Across the southeastern United States longleaf pine woodlands support a wide range of military training activities and provide suitable habitat for diverse communities of plants and animals, including remarkable numbers of threatened, endangered, and at-risk plant and animal species (TER-S). As a result of historical land uses, large areas previously dominated by longleaf pine now support different forest types, especially on wetter, more productive sites, and active forest management is required to restore them. Restoring longleaf pine on poorly drained sites where there is no remaining natural seed source, and doing so without further losses of any remaining native herbaceous vegetation, is arguably one of the most difficult challenges to restoration ecologists.

The objectives of this project were to investigate (1) the effects of selected site preparation methods on ground layer vegetation and on longleaf pine establishment and early growth and (2) the persistent effects of past plantation establishment on the structure and composition of the ground layer vegetation on sites that historically supported longleaf pine.

Technical Approach

A replicated field experiment was installed and data were collected from established plantations and compared to mostly undisturbed longleaf pine reference sites. All work was conducted on Marine Corps Base Camp Lejeune, North Carolina. A randomized block designed experiment with eight low- to moderate-intensity site preparation treatments used or likely to be used at Camp Lejeune was replicated on 5 blocks, for a total of 40 experimental units. Eight site preparation treatments were applied and consisted of a check (no treatment applied), six combinations of two initial vegetation control treatments (chopping or herbicide) with three planting site conditions (flat [no additional treatment], mounding, or bedding), and a more intense treatment of chopping, herbicide, and bedding. The treatments are referenced as follows: flat or check (F), chopping and flat (CF), herbicide and flat (HF), chopping and mounding (CM), herbicide and mounding (HM), chopping and bedding (CB), herbicide and bedding (HB), and chopping, herbicide, and bedding (CHB). Pine seedlings were monitored for survival and growth and ground layer vegetation and species richness were measured. Results of pine seedling growth were combined with existing longleaf pine growth models to investigate possible site preparation effects on the production of stand structure (longleaf pine size and stem density) suitable for red-cockaded woodpecker foraging. Additionally, plantation sites at least 18 years old were sampled using standardized methods and archived data was acquired for reference sites sampled in the same way. Species composition, environmental characteristics, and measure of species diversity were compared for these two groups.


Longleaf seedling survival, about 70% after 2 years, did not vary among site preparation treatments. In contrast, both the competition control method and planting site conditions showed treatment differences. Compared to chopping, herbicide treatments resulted in greater seedling root collar diameter and height. Similarly, raising the planting surface by bedding or mounding enhanced seedling growth relative to growth in flat-planted sites. Analysis of environmental conditions adjacent to individual seedlings indicated that the benefits were likely related to better control of competing vegetation associated with herbicide treatments, bedding, and mounding. Also, results indicated that excess moisture on poorly drained sites is an important limiting factor for root collar growth. Site preparation treatments that improve drainage, as well as reduce competition for light and other resources, can be expected to maximize longleaf pine seedling growth.

With respect to changes in ground layer vegetation, two main patterns emerged: (1) the effects of chemical application persisted through the study but effects of bedding or mounding diminished, and (2) the patterns of treatment effects on total vegetation abundance and to a lesser degree species richness tracked treatment effects on the woody component or the shrub functional group. Both of these observations are related to the facts that shrubs were the most abundant vegetation group on the sites and that the herbicide treatment formulated to control the shrubs was very effective. Overall, results were consistent with previous studies on flatwoods sites, including the finding that 3 years after site preparation there was no significant treatment effect on species richness and, except for two treatments (HB and HM), vegetation cover was not different from the flat planted plots. Measurement of prescribed fire behavior in the study plots 2 years after planting indicated that site preparation treatments can affect how much of the site burns, how hot the fires burn, and how much fuel remains after burning, all factors related to the effectiveness of prescribed fire for maintaining ground layer diversity. Bedding and mounding tended to reduce temperatures and percent of the area burned, while chopped sites had more uniformly distributed higher maximum temperatures.

Key messages to land managers include: (1) bedding or mounding as applied in this study should have no short-term adverse effect on ground layer vegetation cover, richness, or composition; (2) the herbicide formulation used in this study (imazapyr and triclopyr) broadcast before planting longleaf pine seedlings effectively reduced woody plant cover and had no lasting effects on other plant groups; (3) there was a tendency for all herbaceous groups to benefit from the herbicide effect of reducing the shrubby dominance, shrubs were not eliminated by this treatment, but herbaceous cover tended to increase by the third year after site preparation; (4) prescribed burning appears to be critical for maintaining any benefits to the herbaceous community, and prescriptions for management burns may have to be modified to ensure effective burning in site-prepared locations; and (5) although the plant community richness or abundance was not changed much by site preparation in the first 3 years after treatment, the beds and troughs produced by bedding are expected to persist throughout the age of the plantation and to change the microhabitats within the flatwoods by creating both drier and wetter than average conditions that are likely to favor different vegetation.

Models of plantation development based on seedling growth data and previously published longleaf pine growth models indicated that there can be substantial differences (>20 years) in the time needed to achieve the size and density of trees required for red-cockaded woodpecker foraging. On landscapes where foraging habitat is extremely limited, site preparation choices that promote early growth may be necessary, but managers must balance the negative effects of plantation management on the ground layer foraging habitat standard (40% herbaceous cover) that was not achieved in maturing plantations.

Plantations on average showed lower species richness measured at small spatial scales, but at the scale of 0.1 ha, species richness was not affected by plantation management. In plantations, no species were totally lost or added, but the relative abundance of some characteristic species (Aristida stricta and Gaylussacia dumosa) was significantly reduced in plantations. On average, plantation sites had about 20% herbaceous cover in the ground layer vegetation, about half of the standard established for red-cockaded woodpecker foraging habitat. Reference vegetation averaged 43% herbaceous cover overall. The lack of difference in species richness except at the smallest scales indicates that reasonably diverse communities are maintained in plantations and suggests the potential for restoring a diverse groundcover without adding species. However, a few dominant species apparently are sensitive to habitat modifications created during establishment and growth of plantations. Although thinning the canopy and prescribed burning may invigorate the groundcover, researchers predict that the effectiveness of prescribed burning may be limited by the lack of fine fuels resulting from the significantly reduced herbaceous cover in plantations.


Site preparation potentially may affect TER-S habitat management in the short-term by direct impacts on ground layer and prescribed fire behavior and over a longer time by accelerating the rate of plantation development with its associated loss of herbaceous cover in general and of fine fuel producing grasses specifically. Site preparation choices must balance the need to grow trees quickly with the possible adverse effects on fire behavior in particular. The successful establishment of longleaf pine stands that furnish habitat for TER-S on wet, poorly drained sites will give managers flexibility in simultaneously maintaining defense-oriented training and fulfilling stewardship responsibilities.