The natural forest ecosystems of South Florida, USA, support a high biodiversity of plant and animal species and provide valuable ecosystem services. However, these ecosystems remain poorly represented in global studies, primarily due to a paucity of standardized data. Here, we present previously unpublished data from 332 censuses of 54 permanent 1-hectare tree inventory plots in the Racoon Point area of Big Cypress National Preserve, Florida, USA, including a total of nearly 100,000 measurements (diameter or height) of >17,000 individual living trees and palms (with additional measurements of nearly 6,000 dead pine snags) collected sporadically over a 19-year period (1993 – 2012). These data, which were originally collected as part of a project to investigate tree responses to different experimental burning regimes, provide unique insight into the diversity, composition, structure, and dynamics of South Florida’s unique and endangered pine forest ecosystems.  Data files include the species identity, size (dbh = diameter at breast height), and location of all trees ≥5 cm dbh in 54 individual tree plots. Additional data are provided about heights of palm trees, and the location and burn history of each plot. Users are encouraged to cite this dataset.

The Raccoon Point area in eastern Big Cypress National Preserve is the largest remaining area of mature south Florida slash pine (P. elliottii var. densa) forest, having escaped the logging activities of 1900-1960 (Patterson & Robertson, 1981). The pinelands consist of a mosaic of small, slightly elevated and dryer "islands" isolated among wetland swamp areas of dwarf cypress prairies and cypress domes. The isolation and inaccessibility of the Raccoon Point pine islands provided de facto protection against logging. However, the federal government did not acquire the mineral rights in Big Cypress when the preserve was created in 1974 and in 1977 Exxon Company, USA, constructed an all-weather private access road running north from U.S. Highway 41 to multiple active oil well pads in the Raccoon Point oil field in BCNP.

The study area surrounds the Raccoon Point oil field. It is divided into 18 experimental burn units. Each burn unit includes at least 50 ha of pine forest. Within each unit, three permanent 1-ha tree plots were established and repeatedly censused (Snyder & Belles, 2000) starting in the early- to mid-1990s.

The experimental burn treatments consisted of burning at three seasons (spring, or early wet season [May-June], when the largest human-caused or lightning-caused wildfires occur; summer, or mid wet season [July-August], when there are frequent, but generally smaller, lightning-ignited fires; and winter, or mid dry season [December 15- February 15], when conditions are frequently favorable for prescribed burning) and two burn frequencies (every 3 years and every 6 years) for a total of six treatment combinations. Each treatment was replicated three times, with one replicate originally scheduled to burn per year for three years. Actual burn treatments did not always follow the prescribed schedule due to abnormally wet conditions, state-wide burning bans brought on by drought conditions, or the occurrence of natural unplanned fires. All pinelands in the Raccoon Point study area were burned twice by the National Park Service as initializing burns before the start of this study (January 3 - February 9, 1990, and February 27 - April 4, 1994). A table of burn dates is included in the data files.

The permanent 1-hectare tree plots were all established between 1993 and 1995 (Snyder & Belles, 2000). Within each burn unit, the tree plot locations were chosen by overlaying a grid on enlarged aerial photographs and randomly choosing points that fell within areas that appeared to be pineland. If it was subsequently determined during visits to the field that the area was not large enough to contain the 1-ha plot, or if more than 10% of the area was dominated by cypress trees, another point was chosen. The final geographic coordinates of all 54 tree plots are included in the data files (coordinates indicate approximate plot centers).

Plots were oriented in a north-south direction so that the permanent markers would be easier to locate in the future. To set up each 1.0 ha tree plot, an east-west 100 m line was first established by driving 60 or 90 cm pieces of 1.3 cm diameter steel reinforcement bar (i.e., rebar) into the ground at 0, 25, 75, and 100 m. Two north-south 100 m lines were then established from the 25 and 75 m rebars using a double right-angle prism to place 2 additional rebar stakes 50 m apart. The two remaining plot corners were put into place last. A total of 10 rebar stakes were placed in each plot. In most cases a rotary hammer demolition drill was needed to secure the rebar in the very hard Tamiami limestone that is close to the surface throughout BCNP. Round aluminum tags with unit and plot numbers were wired to the four corner stakes of each tree plot. Each tree plot was then divided into quarters to make tree mapping more efficient. The quarters were identified by compass orientation: NW, NE, SW, and SE. Each quarter was further subdivided into halves. A 50 m tape was laid out along the center line that separated these two halves. A right-angle prism was used to locate the position of each tree along the center line and a second tape was used to measure the distance from the center line to the tree. The field distances were subsequently adjusted to a single plot-level XY cartesian coordinate system (Snyder & Belles, 2000).

Within each plot, all tree stems with a diameter at breast height (dbh) ≥ 5.0 cm and palms with a height to the apical bud ≥ 1.4 m were tagged with a round, 3.2 cm diameter, pre-numbered aluminum tag using a 5.4 cm aluminum nail. The tags were placed on live trees at breast height (approximately 1.4 m) and the diameter was measured just above the nail. Tags were placed in the most secure location on palm trees (i.e., a smooth area of the stem without remnant leaf bases). In addition to live trees, dead pine snags were also tagged and measured (Snyder & Belles, 2000).

Data collected on individuals included tag number, XY cartesian coordinates within the plot, dbh, and species identity. For palms, the height to the apical bud was recorded rather than dbh since palm diameters do not increase with age. The heights of palms and some trees were measured using a telescoping pole up to a maximum of ~6 meters; taller trees were not measured (with the exception of some uprooted, but still living, trees that could be measured with measuring tapes).

Tree plots were recensused between 3 and 10 times each, with the last censuses completed in 2012 (the last censuses for most plots were prior to 2003). During the plot recensuses, it was determined whether trees were living or had died since the previous census, and the dbhs of all living individuals were remeasured (along with dead pine snags). Additional notes were taken on features such as basal fire scars on pine trees. Tags were replaced as necessary. In addition, all new recruits with dbh ≥ 5.0 cm were measured and tagged. Attempts were made to census plots immediately before and after prescribed burns.

In 2024, the original tree census datasets were cleaned and reformatted to facilitate data sharing and statistical analyses. Data cleaning consisted primarily of correcting typographic and data entry errors (e.g., remove leading and trialing blank spaces, etc.). In addition, we screened the dbh data for outlier values; if any value appeared to be erroneous based on the other measurements of the same individual tree, we adjusted it accordingly (in the vast majority of these cases the incorrect values had been recorded in the wrong unit - i.e., 10× too small or too large - so the required corrections were obvious). Notable to the reformatting, since the dbhs of palms were not measured or recorded in the censuses, they were set to 35 cm and 10 cm for Sabal palmetto and Serenoa repens, which are the approximate mean dbh’s for these species, respectively (Jones 1995). Also, all cypress trees were reassigned to Taxodium ascendens in accord with updated nomenclature. While the censuses did measure dead pine snags, the dbhs for all dead trees were set to “NA” in the reformatted data files to minimize the risk of including dead trees in aboveground biomass or growth estimates. The original dbh values for all stems are included.

REFERENCES

Black, D. W., & Black, S. (1980). Plants of Big Cypress National Preserve: a preliminary checklist of vascular plants: US National Park Service, South Florida Research Center, Everglades National Park.

Duever, M. J. (2005). Big Cypress regional ecosystem conceptual ecological model. Wetlands, 25(4), 843-853.

Gunderson, L. H., & Loope, L. L. (1982). An Inventory of the Plant Communities in the Levee-28 Tieback Area, Big Cypress National Preserve: National Park Service, South Florida Research Center, Everglades National Park.

Jones, D.L. (1995). Palms throughout the World. Smithsonian Institution Press, Washington DC.

Muss, J. D., Austin, D. F., & Snyder, J. R. (2003). Plants of the Big Cypress National Preserve, Florida. Journal of the Torrey Botanical Society, 119-142.

Patterson, G. A., & Robertson, W. B. (1981). Distribution and habitat of the red-cockaded woodpecker in Big Cypress National Preserve: National Park Service, South Florida Research Center, Everglades National Park.

Snyder, J. R., & Belles, H. (2000). Long-term study of fire season and frequency in pine forest and associated cypress wetlands, Big Cypress National Preserve: project description and preliminary data. US Geological Survey, Big Cypress National Preserve Field Station, Ochopee, Florida.

Snyder, J. R. (1991). Fire regimes in subtropical south Florida. Proceedings of the Tall Timbers Fire Ecology Conference 17, 303-319.