Status of Forested Wetlands

In colonial times (circa 1780) the conterminous United States had approximately 221 million acres of wetlands (BROKEN-LINK BROKEN-LINK Dahl 1990). These wetlands had been, and would continue to be, affected by natural and anthropogenic disturbances. Over the next 200 years (circa 1980) the total wetland area in the country was reduced by over 50 percent to 104 million acres (Table 20.1). Losses are primarily attributable to clearing and draining for agriculture. Frayer and others (1983) suggest that the greatest losses between the 1950s and the 1980s were in freshwater forested wetlands. Abernethy and Turner (1987) estimated losses of forested wetlands were up to five times greater than those of nonforested wetlands between 1940 and 1980. Almost 7 million forested wetland acres were lost in the Lower Mississippi Valley alone.

BROKEN-LINK BROKEN-LINK Hefner and Brown (1985) reported that 47 percent (48.9 million acres) of the wetlands in the conterminous United States occurs in 10 Southeastern States (Kentucky, Tennessee, North Carolina, South Carolina, Georgia, Florida, Alabama, Mississippi, Louisiana, and Arkansas). In addition, 65 percent of all the forested wetlands in the conterminous United States occurred in these 10 Southern States. Table 20.2 provides an estimate of total wetland acres, forested wetland acres, and forested wetland change in Southern States. BROKEN-LINK BROKEN-LINK Hefner and Brown (1985) reported that for the period between the 1950s and 1970s the South sustained the greatest wetland losses in the country. Forested wetland losses were attributed to massive clearing and drainage projects designed to bring wetlands into agricultural production. As of the 1970s BROKEN-LINK BROKEN-LINK Hefner and Brown (1985) reported that 80 percent of the 25 million acres of forested wetland in the Lower Mississippi River Valley had been lost to agriculture. Major losses of pocosins and Carolina Bays in North Carolina were attributed to agriculture and peat mining. Overall, forested wetland acres in the South declined by 16 percent between the 1950s and 1970s (Table 20.1).

Hefner and others (1994) reported that approximately 3.1 million acres (9 percent) of forested wetlands in the South were lost or converted in the 1970s and 1980s (Table 20.1). Forested wetlands in these 10 Southeastern States were lost or converted at an average rate of 276,000 acres per year from the 1950s to 1970s but lost at an average rate of 345,000 acres per year from the 1970s to 1980s (Hefner and others 1994). More than 719,000 acres of forested wetlands were converted to scrub-shrub wetlands from the 1970s to 1980s. Almost 69 percent of the South’s forested wetland losses were recorded in the Gulf-Atlantic Coastal Flats and Lower Mississippi Alluvial Plain. The Gulf-Atlantic Coastal Flats of North Carolina and the Lower Mississippi Alluvial Plain of Louisiana suffered the greatest losses during this time period. Nearly 1.2 million acres were lost in North Carolina, presumably to silviculture and agriculture, and nearly 1 million acres of forested riverine wetlands (bottomland hardwood wetland) were severely affected primarily by agriculture in the Lower Mississippi Alluvial Plain. Although the net rate of wetland loss declined from 386,000 acres per year from the 1950s to 1970s to 259,000 acres per year from the 1970s to 1980s, the rate at which forested wetlands declined accelerated (Hefner and others 1994). The drop in overall wetland loss rate resumed between 1986 and 1993, declining 80 percent to 58,500 acres per year for the conterminous United States (Dahl 2000). The change in forested wetland acres during this time period was approximately 3 percent (Table 20.1). Dahl (2000) estimated that nationally 4 million acres of forested wetland underwent some change in condition between 1986 and 1997. Most were converted to freshwater shrub wetlands by timber harvesting or other processes that removed the tree canopy but retained the wetland character. Table 20.1 indicates forested wetland losses exceed total wetland losses for the 1986–97 time period. This is due to the inclusion of restored wetland acreage in the “total wetland loss” category which reduces the actual losses. Table 20.3 shows a breakdown of the number of palustrine (freshwater) forested wetland acres lost or converted by activity and by State for the period of 1986–97, recorded by NWI, for the 13 Southern States included in the Southern Forest Resource Assessment. Georgia, North Carolina, Mississippi, South Carolina, and Alabama showed the greatest change in forested wetland area—over 300,000 acres per State. In each of the above cases, over 80 percent of the change in wetland type resulted from a conversion from forested wetland to shrub-scrub or emergent wetland. Overall, 90 percent of the change in forested wetland acres in the 13 Southern States resulted from these types of conversions. Ninety-five percent of the conversions of forested wetland were to shrub-scrub or emergent wetland types.

According to NWI, losses (changes from wetland to nonwetland) accounted for 10 percent of the change in forested wetlands in the South or 356,000 acres between 1986 and 1997. Thirty-three percent of the losses were due to urban/rural development, 31 percent to agriculture, and 29 percent to silviculture. The remaining 7 percent of losses of forested wetland were attributed to other land uses. The NWI attributes losses to silviculture, if drainage occurs on any forested site (including those in agricultural or urban landscapes) such that a shift from wetland vegetation to upland vegetation is apparent (Personal communication. 2001. Charles Storrs, National Wetland Coordinator, Southeast Region, U.S. Fish and Wildlife Service, Atlanta, GA) The three States with the greatest reported losses due to silviculture were Louisiana, Georgia, and Arkansas. The three States with the greatest loss due to agriculture are Mississippi, Georgia, and Tennessee. The three States with the greatest losses to development were Florida, Mississippi, and Georgia.

Direct comparisons of various wetland inventories is difficult due to the dynamic nature of wetlands, differences in the time period in which the inventories are made, differences in geographic cover, and differences in sampling and delineation protocols (Shepard and others 1998). However, indirect comparison of the NWI and NRI results are interesting. From 1982 through 1987 the NRI data indicated that urban, industrial, and residential land uses caused 48 percent of the wetland losses in the conterminous United States. Agriculture was responsible for 37 percent of wetland losses, while the remaining 15 percent were converted to barren land, open water, or forest (Brady and Flather 1994). For this time period the NRI data suggest a shift from agriculture to urban development as the major cause of wetland conversion. From 1982 to 1992 NRI data indicate that 55 percent of the total wetland loss in the Nation occurred in the 12 Southern States. During this period, wooded wetlands showed the lowest loss rate in recent decades. According to NRI, 75 percent of the losses from 1982 to 1992 were due to development (Shepard and others 1998). The updated 1997 NRI report shows that 12.5 percent of the losses of wetlands in the South are attributable to silviculture, 18.4 percent to agriculture, 58 percent to development, and 10.1 percent to miscellaneous climatic and hydrologic changes. Differences in definitions for attributing loss are a primary reason for discrepancies in wetland loss and conversion estimates between NWI and NRI (Personal communication. 2001. Charles Storrs, National Wetland Coordinator, Southeast Region, U.S. Fish and Wildlife Service, Atlanta, GA).

Land ownership patterns of forested wetlands have been summarized for 5 of the 13 Southern States by Brown and others (2001). About 60 percent of the wetland timberland in Virginia, North and South Carolina, Georgia, and Florida is privately owned. Forest industry owns 28 percent of the land, and the public owns 12 percent (Brown and others 2001). Data from the other eight Southern States is unavailable. Of the wetland timberland in the five Southern States for which data are available, 62 percent is covered with bottomland hardwoods, 25 percent with pine plantations and natural pine stands, and 10 percent oak-pine stands. Most of these forest types are in private nonindustrial ownership except for pine plantations, which are largely owned by forest industry (68 percent) (Brown and others 2001). The percentage of timberland in wetland and the expected increase in timber harvest in the South (chapter 13) indicate the likelihood of additional wetland modifications due to silvicultural activities.

Likely future of forested wetlands in the South—Projecting changes in forested wetlands in the South is difficult, if not impossible, because of the wide variety of scientific, societal, and economic factors that affect the forested wetland resource. Science has provided a great deal of information on how wetlands function and how human activities affect those functions. However, much information is not known and is difficult to discern. The values that people associate with forested wetlands vary greatly. They range from valuing old-growth forest to the exclusion of timber harvesting to valuing forested wetlands as merchantable timber or nothing more than potential development sites. Economic factors are important because, ultimately, wetlands are lost to development or agriculture or converted to intensive silviculture based upon economics.

This section of the chapter addresses changes in wetland condition, with particular emphasis on silviculture, current policies, and the efficacy of current forested wetland restoration efforts in the South. Additional information about forces of change in southern forests can be gained from other chapters in this Assessment.

Forested wetland types in the South are highly variable, ranging from baldcypress swamps to scrub-shrub bogs that undergo cycles of wildfire. Due to these differences in vegetation, hydrology, landscape position, and degree of alteration, wetlands differ in the functions they perform and their ability to perform those functions (Brinson and Rheinhardt 1998). Wetland functions can be simply described as the things that wetlands do. Many of these functions, such as surface and ground-water conveyance and storage, nutrient cycling, and organic carbon export provide societal benefits, goods, and services, (such as floodwater storage, water-quality enhancement, and wildlife habitat). Because of the large geographic area encompassed in this study (13 States), generalizations about forested wetlands must be made. The HGM (Brinson 1993) and functional assessment approach (Smith and others 1995) provide a means to make these broad generalizations about similar forested wetland types, the functions they perform, and the effects of certain activities on those functions.

The predominant forested wetlands in the South can be classified into four HGM classes: (1) riverine, (2) organic soil flats, (3) mineral-soil flats, and (4) depressions (Brinson 1993). Wetlands in each class occupy similar landscape positions and have similar hydrology. The presumption in HGM classification is that if wetlands occupy similar landscape positions so that the water, which drives wetland functions, comes from similar sources and flows into and out of wetlands in similar ways, the ecological processes (functions) that make wetlands important will be similar. This is a logical simplification that facilitates the discussion of wetland ecological characteristics and processes and human impacts.

In general, southern deepwater swamps, major alluvial floodplains, and minor alluvial floodplains (Messina and Connor 1998) can be combined into the riverine class. Carolina Bays, pondcypress swamps, and mountain fens can all be classified as depressions with similar depressional geomorphology and low-energy surface runoff or ground-water hydrodynamics. Wet pine flatwoods are classified as mineral-soil pine flats due to their soil composition, flat topography, and the predominance of rainfall for their hydrology. Pocosins are classified as organic soil flats. Their topography and hydrology are similar to those of mineral-soil flats, but soil composition is dominated by peat. The flats class encompasses areas dominated by pines and by hardwoods. However, mineral-soil pine flats will be the predominant flats class discussed in this chapter due to their extent, fire ecology, and vulnerability to alteration. Based upon the acreage estimates in Table 20.4, riverine is the predominant HGM class in the South, followed by flatwoods and depressions.

In general, the hydrologic regime is one of the main factors controlling ecosystem functions in all wetlands and differentiating wetland types. The timing, duration, depth, and fluctuations in water level affect biogeochemical processes and plant distribution patterns. The rate, magnitude, and timing of biogeochemical processes are determined by hydrology and the living components of an ecosystem. For instance, primary producers (plants) assimilate nutrients and elements in soil and use energy from sunlight to fix carbon. When they die, they depend upon microbial organisms in soil to transform carbon and nutrients such as nitrogen and phosphorous to forms that are available to other plants. Therefore, wetland conditions that maintain plants and soil microbial populations are those that drive characteristic biogeochemical processes. These processes help to sustain the wetland plant community, which provides much of the structure required by wildlife. The integrated combination of water, soils, and plants sustains the ecosystem and provides many of the values attributed to wetlands.

Riverine wetlands—Riverine wetlands occur in floodplains and riparian corridors in association with stream channels (Brinson 1993). The dominant water source for these wetlands is from the stream channel via overbank flooding or through subsurface connections between the stream channel and the wetland. Riverine wetlands lose surface water in four ways: (1) surface flow of floodwater to the channel, (2) subsurface water flow to the channel, (3) percolation to deeper ground water, and (4) evapotranspiration. Evapotranspiration includes evaporation from soil and water surfaces and movement of water through plants to the atmosphere. Unimpacted southern forested riverine wetlands typically extend perpendicularly from a stream channel to the edge of the stream’s floodplain. They have unaltered soils and a mature tree canopy, and they range from narrow riparian strips in low-order streams to broad alluvial valleys several miles wide (Sharitz and Mitsch 1993). This wetland ecosystem occurs in the Lower Mississippi River Valley as far north as southern Illinois and along many streams that drain the South Atlantic Coastal Plain into the Atlantic Ocean.

The functions of riverine wetlands are closely tied to flooding of adjacent streams and the soil and vegetation which result. Flooding is important both ecologically and societally because floodwaters move sediments and nutrients into and out of the wetlands. Wetlands detain floodwaters and prevent or minimize flood damages downstream (Kellison and others 1998, Mitsch and Gosselink 2000, Sharitz and Mitsch 1993). Riverine wetlands enhance water quality by intercepting sediments, elements, and compounds from upland or aquatic nonpoint sources of pollution. They permanently remove or temporarily immobilize nutrients, metals, and other toxic compounds (Ainslie and others 1999). Hydrologic, soil, and biological factors determine the ability of a riverine wetland to sustain a characteristic plant community. The vegetation of low-gradient alluvial riverine wetlands is extremely diverse (Sharitz and Mitsch 1993). The ability to maintain a characteristic plant community is important because of the intrinsic value of the plants themselves, and the many attributes and processes of riverine wetlands influenced by the plant community. For example, plants influence primary productivity, nutrient cycling, and the ability to provide a variety of habitats necessary to maintain local and regional diversity of animals (Brinson 1990, Gosselink and others 1990, Harris and Gosselink 1990). Riverine wetlands provide habitats for a diversity of terrestrial, semiaquatic, and aquatic organisms. They provide access to and from uplands for completion of aquatic species’ life cycles, provide refuges and habitat for birds, and act as conduits for dispersal of species to other areas. Most wildlife and fish species in riverine wetlands depend on the amount and timing of flooding, the variable topography which allows different plants and animals to become established, forest tree composition and structure, and proximity to other habitats. Riverine wetlands also must be viewed in their landscape context or in relation to the other land uses around them. Generally, the continuity of vegetation, the connection between specific vegetation types, the presence and size of corridors between upland and wetland habitats, and corridors among wetlands all have direct bearing on the movement and behavior of animals that use wetlands.

Depression wetlands—These wetlands occur in topographic depressions that allow the accumulation of surface water (Brinson 1993). Depression wetlands may have a combination of inlets and outlets or lack them completely. Potential water sources are precipitation, overland flow, streams, or ground water/interflow from adjacent uplands. Water typically flows from the outside of the depression to the center. Upward and downward movement of the water table may vary daily to seasonally. Cypress domes and Carolina Bays are typical regional forested wetland types (Messina and Conner 1998) that occur in depressions. Pondcypress domes are poorly drained to permanently wet depressional wetlands that occur in the southeastern Coastal Plain and are abundant in Florida (Ewel 1990). Cypress domes are shallow, circular, nutrient-poor swamps located in depressions on low-relief landscapes. They often have an underlying impervious layer of soil that inhibits downward movement of water. These wetlands are called “domes” because the tallest trees are in the center and the smaller trees near the edge give the appearance of a dome. Domes have long-standing, nutrient-poor water which is often dominated by precipitation and surface inflow (Mitsch and Goselink 2000). Limited plant growth rates are related to both low flow and lack of nutrient availability.

Carolina Bays occur on the Atlantic Coastal Plain from New Jersey to Florida. The water source for Carolina Bays ranges from predominantly precipitation to predominantly ground water. These bays occur in clusters, are commonly elliptical in shape, and are often oriented in a northwesterly to southeasterly direction. Larger, deeper Carolina Bays contain lakes, but the majority of them are wetlands with diverse plant communities ranging from shrub-bog pocosins to marshes to hardwood- or cypress-dominated swamp forests. Many bays may become blanketed by an overgrowth of bog vegetation, which compresses lower layers of peat, making them relatively impervious to water movement. The result is a ponding of water, making the depression saturated for long periods of time. Bays are critical breeding sites for amphibians and habitat for birds and other wildlife. They often host rare or endangered plants.

Detention of runoff water is an important depressional wetland function because runoff, or occasional overbank flooding in riparian depressions, alters flood timing, duration, and magnitude. The result is reduced flood flow downstream. Water storage or detention has significant effects on biogeochemical cycling; plant distribution, composition and abundance; and wildlife populations. Just as in riverine wetlands, nutrient cycling is mediated primarily by two processes: (1) nutrient uptake by plants (primary production), and (2) nutrient release from dead plants for renewed uptake by plants (detrital turnover). Because of their location on the landscape, depressional wetlands, particularly those in lower portions of watersheds, are strategically located to remove and sequester sediments, imported nutrients, contaminants, and other elements and compounds before they can contribute to ground water and surface-water pollution downstream. These contaminants are removed from incoming water by the interaction of water, wetland vegetation, wetland microbes, detrital material, and soil. The primary benefit of this function is that the removal, conversion, and sequestration of compounds by depressional wetlands reduces the load of nutrients and pollutants in ground water and in any surface water leaving the depressional wetland. Not all depressions are positioned or capable of removing these sediments, compounds, and contaminants. For instance, depressions at the top of drainage basins, or those in flat topography, may not receive pollutants from upstream.

Depressional wetlands support many animal populations. They provide habitats within the actual wetland and in conjunction with the surrounding landscape. They maintain regional biodiversity by providing open water, nesting cavities, cover and food chain support for a variety of animals (Ewel 1998). In some regions, Carolina Bays are major and critical focal points for breeding and feeding of a large variety of nonaquatic vertebrate and invertebrate animal species. The biomass of animals in these Carolina Bays is extremely high compared to adjacent terrestrial habitats or more permanent aquatic habitats (Richardson and Gibbons 1993).

Forested wet flats—In the Southern United States, wet flats occur on poorly drained mineral or organic soils in lowland areas (Harms and others 1998, Rheinhardt and others 2002). Wet flats on organic, or peaty, soils are called pocosins. Pocosins differ from mineral-soil flats in both geomorphology and vegetation. Pocosins are located on topographic highs and are dominated by evergreen shrubs, and most burn every 15 to 30 years (Rheinhardt and others 2002, Richardson 1981). The hydrologic regime of pocosins is driven by precipitation, but water flows outward from the center and eventually forms headwater streams near the wetland’s outer boundaries (Brinson 1993). The organic soils of pocosins tend to hold water longer than mineral-soil flats. As a result, frequency of fire is less than in mineral-soil flats.

Mineral-soil flats are most common on areas between rivers, extensive lake bottoms, or large floodplain terraces where the main source of water is abundant precipitation and slow drainage associated with a landscape of low relief (Brinson 1993, Rheinhardt and others 2002). This class predominantly occurs on the Atlantic Coastal Plain from Virginia to Texas (fig. 20.1). There are two subclasses of mineral-soil flats: those dominated by a closed canopy of hardwoods, and those characterized by open savanna with widely scattered pines (Rheinhardt and others 2002). Mineral-soil hardwood flats in the Yazoo Basin of Mississippi occur on former and current floodplains created by the Mississippi River and its tributaries (Smith and Klimas 2002). Mineral-soil flats receive virtually no ground-water discharge. This characteristic distinguishes them from depressions. The dominant direction of water movement is downward through infiltration. These wetlands lose water by evapotranspiration, surface runoff, and seepage to underlying ground water. They are distinguished from flat upland areas by their poor drainage due to impermeable layers (hardpans), and slow lateral drainage. Mineral-soil pine flats will be the focus of the following discussion due to the millions of acres that still exist and their susceptibility to alteration due to fire exclusion, development, and silvicultural conversion to pine plantation.

The pre-European landscape was largely maintained by fires resulting from lightning strikes and Native American burning. However, with the colonization and subsequent management by Europeans, less than 2 percent of the fire-maintained character of mineral-soil pine flats remained by the 1990s. In their least altered condition, wet pine flats have very few trees. When trees are present, longleaf, pond, and occasionally slash and loblolly pines are naturally associated with this wetland type. All four pines can tolerate ground fires by the time they reach 6 to 9 feet in height, but longleaf is the only pine whose seedlings are adapted to tolerate fire. The combined stresses of fire and wetness led to the evolution of an unusually rich flora on many wet pine flats (Rheinhardt and others 2002).

Wet pine flats differ from other wetlands due to a combination of factors that do not occur together in any other wetland type. These factors combine to control the biogeochemical processes characteristic of wet pine flats:

1. The source of water, dominated by precipitation and vertical fluctuations in water level driven by evapotranspiration, is generally low in nutrients.
2. When flooding occurs, it is shallow (10 to 20 cm) and flows slowly.
3. The number of pits and mounds on the ground surface is high and provides a diverse array of aerated and anoxic conditions for soil microbial organisms.
4. Nutrient recycling occurs in pulses following fires, which recur on a frequent basis, thus enabling a rapid turnover of nutrients. These four attributes enable wet pine flats to tightly and rapidly cycle nutrients. As a result, wet pine flats rapidly recover their characteristic biomass and structure after fires (Rheinhardt and others 2002).

Plant communities characteristic of unaltered wet pine flats are maintained by an appropriate hydrologic regime, fire regime, and biogeochemical processes that require intact soil conditions. Under relatively unaltered conditions, these three parameters combine to maintain a grassy savanna with few or no trees. On some sites, the herbaceous plant community is extremely rich. In fact, the herbaceous species richness is the highest recorded in the Western Hemisphere (Walker and Peet 1983). This herbaceous assemblage is extremely sensitive to alteration and, as a consequence, many species associated with this ecosystem are rare or threatened with extinction. Because the herbaceous community of wet pine flats is so sensitive to alteration (fire exclusion, hydrologic alteration, and soil disturbance), its condition provides information on habitat quality. Plant populations in wet pine flats have evolved to both withstand and require frequent fire. Fire stimulates flowering and seed set in many wet savanna species, such as toothache grass and wiregrass. As a result, species composition and spatial habitat structure reflect fire frequency. In the absence of fire, wet pine flat vegetative composition becomes dominated by shrubs or hardwood trees. This is a degraded condition when compared to a fire-maintained wet pine flat.

Animals that use unaltered wet pine flats for all or part of their lives are adapted to habitats maintained by frequent fire. Frequent fire maintains open savanna, which is important to some animal species using wet pine flats. For animal species that utilize both unaltered wet pine flats and other similar fire-maintained landscapes, the total area of fire-maintained landscape (both wetland and upland) is critical. Because fire frequency has been drastically reduced in most areas of the Southeast, many animal species that require habitat maintained by frequent fire are threatened or endangered over most of their historic range. Maintenance of a characteristic animal assemblage depends upon: (1) habitat quality within the site (onsite quality), and (2) the quality of the surrounding landscape that provides supplemental resources (landscape quality). Onsite habitat quality can be inferred from the structure and composition of the plant community.

A number of species rely on fire-maintained pine ecosystems of which wet flats are a part. For example, birds and other wide-ranging animals that rely on fire-maintained systems do not appear to differentiate wet pine flats from uplands, as long as both are fire-maintained. Thus, fire-maintained uplands supplement resources available in fire-maintained wet flats and vice versa.