant association comprised of both species, like that suggested in portions of the palaeoecological record from many Indiana peatlands, is not known to exist at present. In modern peatlands M. triquetra is most commonly associated with Aulacomnium palustre, Bryum pseudotriquetrum, Drepanocladus revolvens, D. vernicosus and Tomenthypnum nitens (Montagnes, 1990). Vascular plant associates include Betula pumila, Carex spp., Menyanthes trifoliata, Potentilla palustris and Salix pedicellaris (Montagnes, 1990).
It is difficult to attribute the demise of Calliergon trifarium and Meesia triquetra in Indiana to climate change. One problem is the presence of suitable microclimates in Indiana's remaining fens that function quite satisfactorily as refugia for other boreal relics. The second major argument against the climate change hypothesis is the fact that subfossils of both C. trifarium and M. triquetra occur within 25 cm of the surface of some Indiana peatlands and radiocarbon dating indicates that they survived to within 100 y of the present. Additionally, a contemporary record of C. trifarium has been reported from Ohio (N. C. Miller, pers. comm.).
Montagnes (1990) attributes the decline of Meesia triquetra in rich fens to cultural disturbance and eutrophication rather than climate. This conclusion seems to be the most likely explanation for the loss of not only M. triquetra but also Calliergon trifarium from Indiana's bryoflora. Most of the fens in the state have been destroyed and a large part of the remainder (perhaps all) have been influenced by drainage, agricultural run-off and erosion.
The vascular flora of the fen mat.--Although the past and present bryofloras of Indiana fens differ significantly, the subfossil assemblages of vascular plants are very similar to the plant communities occupying modern fens in the southern Great Lakes region. Early in the development of the fen mat, Bidens cf B. cernua, Carex cf. C. pseudocyperus, Dulichium arundinaceum, Impatiens capensis, Polygonum sp., Scirpus subterminalis and S. validus/ acutus were represented. All of these species are common in modern fens in the region, especially Dulichium arundinaceum and Scirpus acutus which are important in mat development.
Near the transition from fen to bog, Carex spp., Eleocharis spp., Menyanthes trifoliata, Vaccinium macrocarpon and, occasionally, Andromeda glaucophylla often become frequent. All of these species are common to modern transitional fens. One plant in particular, however, was abundant in only one of the peatlands. Fuirena pumila was found in association with Dulichium arundinaceum, Eleocharis sp. and Scirpus sp. in the later stages of the fen phase in Yost Pond, Lagrange County. Deam (1940) writes of its present habitat and distribution in Indiana:
This sedge is very local, having been found in only a few places in two counties [Porter and Steuben]. It grows in moist sand in interdunal swamps and in wet sand on the borders of lakes.
The Yost Bog specimens are the only subfossil records of Fuirena pumila of which the authors are aware.
The vascular flora characteristic of fens is very important in the physical development of the mat. Subfossil records and observations of extant peatlands in Indiana suggest that the fen mat, rather than the bog mat, is responsible for pioneering open water. In Svoboda Fen a mat floating on 6 m of water is invading the open lake. Only at the shoreward margins of the mat have conditions become favorable to Sphagnum. In bogs where paludification is occurring at the shoreward margins, extremely treacherous floating mats of Dulichium arundinaceum and other sedges lead the colonization of open lagg water. Only after the chemical characteristics of the newly formed substrate have been altered sufficiently by the bryophyte community does Sphagnum and other bog plants follow. The floating mat, by its buoyancy, also helps prevent flooding of alkaliphobic bryophytes with mineral-rich lake water.
In the rare situations where a relatively deep centrally-oriented pond occurs in a bog (Yost Bog and Dutch Street Bog are the only such cases in Indiana of which the authors are aware) the edge of the bog mat (having a flora nearly indistinguishable from the rest of the mat) is not at all treacherous and is quite thick (1-2 m). While bog mats can invade open water (Swan and Gill, 1970), some may grow very slowly, recede or reach equilibrium. In the Indiana examples, however, the fen mat apparently did not encroach upon the entire water surface before bog flora became established when the pH and alkalinity approached favorable values. Filling of the ponds will now likely be controlled simply by sedimentation.
THE BOG PHASE
In fens that (1) lacked inlets and outlets, (2) were buffered from strong groundwater influence by significant basal deposits of clay and silt and (3) had small watersheds, Sphagnum became established and eventually dominated (Swinehart, 1997). The stratigraphic transition from brown moss-dominated fen to Sphagnum-dominated bog is extremely abrupt. Such an implied rapid transition from fen to bog was also concluded by Kuhry et al. (1993). Vitt and Kuhry (1992) attribute the rarity of transitional peatlands (pH between 5 and 6) to the rapidity by which they become acid, Sphagnum-dominated bogs. Three large patches characterized by Sphagnum hummocks and situated over the deepest sub-basins in Binkley Fen, Steuben County, are the only examples of transitional fen observed in this study. Hickory Bog in Noble County, Indiana, first described by Swinehart (1994),is a transitional fen. Little Chapman Bog, Kosciusko County, is post-transitional and has recently become Sphagnum-dominated.
In addition to Sphagnum, the bog phase is characterized by subfossils of Andromeda glaucophylla, Carex spp., Chamaedaphne calyculata, Vaccinium macrocarpon and increased wood fragments. A seed of Sarracenia purpurea and a leaf of Vaccinium corymbosum were recovered from Dutch Street Bog. Yost Bog was the only site where subfossils of Polytrichum strictum were recovered. The occurrence of P. strictum in the upper layers of the bog peat suggests xeric conditions at least on the tops of Sphagnum hummocks. The species is still common on the tops of hummocks in Yost Bog and other leatherleaf bogs.
Unfortunately, Chamaedaphne subfossils were so poorly preserved that their actual abundance and duration in the peatland are probably inaccurately represented. However, subfossil leaves of Andromeda glaucophylla were remarkably well preserved and indicate times when the area about the core site was relatively wet. The species is still found in all of the open Sphagnum bogs studied, but only in the wettest areas.
The richness of subfossil remains in the Sphagnum-peat decreased substantially as compared to the fen peat. This is attributed to dryer conditions on the surface of the peat created in part by peat buildup and hummocks of living Sphagnum projecting above the wet substrate. This condition would increase the chances of decomposition of plant debris on the mat. Evidence for this is the consistent disappearance of Vaccinium macrocarpon leaves in the middle of the bog stratum even though living specimens remain abundant presently. Decrease in subfossil taxa may also be a direct reflection of an actual decrease in plant species richness.
THE FORESTED PHASE
The two forested peatlands included in this study, Little Arethusa Bog and Ropchan Memorial Bog, represent two different developmental pathways; a lowland forest resulting from a mineral-rich fen and a lowland forest resulting from a relatively mineral-poor bog, respectively. Both presently have similar canopy structure and composition (Swinehart, 1997).
The forested phase of the peatlands studied was characterized by subfossils of Larix laricina. Subfossils of Betula alleghaniensis were present in Little Arethusa Bog. In Ropchan Memorial Bog Larix occurred above the Sphagnum peat in the uppermost 25 cm of the profile, suggesting a relatively recent forest invasion. Tamarack Bog, Noble County, Indiana, had a nearly identical pattern of development (Swinehart and Starks, 1994). In contrast, in Little Arethusa Bog, Larix occurred at about the same time that the fen mat developed, suggesting that tamarack pioneered the open mat at an earlier time and formed a canopy as soon as the peat around the margins became grounded. Larix subfossils were not found in any of the peatlands that do not currently harbor living populations (except for a few leaves from the basal, terrestrial sediments of Blueberry Bog, Elkhart County). Both of the forested peatlands are currently dominated by Acer rubrum. Larix trees are being shadedout, and most are depauperate and dying.
THE PEATLAND CLIMAX
The peatlands of the southern Great Lakes region lack black spruce (Picea maraiana) and northern white-cedar (Thuja occidentalis), and therefore do not develop into the muskegs and cedar swamps characteristic to the later stages of peatland development in the North. Regardless of the specifics of their sequential development, all peatlands in this study currently exhibit a trend toward development into lowland forests dominated by Acer rubrum. Other investigations also indicate that A. rubrum dominates the later stages of peatland development in the southern Great Lakes region (Transeau, 1905; Crow, 1969; Sytsma and Pippen, 1982; Swinehart and Starks, 1994). Crum (1988) states that, "[In the southern Great Lakes region], the fens [Sphagnum-dominated and non-Sphagnum-dominated minerotrophic peatlands] give way to tamarack, poison sumac, red maple, and black ash."
Little Arethusa Bog and Ropchan Memorial Bog have already become red maple swamps. The remaining peatlands appear to be heading toward a similar fate, as red maple is present in the predominant vegetation zone (not including the lagg) of 7 of the 8 nonforesred peatlands (Swinehart 1997). In these non-forested peatlands, Acer rubrum has a mean frequency of 19% as trees (DBH [greater than] 5 cm), 26% as shrubs (DBH [less than] 4 cm, height greater than 1 m) and 8% as members of the herbaceous layer (height [less than] 1 m) (Swinehart 1997). The margins (not exclusive to the lagg) of these sites are mostly swamp forests dominated by Acer rubrum, along with Quercus palustris and Ulmus americana. If spatial vegetation composition and structure is an indication of future trends, then the open centers of these peatlands will develop communities much like the presumably older, more swamp-like margins. Unfortunately, subfossil support for this hypothesis may be difficult to obtain because of the poor preservation of A. rubrum remains in swamp forests.
The transition to red maple swamp is likely due to grounding of the mat. This favors red maple by (1) providing a stable substrate to support tree-size vegetation, (2) increasing decomposition, thus releasing greater amounts of nitrogen and phosphorus needed for survival of seedlings (Moizuk and Livingston, 1966) and (3) increasing the potential for flooding of the surface because the mat can no longer rise and fall with changing water levels. Although not necessary for growth, Acer rubrum thrives in areas with seasonal standing water (Moizuk and Livingston, 1966). Additonally, red maple can replace other bog trees such as tamarack because of its tolerance of shade.
Another possible explanation for the increase of red maple in peatlands in the southern Great Lakes region is anthropogenic nutrient-loading from run-off and/or atmospheric fallout. Most of the peatlands investigated in this study occur in agricultural areas and have the potential to receive nutrient-rich agricultural run-off. However, studies conducted on peatlands a century ago, when atmospheric and agricultural pollution in the southern Great Lakes region was much less, show similar trends toward red maple swamps. In the bogs studied by Transeau (1905), red maple was found in abundance in the peatlands:
The further development of these societies under present conditions will bring about a complete change. There can be no doubt that the poplars and red maples are the coming trees, with elm a close third. When these have become sufficiently large and numerous to shade the shrubs, the latter will be killed out, and we shall have in their place the maple-elm forest common to the low grounds (Transeau, 1905).
As Crum (1988) remarked, it is unreasonable to consider any forests climactic in terrain so recently deglaciated as the upper Midwest. However, peatlands in Indiana, having been deglaciated longer than the better-known peatlands of the north, offer a better picture of their possible fate, at least in terms of vegetation structure. The uppermost radiocarbon dates in forested Little Arethusa Bog suggest that productivity is still exceeding decomposition. Up to 25 cm of organic soil has been produced in 59 [pm] 80 y BP. If this trend continues, it is not impossible that this peatland, and others like it, will develop into mesic forest communities as the soil surface becomes further removed from the water table.
While there is much controversy about the nature and existence of a classic hydrosere in northern peatland (Winkler, 1988; Klinger, 1996; and others), lake-fill peatlands in Indiana appear to follow a classic successional development from open water to lowland forest. The opportunity for full-canopy forests to succeed bogs and fens in the southern Great Lakes region is attributed to climate. Due to frequent summer droughts, the climate in the southern Great Lakes region favors greater decomposition and nutrient cycling once the secondary stage in peatland development is reached (Crum, 1988). The result is that once the peatland becomes grounded, Sphagnum is no longer favored and can no longer engineer bog conditions, and thus a swamp rather than an ombrotrophic bog flora prevails.
Acknowledgments.--We thank the Basil S. Turner Foundation of Elkhart, Indiana, for funding the larger study from which the present paper is derived. Lee Casebere of the Indiana Department of Natural Resources and Carolyn McNagny of ACRES Landtrust, Inc., provided site information and granted permission to access many of the sites. Kenneth Mueller of the Purdue Rare Isotope Measure- ment Laboratory provided the radiocarbon dates. ALS thanks Gina M. Warner for invaluable assistance in the field, and Drs. Carole A. Lembi, William R. Chaney and Jonathan M. Harbor for providing helpful comments throughout the study.
(1.) Present address: Department of Biology, Hillsdale College, Hillsdale, Michigan 49242. Telephone (517) 437-7341; e-mail: email@example.com
ANDERSON, R. S., R. B. DAVIS, N. G. MILLER AND R. STUCKENRATH. 1986. History of late- and post-glacial vegetation and disturbance around Upper South Branch Pond, northern Maine. Can. J. Bot., 64:1977-1986.
ARCUS, G. W. AND M. B. DAVIS. 1962. Macrofossils from a late-glacial deposit at Cambridge, Massachu-setts. Am. Midl. Nat., 67:106-117.
BAKER, F. C. 1918. Post-glacial mollusca from the marls of central Illinois. J. Geology, 26:659-671.
-----. 1920. Pleistocene Mollusca from Indiana and Ohio.]. Geology, 28:439-457.
BAKER, R. G., R. S. RHODES II, D. P. SCHWERT, A. C. ASHWORTH, T.J. FREST, G. R. HALLBERG AND J. A. JANSSENS. 1986. A full glacial biota from southeastern Iowa, USA. J. Quaternary Sci., 1:91-107.
BIRKS, H. J. B. 1976. Late-Wisconsin vegetational history at Wolf Creek, Central Minnesota. Ecol. Monogr., 46:395-429 + 3 unpaged foldout figs.
CAIN, S. A. 1928. Plant succession and ecological history of a central Indiana swamp. Bot. Gaz., 86:384-401.
CHENEY, L. S. 1931. More fossil mosses from Wisconsin. Bryologist, 34:93-94.
CRUM, H. A. 1988. A focus on peatlands and peat mosses. University of Michigan Press, Ann Arbor. 306 p.
CROW, G. E. 1969. An ecological analysis of a southern Michigan bog. Mich. Bot., 8:11-27.
DAILY, F. K. 1961. Glacial and post-glacial charophytes from New York and Indiana. Butler Univ. Bot. Stud., 14:39-72.
DEAM, C. C. 1940. Flora of Indiana. Indiana Dept. of Conservation, Indianapolis. 1236 p.
DEXTER, R. W. 1950. Distribution of the mollusks in a basic bog lake and its margins. The Nautilus, 64: 19-26.
FRIESNER, R. C. AND J. E. POTZGER. 1946. The Cabin Creek Raised Bog, Randolph County, Indiana. Butler Univ. Bot. Stud., 8:24-43.
FUTYMA, R. P. AND N. G. MILLER. 1986. Stratigraphy and genesis of the Lake Sixteen Peatland, northern Michigan. Can. J. Bot., 64:3008-3019.
GLEASON, H. A. AND A. CRONQUIST. 1991. Manual of vascular plants of northeastern United States and adjacent Canada. New York Botanical Garden, Bronx. 910 p.
GLIME, J. M., R. G. WETZEL AND B. J. KENNEDY. 1982. The effects of bryophytes on succession from alkaline marsh to Sphagnum bog. Am. Midl. Nat., 108:209-223.
GRAY, H. H. 1989. Quaternary geologic map of Indiana. Indiana Geol. Surv. Misc. Map # 49.
GRIMM, E. C. 1992. Tilia(C) and Tiliae Graph(c) Computer Softeware. Illinois State Museum, Springfield.
HARMEN, W. N. AND C. O. BERG. 1971. The freshwater snails of central New York. SEARCH-Agriculture (Cornell University), 1:1-68.
HOLMAN, J. A. 1990. Vertebrates from the Harper Site and rapid climatic warming in mid-Holocene Michigan. Michigan Academician, 22:205-217.
----- AND R. L. RICHARDS. 1993. Herpetofauna of the Prairie Creek Site, Daviess County, Indiana. Proc. Indiana Acad. Sci., 102:115-131.
Hu, F. S. AND R. B. DAVIS. 1995. Post-glacial development of a Maine bog and paleoenvironmental implications. Can. J. Bot., 73:638-649.
JACKSON, S. T., D. R. WHITEHEAD AND G. D. ELLIS. 1986. Late-glacial and early Holocene vegetational history at the Kolarik Mastodon Site, northwestern Indiana Am. Midl. Nat., 115:361-373.
JANSSENS, J. A. 1983a. Past and extant distribution of Drepanocladus in North America, with notes on the differentiation of fossil fragments. J. Hattori Bot. Lab., 54:251-298.
-----. 1983b. Quaternary fossil bryophytes in North America: new records. Lindbergia, 9:137-151.
-----. 1983c. A quantitative method for stratigraphic analysis of bryophytes. In Holocene peat. J. Ecology, 71:189-196.
----- AND P. H. GLASER. 1986. The bryophyte flora and major peat-forming mosses at Red Lake Peatland, Minnesota. Can. J. Bot., 64:427-442.
KAPP, R. O. 1986. Late-glacial pollen and macrofossils associated with the Rappuhn Mastodont (Lapeer County, Michigan). Am. Midl. Nat., 116:368-377.
KLINGER, L. F. 1996. The myth of the classic hydrosere model of bog succession. Arctic and Alpine Research, 28:1-9.
KRATZ, T. K. AND C. B. DEWITT. 1986. Internal factors controlling peatland-lake ecosystem development. Ecology, 67:100-107.
KUC, M. 1973. Addition to the Arctic moss flora. VI--moss-flora of Masik River Valley (Banks Island) and its relationship with plant formations and the post-glacial history. Rev. Bryol. Lichenol., 39: 253-264.
KUBIW, H., M. HICKMAN AND D. H. VITT. 1989. The developmental history of peatlands at Muskiki and Marguerite Lakes, Alberta. Can. J. Bot., 67:3534-3544.
KUHRY, P, B.J. NICHOLSON, L. D. GIGNAC, D. H. VITT AND S. E. BAYLEY. 1993. Development of Sphagnum-dominated peatlands in boreal continental Canada. Can. J. Bot., 71:10-22.
LA ROCQUE, A. 1952. Molluscan faunas of the Orleton Mastodon Site, Madison County, Ohio. Ohio J. Sci., 52:10-28.
----- AND J. FORSYTH. 1957. Pleistocene molluscan faunules of the Sidney cut, Shelby County, Ohio. Ohio J. Sci., 57:81-89.
LEPPER, B. T., T. A. FROKLING, D. C. FISHER, G. GOLDSTEIN, J. E. SANGER, D. A. WYMER, J. G. OGDEN AND P. E. HOOGE. 1991. Intestinal Contents of a late Pleistocene mastodont from mid-continental North America. Quaternary Research, 36:120-125.
LOWNDON, J. A., I. M. ROBERTSON AND W. BLAKE, JR. 1971. Geological survey of Canada radiocarbon dates IX. Radiocarbon, 12:46-86.
LUCAS, R. E. 1982. Organic soils (histosols)--formation, distribution, physical and chemical properties, and management for crop production. Michigan State University Agricultural Experiment Station and Cooperative Extension Service, Research Report 435 Farm Science. 77 p.
MARKLE, M. S. 1916. Phytecology of peat bogs near Richmond, Indiana. Proc. Indiana Acad. Sci., 1915: 359-375.
MCQUEEN, C. B. 1975. Comparison of the present and past bryophyte flora of Cedar Bog. Ohio J. Sci., 75:188-190.
MELHORN, W. N. 1997. Indiana on ice: the late Tertiary and ice age history of Indiana landscapes, p. 15-27. In: M. T. Jackson (ed.). The Natural Heritage of Indiana. Indiana University Press, Bloomington.
MILLER, N. G. 1973. Lateglacial plants and plant communities in northwestern New York State. J. Arnold Arboretum, 54:123-159.
----- 1987. Late Quaternary fossil moss floras of eastern North America: evidence of major floristic changes during the late-Pleistocene-early Holocene transition. Symposia Biologica Hungarica, 35:343-360.
----- 1992. A contribution toward a history of the Arctic moss flora. Contr. Univ. Mich. Herb., 18: 73-86.
MILLER, N. G. AND R. P. FUTYMA. 1987. Paleohydrological implications of Holocene peatland development in northern Michigan. Quat. Res., 27:297-311.
MOIZUK, G. A. AND R. B. LIVINGSTON. 1966. Ecology of red maple (Acer rubrum L.) in a Massachusetts upland bog. Ecology, 47:942-950.
MONTAGNES, R. J. S. 1990. The habitat and distribution of Meesia triquetra in North America and Greenland. The Bryologist, 93:349-352.
MUENSCHER, W. C, 1944. Aquatic plants of the United States. Comstock Publishing Company, Ithaca, New York. 374 p.
NICHOLSON, B.J. AND D. H. VTT. 1990. The paleoecology of a peatland complex in continental western Canada, Can. J. Bot., 68:121-138.
OLTZ, D. E AND R. O. KAPP. 1963. Plant remains associated with mastodon and mammoth remains in central Michigan. Am. Midl. Nat., 70:339-346.
PAGE, L. M. AND B. M. BURR. 1991. A field guide to freshwater fishes. Houghton Mifflin Company, Boston. 432 p.
PERSSON, H. AND H. SJORS. 1960. Some bryophytes of the Hudson Bay Lowland of Ontario. Su. Bot. Tidskr, 54:247-268.
POTZGER,J. E. 1936. Post-pleistocene fossil records in peat of the upper Blue River valley, Henry County, Indiana. Proc. Indiana Acad. Sci., 45:65-68.
REYNOLDS, M. B. 1959. Pleistocene molluscan faunas of the Humboldt deposit, Ross County, Ohio. Ohio J. Sci., 59:152-1 66.
ROSENDAHL, C. O. 1948. A contribution to the knowledge of the Pleistocene flora of Minnesota. Ecology, 29:284-315.
SCOTT, W. 1916. Report on the lakes of the Tippecanoe Basin (Indiana). Indiana Univ. Stud. 3(31):39 p., 13 maps.
SWAN,J. M. A. AND A. M. GILL. 1970. The origins, spread, and consolidation of a floating bog in Harvard Pond, Petersham, Massachusetts. Ecology, 5:829-840.
SWINEHART, A. L. 1994. An ecological investigation of three Sphagnum bogs in Noble County, Indiana: remnant communities of a region in transition. M.S. Thesis, Dept. of Biology, Central Michigan University, Mt. Pleasant, Michigan. 224 p.
----- AND G. D. STARKS 1994. A record of the natural history and anthropogenic senescence of an Indiana tamarack bog. Proc. Indiana Acad. Sci., 103:225-239.
-----. 1995a. Paleoecology of an alkaline peatland in Elkhart County, Indiana. Proc. Indiana Aced. Sci., 104:43-46.
-----. 1995b. Subfossils of the boreal mosses Calliergon trifarium and Meesia triquetra from an Indiana peatland. Ohio J. Sci., 95:278-280.
-----. 1996. Palacoecology of the Wilkinson Giant Beaver Locality, Whitley County, Indiana. Report to the Indiana Dept. Nat. Res., Indiana State Museum. 19 p., 5 figs., 2 tabs.
-----. 1997. The development and ecology of peatlands in Indiana. Ph.D. Dissertation, Department of Forestry and Natural Resources, Purdue University, West Lafayette, Indiana. i-xx + 303 p.
SWINK, F. AND G. WILHELM. 1994. Plants of the Chicago Region, 4th ed. Indiana Academy of Science, Indianapolis. 921 p.
SYTSMA, K. J. AND R. W. PIPPEN, 1982. The Hampton Creek wetland complex in southwestern Michigan III. Structure and succession of tamarack forests. Mich. Bot., 21:67-76.
TRANSEAU, E. N. 1905-1906. The bogs and bog flora of the Huron River Valley. Bot. Gaz., 40:351-375, 418-448 (1905); 41:17-42 (1906).
VITT, D. H. AND P. KUHRY. 1992. Changes in moss-dominated wetland ecosystems, p. 178-210. In:J. W. Bates and A. M. Farmer (eds.). Bryophytes and lichens in a changing environment. Oxford University Press, Oxford.
-----. H. A. CRUM AND J. A. SNIDER. 1975. The vertical zonation of Sphagnum species in hummockhollow complexes in northern Michigan. Mich Bot., 14:190-200.
Voss, E. G. 1972. Michigan flora, Vol. 2. Cranbrook Institute of Science, Bloomfield Hills, Michigan. 724 p.
WATTS, W. A. AND T. C. WINTER. 1966. Plant macrofossils from Kirchner Marsh, Minnesota--a paleoecological study. Geol. Soc. Amer. Bull., 77:1339-1360.
WAYNE, W. J. 1956. Map showing thickness of drift in Indiana north of Wisconsin glacial boundary. Indiana Dept. of Conservation, Report of Progress 7, Plate 1.
WELCH, P. S. 1952. Limnology, 2nd ed. McGraw Hill Book Company, New York. 538 p.
WELCH, W. H. 1957. Mosses of Indiana. Indiana Department of Conservation, Indianapolis. 478 p.
----- 1962. Bryophytes of Cabin Creek Raised Bog. Proc. Indiana Acad. Sci., 72:105-107.
WETZEL, R. G. 1983. Limnology, 2nd ad. Saunders College Publishing, Fort Worth, 767 p.
WHITEHEAD, D. R., S. T. JACKSON, M. C. SHEENAN AND B. W. LEYDEN. 1982. Late-glacial vegetation associated with caribou and mastodon in central Indiana. Quaternary Research, 17:241-257.
WILSON, L. R. 1938. The postglacial history of vegetation in northwestern Wisconsin. Rhodora, 40:137-175.
WINKLER, M. G. 1988. Effect of climate on development of two Sphagnum bogs in south-central Wisconsin. Ecology, 69:1032-1043.
WORLEY, I. A. 1981. Maine peatlands. Their abundance, ecology, and relevance to the Critical Areas Program. Vermont Agricultural Experiment Station, Bulletin 687. i-v + 387 p.
WYNNE F. E. 1945. Studies in Calliergon and related genera. Bryologist, 48:131-155.
ZIMMERMAN, J. A, 1960. Pleistocene molluscan faunas of the Newell Lake Deposit, Logan County, Ohio. Ohio J. Sci., 60:13-39.
ZOLTAI, S. C. AND J. D. JOHNSON. 1985. Development of a treed bog island in minerotrophic fen. Can. J. Bot., 63:1076-1085.
|Printer friendly Cite/link Email Feedback|
|Author:||SWINEHART, ANTHONY L.; PARKER, GEORGE R.|
|Publication:||The American Midland Naturalist|
|Date:||Apr 1, 2000|
|Previous Article:||Age-related Fecundity in Four Taxa of Western Shrews (Sorex spp.).|
|Next Article:||Dynamics of Wetland and Upland Subshrubs at the Salt Marsh-Coastal Sage Scrub Ecotone.|