Wetlands serve an impressive array of vital functions. Shallow water depth allows an intimate interface of air, water, and soil, and creates an ideal substrate for a multitude of physical, chemical, and biological processes (Mitsch 1992). Wetlands store floodwaters and slow runoff, allowing evaporation or slow percolation of water through the soil. Impurities and pollutants are stored or transformed into less harmful substances. For example, nitrates from agricultural runoff can be converted by microbial action to nitrogen gas, or incorporated into plant and animal tissues. Other pollutants or toxic metals may be bound in sediments or taken up by plants. Large amounts of carbon are stored in plant biomass. Wetlands can also be very fertile ecosystems, providing a source of nutrients for communities downstream. Sunlight, air and water may generate a productive system that exports organic matter in plant wastes. Seeds and other propagules are broadcast in this way also. The interface of air, water, and soil also provides a rich habitat for animal life, from unicellular organisms to crayfish, mollusks, fish, ducks, otters, and black bears. These interdependent influences create some of our most concentrated zones of biodiversity.
Considering their richness and productivity, it is no surprise that wetlands on the North Carolina Coastal Plain have been used for forestry and agriculture for hundreds of years. Forestry and farming, along with other minor uses, have caused the loss or alteration of at least 51% of North Carolina's coastal plain wetlands (Cashin et al. 1992). Rader and Babcock (1989) estimate that 70% of North Carolina's rare and endangered animals and plants are linked to wetlands. In response to growing awareness of the importance of wetlands to human and natural communities, and as the extent of their loss became known, the USDA created the Wetland Reserve Program. This program offers financial assistance to farmers who wish to restore their agricultural fields to wetlands and offers compensation for conservation easements. It represents an important turning point in wetland policy in the United States. This project is intended to develop practical methods to assist in the restoration process.
Microtopography and wetlands restoration
Microtopography in forest floors, and its effect on soil properties and vegetation, has been noted and written about for well over 100 years (see Lyford and MacLean, 1966). It has been shown that this type of disturbance, usually caused by windthrow of trees, is an integral factor of the forest ecosystem. It contributes to both the unique properties of forest soils and the patterns and richness of plant species in forests (Lutz 1940, Barry et al. 1996, Bratton 1976, Ehrenfeld 1995, Stephens 1956). Unevenness in the soil surface may also be also caused by anthropogenic influences such as logging, road ruts, building construction, and other disturbances lost to history. Microtopography is a complex gradient, as is the topography of mountains and valleys. It influences a number of other factors important to plant establishment and survival, such as temperature, moisture, soil chemistry and structure, and degree of exposure and radiation. The mixing of soil organic and mineral layers that occurs as tree roots pull free from the ground has profound implications for plant establishment and survival. There is much to be learned about how this complex interaction of factors can be used to enhance the craft of ecosystem restoration.
Only recently has the use of microtopography been applied to the science of wetlands restoration (McCuskey et al. 1994, Dietz et al. 1996, Barry et al. 1996). A major purpose of this study is to contribute to the body of information available to those who undertake the restoration of farmed wetlands. In particular, methods are developed which farmers who wish to restore agricultural fields, using standard farm equipment, can easily use.
An important process to be observed is the course of vegetative succession and the establishment of a natural plant community. Analyzing the component species and their distribution over time gives direct evidence of the success of the restoration process. Therefore, vegetation analysis also comprises a major aspect of this project.
This study involves two wetlands in the North Carolina coastal plain that had been altered for agriculture. One, near Aurora, N.C. in Beaufort County, is owned by the PCS Phosphate Company. About 10 ha in extent, it has four field ditches, about one m deep and approximately 68 m apart. On one side of the site lies an active agricultural field. The other three sides are surrounded by second-growth hardwood forest with an overstory of swamp chestnut oak (Quercus michauxii), laurel oak (Q. laurifolia), water oak (Q. nigra), yellow-poplar (Liquidambar styraciflua), and white ash (Fraxinus pennsylvanica), among other species.
The other site is near Vanceboro, Craven County. The landowner is enrolled in the Wetlands Reserve Program. The site is 18 ha in extent, with four field ditches approximately 85 m apart and 0.75 m deep. Loblolly pine (Pinus taeda) plantations surround the Vanceboro site on all four sides
Both sites are less than 20 m in elevation and are based on mineral Ultisols. The soil at Vanceboro is Leaf very fine sandy loam, and at Aurora the soil is Roanoke sandy loam. The USDA classifies both soils as hydric, with water input provided by precipitation. Agricultural land modification practices included extensive ditching and draining, forest clearing, and smoothing and grading of the surface to enhance runoff. The soil was then fertilized and prepared for crops such as corn, wheat, and soybeans (Lilly 1981). Both sites had been cropped for many years.
In 1995, earthen clay dams were constructed around each site, and each was divided into eight plots. Each plot is separated by a drainage ditch on one side and a berm on the other. The drainage ditches were sealed and fitted with weirs; water flow is still being controlled and monitored. Two different water table regimes were established at each site, with one half being held at 15 cm above the soil surface, and the other half allowing runoff to 15 cm below the surface.
Within the two water table areas at each site, two of the plots were left with the smooth grade established by agriculture, and the other two were manipulated to produce a rough microtopography surface. This was accomplished with a farm tractor pulling modified disks. Normally, two rows of circular disks are used to prepare the soil; the leading set mounds the soil surface and the trailing set replaces the soil in the resulting depressions, leaving a smooth surface. In preparing the microtopography treatment for this project, the second set of disks was simply removed so that the initial mounds and furrows were left intact. The tractor driver was left to install these in a haphazard manner, creating a fairly random pattern of ridges and pits. This treatment was intended to simulate the microtopographic relief ubiquitous in natural woodlands.
Half of the plots were planted with indigenous hardwood tree seedlings. These included yellow-poplar, water oak, cherrybark oak (Quercus falcata var. pagodaefolia), and swamp blackgum (Nyssa sylvatica var. biflora). The density of the plantings is about 269 trees per hectare. These plantings were intended to enhance the establishment of heavy-seeded species which otherwise might not appear for many years.
In sum, each site has one replicate of about 500 by 40 m in size, of each combination of the treatments: water level, contour, and tree planting.
Figure 1: Experimental site layout
The construction of the study plots cleared the ground, and the course of vegetative succession was set back to its earliest stages. At both sites, long-term cropping depleted the soil bank of "weedy" species, so propagules come largely from surrounding areas. At the Aurora site, early and later-successional species are available in the surrounding fields, roadsides, and woods. Loblolly pine plantations surround the Vanceboro site on all sides. This factor may restrict its complement of successional species. Numerous loblolly seedlings have sprouted on this site. Of course, propagules from remote areas will also arrive through dispersal by wind, water, and animals.
In both sites, it is expected that restoration of hydrology will enable a greater component of wetland species, as opposed to upland species, to regenerate. This expectation has been met by a trend toward wetland species in the first four years of the project (Shear et al., in press, McKinney 1997). Both sites showed treatment effects on vegetation, with more hydric species in troughs and mesic species on mounds. Installation of mounds and pits also created greater variation in soil moisture and hydroperiod within the plots (Tweedy 1988). This variation is expected to produce a more species-rich community in the contoured plots than in the flat ones. McKinneys (1997) analysis indicated that this was true at Vanceboro, but not at Aurora. It is possible that the level of water maintained at Aurora was so high that there was less practical difference between mounds, troughs, and flat areas than at the drier plots at Vanceboro. It remains to be seen whether these trends in species richness have continued.
Previous studies have found that both the hydrologic control and the restoration of microrelief reduce the intensity of surface runoff and soil erosion at these sites. Wetland hydrology restoration was successful at the Aurora site in the first two years of the project, but not at the Vanceboro site (Tweedy 1998, Smith 1998). Possible causes were soil characteristics and efficient, channelized stream drainage at Vanceboro. Tweedy concluded that the influence of microtopography on hydroperiod and water storage was more pronounced in the lower water table treatments.
Although many studies have been made of forest microtopography in general, there was no clear picture of the extent and magnitude of the topography local to these particular study sites. Detailed measurements were made to provide a baseline for determining if the created mounds and troughs are similar to local, natural surfaces. Possibly, indigenous plant species are adapted to a particular degree of soil roughness or elevation. If the created microtopography is significantly different from that common in the area, the data will give guidelines for adjustment of the farm equipment used in microtopography restoration. The disks could be raised or lowered, or a larger or smaller sized implement could be used or perhaps fabricated. In this way, the mound and trough construction methodology can be fine-tuned.
Mature wooded areas adjacent to the study sites were searched for examples of microtopographic relief. Two sites were selected, and each was designated as a control area corresponding to one of the experimental sites. Ten-meter square plots were laid out in the experimental sites, configured to avoid ditches, berms, and study transects. The configuration was duplicated in the forest plots, to minimize variation between the experimental and control sites. In addition, four 10 ´ 10 plots were set up in the flat treatments of each experimental site. These plots were used to compare with the microtopography treatment plots, to further test the effectiveness and magnitude of the contouring techniques.
At each tenth-hectare plot, a tripod with a laser chalkline (revolving laser dot) was set up. A survey rod was used to measure the distance from ground surface to the reference height indicated by the laser. A grid was laid out in 1/3 by 1/3 meter increments, and elevation points were taken at each 0.33 m intersection. In this way, the magnitude of heights and depths of mounds and troughs, and their positions, extents, and locations within the plots were recorded. Approximately 38,000 data points were gathered.
Previous vegetation data were collected in May 1995, and in August 1995, 1996, and 1998. Collection was continued in this phase of the study. Transects of 0.2 m2 quadrats were established on the experimental sites, with two transects laid out across the flat treatment plots, and four across the contoured treatments. The size of the quadrats was calculated to include only one microtopographic feature. A surveyors level was used to assign each quadrat to one of the following six categories: mound, lower mound, flat, upper pit, pit, and puddle (a pit area with no outlet for water flow). A total of 908 quadrats was set up at Aurora and 928 at Vanceboro. The same transects and quadrats have been used in all analyses.
Within each quadrat, plant species were identified and percent cover recorded. In addition, the percent cover of bare ground and dead plant material was recorded.
Vegetation data will show the degree of preference for different elevations of the soil surface. If plants are non-randomly distributed across the mounds, pits, and flat areas, then there is a significant influence of microtopography on plant species distribution. Comparison of species distributions between rough and flat topography treatments will also indicate degree of influence of the topography treatments. In addition, species richness between rough and flat areas will be closely examined.
The component of wetland species (as defined by the USFWS) will also be assessed from these data. Comparison will be made to data gathered up to 1998. The progress of development of a wetland plant community will be analyzed using the combined data.
Initial analysis of microtopography data involves formatting for visual inspection. General differences in experimental contour and flat plots and control forest plots are often quite visible, and give an intuitive idea of where to proceed with further analysis. The following graphics suggest the scope of variation in the ground surfaces.
Some questions that are to be explored are: Do experimental and natural plots have the same size, number, and distribution of mounds and pits? Are the mounds and pits of the same magnitude in height and depth? And finally, by comparing experimental contour and flat plots, it can be seen whether the contour treatments actually made a significant difference in the soil surface.
Figure 5: Tree positions in relation to microtopography in forest plots. Trees are marked in red. Tick marks are in meters.
As an additional study, the positions and diameters of the trees in the forest topography plots were recorded. By superimposing this information onto two-dimensional contour graphs, the positions of the trees in relation to microtopography can be assessed. Trees are expected to show a preference for mounds in moist soils.
Barry, W.J., A.S. Garlo, and C.A. Wood 1996. Duplicating the mound-and-pool microtopography of forested wetlands. Restoration and Management Notes 14(1):15-21.
Beatty, S.W. 1984. Influence of microtopography and canopy species on spatial patterns of forest understory plants. Ecology 62: 1406-1419.
Bratton, S.P. 1976. Resource division in an understory herb community: responses to temporal and microtopographic gradients. The American Naturalist 110 (974):679-693.
Cashin, G.E., J.R. Dorney, and C.J. Richardson 1992. Wetland alteration trends on the North Carolina coastal plain. Wetlands 12(2):63-71.
Dietz, K.B., J.A. O'Reilly, G.S. Podniesinski, and D.J. Leopold 1996. Rebuilding microtopography and planting woody species restores abandoned agricultural land. Restoration and Management Notes 14(2):171-172
Ehrenfeld, J.G. 1995. Microsite differences in surface substrate characteristics in Chamaecyparis swamps of the New Jersey pinelands. Wetlands 15(2):183-189.
Lilly, J.P. 1981. The blackland soils of North Carolina: Their characteristics and management for agriculture. North Carolina Agricultural Research Service Technical Bulletin No. 270.
Lutz, H.J. 1940. Disturbance of forest soil resulting from the uprooting of trees. Yale University School of Forstry Bulletin No. 45.
Lyford, W.H. and D.W. MacLean 1966. Mound and pit microrelief in relation to soil disturbance and tree distribution in New Brunswick, Canada. Harvard Forest Paper 15.
McCusky, S.A., A.W. Conger, and H.O. Hillestead 1994. Design and implementation of functional wetland mitigation: case studies in Ohio and South Carolina. Water, Air, and Soil Pollution 77(3):513-532.
McKinney, J.D. 1997. Effect of microtopography on the natural regeneration of prior converted wetlands. M.S. thesis, Department of Forestry, North Carolina State University.
Mitsch, W.J. 1992. Application of ecotechnology to the creation and rehabilitation of temperate wetlands. In: Ecosystem Rehabilitation, volume 2: Ecosystem Analysis and Synthesis. M. K. Wali (ed.) SPB Academic Publishing, The Hague, The Netherlands. pp. 309-331.
Rader, D.L. and E.L. Babcock 1989. Carolina wetlands: our vanishing resource. North Carolina Environmental Defense Fund. EDF, Inc.
Shear, T., J.D. McKinney, and S. Fraver. Microtopography and plant establishment on restored forested wetlands of North Carolina's coastal plain: preliminary results. Restoration Ecology, in press.
Smith, B.M. 1998. The effect of microtopographic relief and soil temperature on the restoration of prior converted wetlands. M.S. thesis, Department of Soil Science, North Carolina State University.
Stephens, E.P. 1956. The uprooting of trees: a forest process. Soil Science Society of America Proceedings 20:113-116
Tweedy, K.L. 1998. Hydrologic characterization and modeling of two prior converted wetlands in eastern North Carolina. M.S. thesis, Biological and Agricultural Engineering, North Carolina State University