black bar with title, National Park Service, U.S. Department of the Interior and the arrowhead

History & Culture Home

History & Culture
History and Culture title graphic

Earthwork Husbandry:

Assessment of the Principal Earthworks:

The Federal "Fish Hook" Line, Petersburg, VA


Table of Contents | Next Section

For several years, resource managers of the National Park Service have been involved in an internal debate to determine the "best" way to preserve and manage military earthworks. Before Andropogon's Earthworks Landscape Management Manual appeared in 1989, little else had been published on the subject. Andropogon's work represented a first attempt to define the debate and to offer recommendations for preservation and stabilization. The report stated that "Forest and Tall Grass, which provide the greatest level of stabilization and require the least maintenance, should become the most prevalent cover types over time."(1) The use of high maintenance, non-native grasses (turf grasses) was discouraged. The two types of preferred ground cover-forest and native grasses-are at extremes of the resource management spectrum and dictate vastly different approaches to maintenance and upkeep. This has engendered further discussion over when it is advisable to clear trees and plant grasses or when to leave forest intact.

Our work at the Fish Hook in Petersburg confirms mounting evidence that mature forest cover provides the best protection for the resources. The profile, definition, and clarity of surviving detail found at Fort Welch, the Siege Battery, Fort Fisher, and, to a lesser extent, forts Gregg and Wheaton, are clearly superior to that observed where earthworks have been cleared of trees. This does not imply that all earthworks should be maintained in forest cover, only that the integrity of the resources should be considered before any decision to clear away trees is made. Any abrupt change of land cover will damage the resources. How much damage, or how little, will depend upon many factors, but principally on the impact of the clearing and replanting practices and how long the parapet surface is bared to the elements before a new surface cover is established.

The principal surface cover under forest conditions is leaf litter, which effectively inhibits erosion if left undisturbed. The greatest erosion occurs when all large trees are removed without providing a new parapet cover. Fort Conahey and Fort Urmston are unfortunate examples of earthworks that have suffered erosion as a result of tree removal. With the canopy removed, the leaf litter cover is washed away with the next storm. Erosion sets in, immediately softening the profile, blurring and eventually destroying features, such as gun platforms and embrasures. In extreme cases, one can actually observe segments of the parapet slowly "melting" into the ditch. The erosion ends only when a new cover is established.

Large trees growing on or adjacent to the parapet, however, can uproot and damage the earthwork. Tree throw on the Fish Hook line was not found to be a problem generally, but storm damage could be catastrophic. This suggests that earthworks in mature forest cover cannot simply be left alone as suggested by the manual, but must be actively maintained by selectively removing trees. High-risk trees would be thinned and removed, while low risk trees would be encouraged to maintain enough canopy to shed rainwater and replenish the leaf litter cover. This might be described as a "tree husbandry" approach to maintenance.

Managing Earthworks under Forest Cover (draft) by James Johnson offers some ideas about how to identify high-risk trees. Logically, based on "root zone geography," trees located on the ends of parapets pose the greatest risk, those on top of the parapet are next, and those on the side pose the least threat.(2) From our observations in the field, however, the location of thrown trees appears entirely random. Certain tree species, such as white pine, appear to cause more damage than others when uprooting. More research might determine species and sizes of trees that pose a higher risk, although it may prove that risk factors are related directly to the health of individual trees, which is more difficult to assess over a large area.

Establishing a replacement Tall Grass cover is more problematic. Andropogon suggested invasive methods, such as plugging, staking, or implanting fascines to limit erosion while the new cover "took." These approaches compromise earthwork integrity and should be rejected except in situations where erosion from other causes is already far advanced. The 1996 revision of the manual, entitled Earthworks Landscape Management Field Handbook (draft)(3), recommends an approach that would establish vegetative cover, particularly tall grasses, on the ditch and parapet by thinning woody growth and encouraging native plants. This echoes Andropogon's recommendation to establish "dense stands of native grasses, primarily little bluestem, under a light tree canopy"(4) where appropriate. This "plant husbandry" approach suggests mowing and burning to encourage native grasses in a light forest setting. While this approach may be perfectly acceptable for earthworks that already have been cleared and are maintained in some manner, serious problems may be encountered when moving directly from leaf litter cover to vegetation. First, if the acidic leaf litter remains in place, it will smother the growth of desirable grasses, whether the area is mown or not. The protective cover must be removed before grasses will take. Erosion will occur. Second, burning removes 100% of the protective leaf litter, opening the parapet to erosion on a large scale. This could be disastrous with some soil types if a replacement cover for the parapet, such as mulch, is not immediately provided.

The transition from leaf litter cover to live vegetative cover is the most dangerous time of the process. Grasses need to be sown and cultivated, while maintaining a protective surface cover. The only method that would appear to accomplish both of these objectives simultaneously is hydro seeding. Published research on the efficacy of hydro seeding would make a valuable contribution to earthworks management in the national parks.

Both Morrison and Johnson identify three management categories of earthworks: cleared; intermediately managed; and protected (in forest cover). Management practices vary according to the category. A slight revision of these categories would clarify management intent:

1) Cleared or Display Earthworks - sites that typically are close to visitor facilities or busy parking areas and that receive many visitors. Without some form of maintained cover and ongoing care, the resource would soon erode beyond recovery. In this case, planting in grass cover is a way to retard the deterioration of the earthwork. At the same time, it opens the area to display and more intensive interpretation, appropriate for a heavily visited site. Seriously degraded earthworks may as well be cleared of trees and sown with grasses, as it may be the best means to salvage remaining integrity. In extreme cases, some reconstruction of display earthworks may be justified-rebuilding revetments, gun platforms, and embrasures. This reconstruction is a form of protection in that it covers the resource completely and prevents further erosion. While such remedial work has some interpretive value, the overall effect is no substitute for the authenticity of a well-preserved and unreconstructed earthwork.

2) Selectively Interpreted (intermediately managed) Earthworks - sites that receive moderate, self-guided visitation and that probably are not well patrolled. These earthworks might best be maintained in a mixed cover. Strategic areas would be cleared, planted in grasses, and connected by wood chip trails to guide visitors through the site and away from sensitive areas. The rest of the site would be managed under forest cover to preserve resources, with underbrush serving to limit or discourage access. Over time, features in the cleared areas could be compared to those preserved under forest cover to determine the effect of clearing and visitation upon the resources. Access to these sites might be restricted from time to time to allow vegetation to recover from trampling.

3) Forested (protected) Earthworks - sites that currently are well preserved due to "benign neglect," that receive little visitation, and that would continue in this state with minimal intervention. These sites are set aside for future study and would be maintained under a tree husbandry regime. Forested earthworks would be interpreted, but with guides rather than signage. Guides could conduct small groups of visitors to these sites during winter months when visibility is good and the ground frozen. During summer months, undergrowth would discourage casual visitation.

With the goal of preserving the largest number of earthworks in the best possible condition across all park service holdings, every park with earthworks should attempt to maintain a balance among these three categories. Cleared earthworks are easier to interpret and maintain in areas of high visitation, but every change of land cover exacts some toll on the resources. Forest cover best preserves resource definition and clarity of detail. These details communicate the immediacy and authenticity of the surviving resources. Selective interpretation is a compromise that encourages some visitation, while emphasizing resource protection.

GPS Survey in Fort Fisher
Survey teams from Cultural Resources GIS mapped the earthworks on Petersburg National Battlefield's Fish Hook line in April 1998. The teams used Trimble Pro-XR and Pro-XL Global Positioning Systems receivers that achieve an accuracy of +/- 1 meter after differential correction for 90% of collected positions. The hand held computer was loaded with a data dictionary that enabled surveyors to enter essential attributes of the earthworks features, while capturing spatial coordinates. The field data were used to create thematic map layers in an ArcInfo GIS (Geographic Information Systems) database. Parapet lines were buffered using the recorded average width to generate a polygon coverage, which is the basis for the site maps and area statistics included in this report. These fort polygons are models and may be subject to revision. Draft maps were field checked in May.

1. Andropogon, Earthworks Landscape Management Manual. Washington, DC: Park Historic Architecture Division, National Park Service, 1989; p. D-1
2. James E. Johnson, Managing Earthworks under Forest Cover. National Park Service and Georgia Trust for Historic Preservation [Draft January 1998], p.14.
3. Darrel Morrison, Earthworks Landscape Management Field Handbook. National Park Service and Georgia Trust for Historic Preservation, [Draft, July 17, 1996]
4. Andropogon, p. D-6

Table of Contents | Next Section