Weimin Xi1, Szu-Hung Chen1, Andrew G. Birt1, John D. Waldron2, Charles W. Lafon3, David M. Cairns3,  Maria D. Tchakerian1, Kier D. Klepzig4, and Robert N. Coulson1

1Knowledge Engineering Laboratory, Department of Entomology, Texas A&M University, College Station, TX 77843
2Department of Environmental Studies, University of West Florida, Ft. Walton Beach, FL 32547
Department of Geography, Texas A&M University, College Station, TX 77843
4USDA Forest Service Southern Research Station, Asheville, NC 28804


The southern Appalachian forests are subject to under multiple environmental threats, including periodic fires, insect outbreaks, and more recently, invasion by exotic plants. Past studies suggest that these multiple disturbances interact to shape the species-rich forest landscapes, and hypothesize that altered fire regimes, interacting with increased landscape fragmentation, may exert complex influences on the pattern and process of invasions. However, complete knowledge of the landscape-scale processes that drive invasions, and therefore sound forest management practices to reduce damage, are still unclear.

We have developed a modeling approach to explore the synergistic effects of wildfires and landscape fragmentation on spread of two exotic invasive plants, Princess Tree (Paulownia tomentosa) and Tree of Heaven (Ailanthus altissima). LANDIS-II, a spatial explicit forest succession model, was used to simulate forest dynamics and plant invasion in a xeric pine-oak landscape, which presents the predominant vegetation in southern Appalachian Mountains.

  • The southern Appalachian mountains is one of the most biologically diverse regions in the United States. For any invasive species, it could equally be argued that either:
  • the area might be a barrier to invasion because of increased inter species competition or
  • be at greater risk of invasion because of a greater potential for suitable habitat niches.

  • In the last decade, plant invasion has been an increasing concern. We focused on the low to middle elevation xeric eco-zone, which usually includes xeric pine, pine-oak, and oak forests typically dominated by an evergreen understory.
  • We focused on Princess Tree and Tree of Heaven that are currently considered invasive in this area. Both plant species are prolific seed producers, grow rapidly, and can overrun native vegetation in burned areas once they become established.
The LANDIS-II Model: LANDIS-II is a spatially-explicit computer model designed to simulate forest succession and disturbance across broad spatial and temporal scales (ca. 1~1,000s km2). The model was calibrated to simulate changes in abundance of Princess Tree and Tree of Heaven over a 200-yr period along the gradients of fire frequency and landscape fragmentation in a hypothetical xeric pine-oak landscape. We parameterize a pool of the 33 most dominant tree species plus three invasive tree species.

Fire Parameterization: In LANDIS-II, fire regimes are defined by a combination of fire frequency (fire rotation period) and fire event size. To examine the relationship between fire frequency and species abundance over time, we created 17 different fire frequencies (i.e., fire spread age at 10, 25, 75, 100, 125, 150, 200, 250, 300, 400, 500, 600, 700, 800 ,900, 1000 yrs, no fire), and parameterized fire event size as 75 ha based on published literature and communications with fire experts.

Landscape Creation: We used 3 different proportions of forest cover (10%, 50% & 80%) and used different numbers of forest patches under each proportion setting. Hence, we created 9 initial-community landscapes with different levels of landscape fragmentation. The 36 selected species were randomly assigned to the forest patches (Fig. 2, left plate).

Fig. 1 Comparison (via regression) between the observation values from the field plot data and the simulated abundance using LANDIS-II. The regression slope (1.0669) and R2 (0.96) indicate a good match, and confidence in the species establishment coefficients.

For LANDIS model calibration, appropriate species establishment coefficients (SECs) need to be derived. Field plots (from the Carolina Vegetation Survey) were used to calculate abundance values for each species (i.e. how many plots contained species x). Values were then rescaled and used as SECs in the simulations.




Fig.2. Left plate: The 9 initial-community landscapes with 3 levels of forest covers (10%, 50%, and 80%) and patch numbers (1, 2, 5). The green areas are forest patches; the red areas are non-forest patches. Right plate: An example of LANDIS-II output of abundance of Princess Tree in the 9 landscape settings over 200 years.  
  • Intermediate levels of fire frequencies (ca. between 10 and 300 yrs frequency) promoted spread of the invasive species in the landscape, while both low and high extremes of the fire frequency spectrum limited the spread of the species (Fig.3).

  • Increasing fire frequency promotes the growth of dominant pine species (Pitch pine, Table Mountain Pine & Shortleaf pine), while reducing less fire tolerant White pine (Fig.4).
  • Under the same level of fragmentation, a higher proportion of initial forested patches resulted in an increase in abundance of the invasive species in the landscapes, because of greater seed availability.
  • The synergistic effects of wildfires and landscape fragmentation are complex. While increasing fire frequency (or re-introducing fires) could maintain xeric pines, it may also promote the plant invasion.
  • Landscape models such as LANDIS-II are promising tools for evaluating forest management strategies for invasive plants.
  • Under increasing fragmentation, a trade-off between (low) fire frequencies to minimize plant invasion, and (high frequency) to promote pines in the landscape may be required.
  • Future work will focus on exploring: (1) optimal fire regimes; (2) additional invasive species and other disturbances such as southern pine beetle outbreaks; and (3) the reciprocal interaction of plant invasion and forest composition and diversity.

Carins, M.D., C.W. Lafon, J.D. Waldron, M.D.Tchakerian, R.N. Coulson, K.D. Klepzig, A.G. Birt, and W. Xi. 2008. Simulating the reciprocal interaction of forest landscape structure and southern pine beetle herbivory using LANDIS. Landscape Ecol. 23: 403-415.

Scheller, R.M., J.B. Domingo, B.R. Sturtevant, J.S. Williams, A. Rudy, D.J. Mladenoff, and E.J. Gustafson. 2007. Introducing LANDIS-II: design and development of a collaborative landscape simulation model with flexible spatial and temporal scales. Ecological Modelling 201:409-419.

Xi, W, J.D. Waldron, C.W. Lafon, D.M. Cairns, A.G. Birt, M.D. Tchakerian, R.N. Coulson, and K.D. Klepzig. 2009. Modeling long-term effects of altered fire regimes following southern pine beetle outbreaks. Ecological Restoration 27(1) : 24-26.

Zouhar, K., J.K. Smith, S. Sutherland, and M.L. Brooks. 2008. Wildland fire in ecosystems: fire and nonnative invasive plants Gen. Tech. Rep. RMRS-GTR-42-vol. 6. Ogden, UT: U.S. Department of Agriculture, Forest Service, Rocky Mountain Research Station. 355 p.