South Florida Slash Pine (Pinus elliottii var. densa)
Mortality Following Three Seasons of Prescribed Burns

James R. Snyder and Holly A. Belles
U.S. Geological Survey
Florida Integrated Science Center

Presented a the 45th Symposium of the International Association for Vegetation Science,
March 3-8, 2002, Porto Alegre, Brazil

Location

• This study is being conducted in Big Cypress National Preserve, which covers 295,000 ha of subtropical southern Florida, USA (Figure 1).  Big Cypress is contiguous with Everglades National Park to the south.
   
• Big Cypress has a large prescribed fire program, averaging over 15,000 ha of management-ignited fire per year.
   
• The vegetation in the study area consists of a mosaic of pinelands and cypress domes.
   
• The soils are shallow sands (pH 6.7-6.9) overlying limestone. 
   
• The pine forests are often called wet pine flatwoods because there is standing water for a few months during the rainy season.
   
• The sole overstory tree of the pinelands is South Florida slash pine (Pinus elliottii var. densa). This variety is more fire-adapted than the more widespread and commercially important var. elliottii and shows ecological similarities to longleaf pine (P. palustris).
   
• Cabbage palm (Sabal palmetto) and saw palmetto (Serenoa repens) can dominate the pineland shrub layer. Hardwoods are relatively sparse. Cypress trees (Taxodium ascendens) form dome-shaped stands in bowl-shaped depressions in the limestone and are also scattered in pinelands.
   
• The herb layer in the pinelands is dominated by grasses such as Muhlenbergia capillaris, Schizachyrium rhizomatum, and Paspalum monostachyum.
   

Figure 1.  Location of the study area.

South Florida Fire Environment

• South Florida experiences highly seasonal rainfall with >75% of annual mean of 135 cm falling during the May to October rainy season.  Progressive drying during the winter leads to the driest conditions in April and May. Lightning activity begins in May with advent of summer thunderstorms, peaks in mid-summer, and ends in October. 
   
• The greatest area burned by lightning-caused fires is early in the rainy season (Figure 2).
   
• Human-caused wildfires increase in number and size during the dry season and also burn the greatest area in May.
   
• Humans have inhabited South Florida for at least 5,000 years and probably supplemented lightning-ignited fires the entire time.
   
• Wildland fires can occur at any time of the year, but the least area burns in September and October when water levels are highest.
   

Figure 2.  Monthly distribution of area burned by cause. Data from Everglades National Park, 1948-1997.

Study Objective

• A wide range of potential fire regimes is feasible in South Florida pinelands. The main objective of this project is to establish the baseline vegetation conditions and begin the experimental treatments for a long-term study of season and frequency of burning. 
   
• The research will provide detailed data on responses, such as changes in species composition and dominance relationships, to different burning regimes.
   
• The results will be considered along with wildlife, public safety, and other management concerns in refining the prescribed burning program at Big Cypress National Preserve and other South Florida public lands. 
   
• Here we report some early results on fire-induced mortality of pines.
   

Methods

• The long-term study involves six experimental treatments, combinations of 3 seasons and 2 frequencies of prescribed burning.
   
• The three seasons are winter (January-February, dry season, plants most dormant), spring (May-June, early in wet season, conditions relatively dry), and summer (July-August, middle of wet season, conditions wet).
   
• The short frequency is every three years, and the long frequency is every six years.
   
• The study area is divided into 18 experimental units, each containing at least 50 ha of pineland.  There are three replicates of each treatment, one replicate burned per year for three consecutive years.
   
• Each unit has three 1.0 ha tree plots with all trees >5 cm diameter and palms with stems >1.4 m tall mapped and tagged.
   
• Experimental burns are carried out by the National Park Service fire management staff.
   
• Fire behavior measurements to characterize burns include pre- and post-burn fuel sampling to calculate fuel consumption and measurement of fire temperatures with temperature-sensitive paints.  We also estimate height of bark char and percentage of the needles scorched on all the pines.
   
• Pine tree mortality is assessed one year after the burns.
   

Results

• The prescribed burns began in spring 1996, all 18 of the experimental units have received their initial treatment burns, and the second application of the frequent treatments is underway.
   
• A total of 16,370 trees have been tagged and mapped.
   
• The mean density of pines is 227 ha^-1 (range 72-438). The plots also contain an average of 52 cabbage palms and 24 cypress trees.
   
• The largest pine tree in each plot ranged from 32 to 52 cm dbh.  The mean pine basal area was 8.4 m²/ha (range 4.2 to 14.0).
   
• Study-wide, 2.8% of the pine trees were found dead one year after burning, excluding those obviously killed by other means such as lightning or windthrow.  Due to high variability from plot to plot, seasonal differences in mortality were non-significant (Table 1).
   
• There were significant differences among seasons in several measurements of fire behavior (Table 1).  Winter burns had the most severe values in all cases. In spite of this it does not appear that these burns are most stressful to pine trees. Summer burns had the lowest measures of fire severity, except for crown scorch. One of the problems with summer burning in South Florida is the lack of dependable winds.
   
• While mortality did not show a significant relationship to season of burning, mortality was significantly related to several measures of fire behavior at the plot level (Table 2).
   

Figure   4.  Size-class distribution of pines before burns and those dead one year after burning.

• As expected, smaller trees were more likely to die than larger trees (Figure 4).  The mean diameter of trees that died was 12.0 cm compared to 20.3 cm for those trees still alive.
   
• Crown scorch appears to be a better predictor of mortality than height of bark charring. At the individual tree level, dead trees averaged 90% scorch while those that survived averaged only 40% scorching. South Florida slash pine has a high tolerance for scorching--there was nearly 90% survival of the trees whose entire complement of needles was killed in the fire.
   
• When the data from additional burns are available, a more thorough analysis accounting for variability in size-class distribution in different units may show a significant effect of season.  In the interim, there is little indication that impacts on mature trees should influence the season in which prescribed burning is conducted in South Florida pinelands.
   

Acknowledgements

This project is a cooperative effort between the U.S. Geological Survey, Biological Resources, and the National Park Service, Big Cypress National Preserve. Marlena Hovorka assisted with all field portions of the study and Frank Partridge, Dana Schulze, and Scott Cooper assisted in poster production.