Introduction

To “design with nature” is a concept first introduced by Ian McHarg in 1969 (McHarg 1992). He eloquently advocates full understanding, analysis, and consideration of every aspect of nature and ecosystems during the planning process. “[Man] must become the steward of the biosphere. To do this he must design with nature” (McHarg 1992). This concept should be taken beyond general development and planning to include landscape design. Often, soils in urban areas, roadside corridors, and other anthropogenically altered sites do not receive adequate attention in relation to landscape planting (Jim 1998). Before the design process begins, a thorough analysis of the nature of the site must evaluate historic land use and soil characteristics in order to efficiently and effectively choose plant material that will develop into a sustainable, successful design. Soils in urban or disturbed settings need special consideration because their unique, unnatural characteristics often pose potential stresses on landscape plants (Pickett et al. 2001; Craul 1992). For example, a study in urban Hong Kong evaluated soil samples of varied natural and disturbance history in roadside tree-pit sites with histories of poor tree health and survivability (Jim 1998). Baseline information was established on soil limitations to amenity landscape vegetation based on assessments of extensive ranges of physical and chemical properties. The study concluded that the “fundamental requirements of plant growth are the same for natural or human-modified soils, for they do not recognize artificial city-countryside boundaries” (Jim 1998). Therefore, one must take into consideration the unique limitations of urban soils when choosing plant materials for landscape designs and planning management practices. Although choosing native plants for restoration or revegetation efforts is usually encouraged, at disturbed sites plants “native” to a certain region may not be the most suitable choice. For example, altered soil conditions in urban environments are often typified by extreme compaction, which limits the ability of many plant species to survive. Therefore, characteristics and tolerances of both native and exotic species should be considered when designing on altered landscapes.

Any time soil is drastically disturbed (e.g. grading, construction), natural soil horizons may be modified or removed from the site, resulting in lower quality soil for ornamental plant development (Campbell and Varga 1993). Soil testing should be conducted prior to plant selection to determine soil physical and chemical parameters that may affect plant survivability (Cogliastro et al. 1997). After a full examination of soil and environmental conditions, plants most adapted to the soil conditions may be selected and effective management practices (e.g. fertilization, aeration, irrigation) may be planned to produce high quality, aesthetically pleasing, sustainable landscapes. Analysis of soil physical and chemical characteristics prior to landscape or crop plantings is essential not only to ensure optimal plant health, quality, and yield, but also to protect environmental quality.

Poor agricultural practices in the Piedmont during cotton farming led S.W. Trimble in 1972 to characterize the southern portion of the Piedmont as one of the most severely eroded agricultural areas in the United States (Porcher and Rayner 2001). Cultivation year after year and poor tillage methods, coupled with limited soil conservation management practices, quickly advanced soil erosion rates. Erosion was further enhanced by the general topography of the Piedmont region, with slopes generally ranging from 6 up to as high as 25%. Based on comparisons of existing versus expected soil profiles, Trimble estimated that an average of 25 cm of soil had eroded in South Carolina, with more than 30 cm removed in some larger areas. Ultisols of the Piedmont are by nature easily erodible because they are highly weathered and friable. Abundant rainfall in the region throughout the year, including times when the land is unprotected by vegetation, further promote erosion. Trimble also estimated that 40% of the Piedmont in South Carolina was so severely eroded leaving the topsoil so depleted that it was rendered useless for agriculture (Porcher and Rayner 2001; Richter and Markewitz 2001). The highly eroded, infertile soils were, however, suitable for growing pines, especially loblolly and shortleaf. Unfortunately, the altered soils proved to be too poor for the natural regeneration of the oak-history forest ecosystem that once characterized the landscape.

This site analysis was conducted to examine historical land use change and its effects on soil properties as a growing medium for a greenway planting. After taking into consideration soil characteristics and limitations, a sustainable, site-appropriate landscape design for a public-use greenway was created and implemented to provide an effective and attractive buffer. The greenway will provide seasonal interest with a variety of ornamental trees and shrubs, as well as native grasses and forbs that will provide an attractive entrance to Clemson University. The greenway will also function as a visual buffer for future cuttings of the loblolly pine stand, and provide a future pedestrian pathway incorporating interdisciplinary educational opportunities, and wildlife habitat.

Materials and methods

Description of the site

The site is along the western edge of New Hope Road, in Pickens County, South Carolina (34° 39′ 38″ N, 82° 48′ 58″ W; elevation 225–244 m (740–800 ft) MSL). The road is approximately 610 m long and connects Old Stone Church Road with US Hwy 76/SC Hwy 28 (Fig. 1). Loblolly pine (Pinus taeda) was the dominant vegetation on the site from 1941 to 2003, when the mature tree stand was clear-cut harvested. Although most of the harvested area was replanted with loblolly pine in March 2005, a 30 to 45 m wide roadside buffer corridor was allocated for a public greenway and streetscape. The scenic corridor runs the length of the road and allows for an effective buffer comprising of evergreen and deciduous tree and shrub species, native grasses and forbs.

Fig. 1
figure 1

Photographs of the New Hope Road corridor site (2005)

Full size image

The Piedmont area of South Carolina has a temperate climate, receiving an average of 13,200 mm of rainfall per year (Soil Survey Staff 2006). Ultisols are the dominant soil order present in the Piedmont region (Soil Survey Staff 2006). The soil series present at the site are Cecil clay loam (fine, kaolinitic, thermic Typic Kanhapludults), characterized by severely eroded soils with 6 to 10% slopes; Cecil sandy loam (fine, kaolinitic, thermic Typic Kanhapludults), eroded with 6 to 10% slopes; Pacolet fine sandy loam (fine, kaolinitic, thermic Typic Kanhapludults), eroded with 10 to 25% slopes; and Pacolet clay loam (fine, kaolinitic, thermic Typic Kanhapludults), severely eroded with 10 to 25% slopes (Fig. 2; Soil Survey Staff 2006). The Cecil and Pacolet series consist of very deep soils, formed in residuum weathered from felsic, igneous and high-grade metamorphic rocks, on the ridges and side slopes of the Piedmont uplands (Soil Survey Staff 2006). The Pacolet and Cecil soil series are well drained, with medium to rapid runoff, moderate permeability, and medium internal drainage (Soil Survey Staff 2006). Both soil series are characterized by their very strongly acidic to slightly acidic soil horizons (Soil Survey Staff 2006).

Fig. 2
figure 2

Sample locations and soil map generated from the National Resource Conservation Service (NRCS) Soil Survey Geographic (SSURGO) database

Full size image

Soil sampling

Soil samples were collected from the site at nine different locations (Fig. 2), with ten sub-samples collected at each location. Samples 1 through 8 were located within the disturbed planting site, while Sample 9 was located nearby in an unaltered, mixed hardwood and pine forest to serve as a comparative undisturbed sample. The sampling locations were marked with flags and each sample point was recorded using a Garmin GPSMAP 76 Global Positioning System (GPS) unit (Garmin International Inc., Olathe, KS). Data from the GPS unit were downloaded and converted to a Geographical Information System (GIS) point file format with a free software package, Garmin GPS Extension, developed by the Minnesota Department of Natural Resources (2005). Maps were produced from aerial photography and spatial data layers from Pickens County GIS Department using Version 9.1 of the ArcGIS software package (ESRI 2004, Redlands, CA). Soil samples were collected at the surface from a depth of 15 to 20 cm. The ten sub-samples from each sampling location were combined in a clean plastic bucket (one per sampling location), mixed thoroughly to homogenize each composite sample, and allowed to air-dry overnight. Approximately 200 g of soil was removed from each bucket for subsequent soil chemistry analysis at the Clemson University Agricultural Services Soil Testing Laboratory.

Soil analysis

Laboratory soil tests were conducted to determine soil and buffer pH; organic carbon and nitrogen contents; nitrate–nitrogen; extractable phosphorus, potassium, calcium, magnesium, zinc, manganese, copper, boron, and sodium; lime requirements and recommendations; cation exchange capacity (CEC); acidity; and percent base saturation. An extensive description of all soil test and quality control procedures is available via the test laboratory website (http://www.clemson.edu/agsrvlb/procedures2/interest.htm). After drying, the samples were crushed and homogenized by grinding and screening through a 10-mesh (2-mm) screen before analysis.

The pH of all soil samples was determined by equilibrating 20 g of each soil with 20 ml of deionized water for a minimum of 1 h and then measuring the pH with a calibrated AS-3000 Dual pH Analyser. Buffer pH (i.e. lime requirement) was determined for these same samples using the Adams–Evans buffer method (Moore and Franklin 2002) and the pH analyzer. Extractable nitrate–nitrogen (NO3−N) was quantified using an aluminum sulfate solution and subsequent determination of nitrate with an ion specific electrode as explained by Dahnke (1971). The nitrate extracting solution was prepared by dissolving 173.2 g Al2(SO4)3·18H2O, 12.8 g H3BO3, and 0.7222 g KNO3 in 10 l of deionized water with the final pH adjusted to 3.0 with NaOH. Twenty grams of each soil sample was equilibrated with 40 ml of the nitrate extracting solution for a minimum of 1 h before filtering the extracts and measuring for NO3−N concentrations using a calibrated specific nitrate ion electrode. Mineral analyses (P, K, Ca, Mg, Na, Zn, Mn, Cu, B) were determined using a Mehlich No. 1 extraction solution and element quantification by inductively coupled plasma atomic emission spectroscopy (ICP-AES) (Jones 2001). Total carbon (C) and nitrogen (N), equivalent to organic C and N for these particular soil samples, were measured on 200-mg samples using a combustion analyzer (Elementar Vario Macro, Mt. Laurel, NJ).

Historic land use analysis

The New Hope Project is located in the Clemson Experimental Forest, which is the result of one of the federal land reclamation projects initiated by the Roosevelt administration in the 1930s. Aerial photography associated with the Land Use project taken June 5, 1938 (PI-2-5) shows the areas of the New Hope Project as cultivated land. Forest and stand records indicate that the tract was planted to loblolly pine in the early 1940s and managed for timber production thereafter. A clearcut harvest of the stand was completed in 2003 and the stand was again planted to loblolly pine except for a buffer along New Hope Road to be developed as a scenic buffer of hardwoods and shrubs. A scanned section of the 1938 photo was geo-referenced, using the Geographic Information Systems (GIS) software ArcMap 9.1, to 2002 color digital ortho-photography acquired from Pikens County, SC. Photographs were taken from a light aircraft in November, 2005 and similarly georeferenced to the Pickens County photos. In addition to the historical land use information available from archival documents and literature, aerial photographs were taken in 2002 and 2005 and compared with archived photographs to determine the land use history of the site as it changed from agricultural to forestry, with anthropogenic disturbances increasing with time (Fig. 3).

Fig. 3
figure 3

Aerial photographs of the New Hope Road corridor site from the years 1938, 2002, and 2005

Full size image

Analysis of potential landscape plant species

Extensive research was conducted on several promising plant species for potential incorporation in the landscape design of the site. In order to select potential landscape plant species capable of surviving the site’s harsh, disturbed conditions, respectable sources (Adams 1994; Burns and Honkala 1990; Darke 1999; Porcher and Rayner 2001) and the USDA PLANTS database were used to compile phytogeographical and growth characteristics information for the trees, shrubs, native grasses and forbs chosen in the preliminary design (USDA, NRCS 2005). The designer used native or naturalized plants when available and suitable for the site. Some nursery cultivars developed to survive harsh conditions and provide more attractive flowering and foliage were also used. The plant list and layout were reviewed by several faculty and landscape professionals on campus. The plant list was also checked against the USDA’s invasive plant list to reduce the potential of introducing or spreading problem plants. To predict the ultimate survivability of the various plant species in the landscape design, their critical soil and environmental requirements were evaluated. Since the site varies in levels of soil compaction, soil types, and moisture levels and drainage capabilities, plant anaerobic tolerance, drought tolerance, average root depth (in optimal conditions), CaCO3 tolerance, and moisture requirements were recorded. Other characteristics noted were nativity, common soil orders, optimal pH, and landscape features. Final selection of plant species and location in the design were determined based on the plants’ suitability for the site’s various microclimates and soil characteristics. A combination of exotic and native trees, shrubs, grasses and forbs were incorporated into the design to create a dramatic entrance to Clemson University and provide a beautiful, yet functional greenway buffer.

Results and discussion

To effectively “design with nature,” an interdisciplinary approach must be coordinated to adequately analyze the nature of the site. Both the natural and anthropogenically-altered characteristics of an environment should be evaluated to gain a full understanding of the site’s capabilities and limitations. Effective communication among the interdisciplinary team members provides a framework that demonstrates the holistic qualities of an environment so that a successful design can be created that both meets user needs and respects the natural qualities of the site. Figure 4 demonstrates how experts in the fields of soil science, horticulture, geographic information systems (GIS), landscape architecture and planning, native plant ecology, and forestry have worked together on this site to combine historical land use data and soil and environmental characteristics to develop a sustainable landscape design that will function as a vegetative buffer as well as an attractive public greenway and educational site.

Fig. 4
figure 4

Theoretical framework to “design with nature”

Full size image

The Clemson Experimental Forest, along with some 250 other federal land conservation projects, resulted from a plan to rehabilitate sub-marginal farmlands and put them into more compatible uses such as forests. These projects were initiated in the 1930s and consolidated under the “Bankhead–Jones Farm Tenant Act” (Title 7, Chapter 33, Sec. 1010). Clemson University became custodian and took responsibility for the Clemson Land Use Project by deed in 1954. The subject property for the scenic corridor landscape design project is a part of agricultural land purchased from Clyde Patterson, Gdn. and Birdie L. Patterson, et al. January 26, 1940 and recorded in Oconee County Deed Book 4V page 520 and identified as SC-LU-3-17012 and contains approximately 34 ha (84 ac). The Works Progress Administration planted old-field portions of the tract with loblolly pine in 1941, shortly after purchase. Since 1941, the site has been forested with a loblolly pine stand for timber production, wildlife habitat, recreation, and education. The pine plantation was thinned in 1959, 1964, 1970, and 1998. Southern pine beetle (Dendroctonus frontalis) infestation coupled with severe drought influenced the decision to clear-cut harvest the 11 ha (27 ac) mature loblolly pine stand adjacent to New Hope Road in 2003. The harvested area, with the exception of the scenic corridor project site, was replanted with loblolly pine in March 2005.

In addition to researching historical documents and other literature, aerial photographs were analyzed to document the land use history of the site. As shown in Fig. 3, in 1938 the site was used for agriculture, primarily cotton farming. Terraces, which are commonly used in agricultural fields with significant slopes, are clearly shown in the aerial photograph. Poor soil conservation practices and intensive cultivation practices eventually left the site with severely eroded and infertile soils, common to most agricultural sites in the region.

Soil test results for the disturbed and undisturbed soil samples are shown in Table 1. The mean soil pH of the eight disturbed samples was acidic at 4.7, and all of them exhibited a pH ≤ 4.9. Conversely, the undisturbed sample had an acidic but significantly higher pH (5.1). Buffer pH values of all samples were consistently 7.6–7.7, with no statistical difference between the undisturbed and disturbed samples. The eight disturbed soil samples all had very low extractable P, with a mean value of 4.8 kg/ha. Phosphorus levels for the undisturbed soil, however, were nearly eight times higher. Potassium levels were generally low for all samples, with only Sample 7 exhibiting a medium level of K. Calcium levels were also generally low for the disturbed soil samples, but were very high (e.g., 7 to 15 times higher) for the undisturbed soil sample. Mg levels for all samples ranged from medium to high with no statistical difference between the undisturbed and disturbed soils. Although the C and N contents of the undisturbed soil sample were slightly higher than the corresponding mean values for the disturbed soil samples, the values were not statistically different. Many of the other measured parameters showed no statistical differences between the undisturbed versus disturbed soil samples (Table 1). Interestingly, the level of Mn in the undisturbed soil sample was statistically lower than the mean value for the disturbed soil samples. From the results presented in Table 1, the most noticeable differences between the undisturbed and disturbed soil samples were the statistically higher values for Ca, P, and CEC for the undisturbed sample. Taken together, the data demonstrate the acidic and relatively low fertility of the surface soils at the site.

Table 1 Soil testing results
Full size table

Examination of deep soil profiles revealed the lack of a highly developed topsoil layer due to erosion, and subsoil layers dominated by highly compacted, heavy clays. These qualities make growing conditions extremely difficult for most vegetation, especially transplanted nursery stock material that will be used to landscape the greenway. Native and exotic trees, shrubs, grasses and forbs that are more tolerant of these harsh, compacted conditions were chosen for the landscape design to minimize the risk of poor plant vitality and death after planting. It was imperative to select high quality trees from the nursery to ensure good plant health and to practice appropriate planting techniques during installation to help ensure the viability of the installed plants. Prior to planting the trees, wide, shallow holes were dug five to six times the width of the root balls. The recommended width of the planting hole should be at least two to three times the width of the root ball. The expanded width of the planting holes at this site will aid in the establishment of the trees because the native backfill soil is much less compacted than the surrounding site. After planting, all trees and shrubs were mulched with 8 cm of organic mulch to minimize moisture loss and compaction, minimize weed competition, regulate soil temperatures, and improve soil conditions. During the establishment phase, it will be imperative to maintain regular moisture levels by irrigating or manually watering the vegetation thoroughly. Similarly, it will be important to provide the planted species with adequate nutrients, particularly for the low fertility soil conditions associated with this particular site. For example, Table 2 shows fertilizer recommendations based on the soil test results at each sample location for trees, shrubs, and perennial flowers as reported by the Clemson University Agricultural Services Soil Testing Laboratory. Once established, species adapted to soil and environmental characteristics comparable to the site should survive, precluding certain climatic abnormalities (e.g. extended drought, extremely wet seasons).

Table 2 Fertilizer recommendations for disturbed soil sample locationsa
Full size table

Based on historical land use, soil analysis, and plant species’ characteristics, a landscape design was created addressing environmental opportunities and constraints (Table 3). The landscape design for the New Hope Road corridor was created to provide an attractive streetscape and recreational corridor that will provide bicycle and pedestrian trails, seasonal interest, and visual appeal while helping buffer the harvest and future forest management activities on Clemson University Experimental Forest land adjacent to the road (Fig. 5). The design integrates both user and site needs to develop an aesthetically pleasing greenway buffer comprised of plant species and elements of design that enhance the site while simultaneously working with its unique limitations. The development of the design was based on McHarg’s “design with nature” philosophy (McHarg 1992), in which urban and landscape planning revolve around a sites’ individual characteristics, and respect is given for these factors when designing a space unique to each site. A preliminary landscape design was developed for the site incorporating conceptual ideas from an undergraduate landscape design class and Experimental Forest land managers (Fig. 5). Large trees were utilized to provide shade, screening, accent, and scale along the roadway. Smaller understory trees and shrubs add seasonal interest through flowering, fruiting, leaf color, bark color, and texture. A variety of native and exotic species provide visual interest throughout the seasons. The naturalistic, curvilinear design form reflects the rural character of the surroundings and becomes more structured near intersections and buildings. The design is intended to be experienced by motorists on New Hope Road as well as by bicyclists and pedestrians utilizing the path system which weaves throughout the plant groupings. Trees were placed to alternately screen, buffer, and provide glimpses into the newly reestablished site while framing views of the dairy barns, cow pastures, and uncut forest. Both woody plants and native grasses have been incorporated into the design for erosion control, wildlife food and habitat, and visual interest throughout the year. After the site’s inherent qualities were fully documented by the interdisciplinary team, the preliminary plant list was thoroughly analyzed and altered to meet the needs of the land and its users. Certain needs of the plants in the design, such as optimal pH, soil moisture, and anaerobic and drought tolerances, were analyzed to determine their suitability at the site (Table 3).

Fig. 5
figure 5

Preliminary New Hope Road corridor landscape design

Full size image
Table 3 Plant species considered and their optimal soil property requirements (compiled from Adams 1994; Burns and Honkala 1990; USDA and NRCS 2005; Porcher and Rayner 2001; Darke 1999)
Full size table

Most of the trees and shrubs listed in Table 3 have optimal pH ranges that are inclusive of the average pH at the site (4.7), with only a few having optimal pH ranges slightly above this value. Only Dawn Redwood (Metasequoia glyptostroboides) and Deodar Cedar (Cedrus deodara) prefer pH levels above 6. Therefore, it was decided to replace these species with Eastern red cedar (Juniperus virginiana) and American Holly (Ilex opaca), native evergreen species that better tolerate the more acidic soils. The native grasses and forbs generally prefer soil pH levels between 5 and 6.

The majority of tree species originally considered prefer moist, well-drained soils. For example, Bald cypress (Taxodium distichum), Dawn redwood (Metasequoia glyptostroboides), Fringetree (Chionanthus virginicus), River birch (Betula nigra), and Bushy bluestem (Andropogon glomeratus) prefer higher moisture levels. In contrast, Crepe Myrtle (Lagerstroemia indica) and Deodar Cedar (Cedrus deodara) favor drier soils. The native grasses and forbs generally require moderate moisture levels, but can tolerate drought once established. Moisture levels and aerobic conditions vary throughout the site. More moist, anaerobic conditions occur in the low-lying zones and areas near drainage pipes that convey storm water runoff from the northern side of New Hope Road onto the site, while drier, aerobic conditions characterize the areas with higher elevation and/or coarser soil texture. Therefore, species tolerant to anaerobic conditions and requiring more moisture were grouped in the areas that naturally retain moisture, while the species lacking tolerance to anaerobic conditions and requiring drier soils were grouped throughout the naturally drier sites to benefit the health of the plants and reduce onsite maintenance. While most of the planting site is eroded uplands, the contour is such that water drains across the site at several locations. Two culverts under New Hope Road add additional water to these areas. Plants needing mesic conditions were concentrated in these drainage areas.

Conclusions

A history of intensive agricultural land use and managed forestry along New Hope Road in Pickens County, SC, has noticeably altered the soil qualities. Ultisols are prevalent throughout the Piedmont region and are characterized by their low nutrient and high clay contents, high bulk density, and acidic nature. Severe soil erosion and compaction brought about by a combination of anthropogenic and natural causes has left much of the Piedmont with infertile soils lacking a true topsoil layer. These qualities present numerous challenges when attempting to modify and install vegetation. The existing natural conditions must be carefully analyzed and respected when “designing with nature” to create successful, aesthetically pleasing, and sustainable environments. Soil analyses provide site- and plant-specific fertilizer recommendations for successful establishment and maintenance of a landscape design.