Individual study: The effect of controlled burns on fruiting of saw palmetto Serenoa repens in Collier, Sarasota, Highlands and Polk Counties, Florida, USA
Carrington M.E. & Mullahey J.J. (2006) Effects of burning season and frequency on saw palmetto (Serenoa repens) flowering and fruiting. Forest Ecology and Management, 230, 69-78
Medicinal use of saw palmetto Serenoa repens (endemic to the southeastern coastal plain of the United States) fruits has led to a recent sharp increase in fruit harvesting. This palmetto is also often considered a keystone species, serving as habitat or a food source for many associated animals. Due to harvesting pressure on this non-cultivated species, practices that produce sustainable harvests are needed. As part of a research project investigating controlled burning as a management tool, a study of the effects of burning season and frequency on saw palmetto flowering and fruiting in pine flatwoods in Florida in 1997 to 1999, was undertaken. Fire is the predominant natural disturbance in most ecosystems in which saw palmettos occur, but fire is also used as a management tool in these areas to maintain land for grazing. Saw palmetto is a fire-adapted species, with strong post-fire resprouting and flowering responses.
The objective of this study was to characterize saw palmetto flowering and fruiting responses to different burning seasons and past burning frequencies, and to facilitate development of controlled burning as a management practice.
Study sites: The study was undertaken in two stages, the first experiment conducted in 16 pine flatwoods sites during 1997–1998, and the second using 16 different pine flatwoods sites in 1998–1999. The sites were located in Collier, Sarasota, Highlands and Polk Counties in southwest and central Florida. The study flatwoods communities were dominated by longleaf pine Pinus palustris and/or slash pines Pinus elliottii. Pine densities ranged from 0 trees/ha to 227 trees/ha. Average density of saw palmetto ramets in the understory was 14,000 (± 1,100) ramets/ha.
The climate in the area is humid and mild, with an average annual temperature of 21–23 °C, annual rainfall is 130–150 cm, with most (approximately 60%) occurring from June to September.
Burning season & burn frequency: Effects of burning season ('growing season' or 'winter') and burn frequency ('frequent' or 'infrequent') were investigated. Growing season burns had occurred in April–July 1996, and winter burns during November 1996–February 1997. Frequent burn sites had been burned every 2–3 years for the past 15 years; infrequent burn sites had been burned less often than every 5 years for the past 15 years. Flowering and fruiting levels in all sites were estimated in 1997 and 1998. The experiment was repeated in 1998 1999 at16 new sites in Florida Panther Wildlife Refuge, Picayune Strand State Forest, Southwest Florida Research and Education Center, Myakka River State Park, Lake Wales Ridge State Forest and Avon Park Air Force Range.
Palmetto measurements: During April (the peak saw palmetto flowering month in the study area) flowering level estimates were made by quantifying proportions of saw palmettos flowering in 10, 5-m line transects at each site (positioned in a stratified random manner, each 5–30 m apart). In the first year following a burn, any flowering prior to April was ascertained by looking for old, unburned flowering spadices on palmettos in the transects.
During August and September (peak fruit harvesting months) estimates of proportions of palmettos fruiting (using the transects and methodology), fruit yield per ramet (by picking and weighing all fruits from each fruiting ramet) and ramet density per site were ade. To estimate ramet density per site, saw palmetto ramets were counted in 10, 2 m × 5 m quadrats in each site. The quadrats were positioned in a stratified random manner. Locations of transects and quadrats were re-randomised for the second year of data collection in each site.
Data analysis: For each year in each site, proportions of saw palmettos flowering and fruiting were calculated by averaging proportions flowering or fruiting over the 10 transects. To calculate average fruit yield per ramet (kg fruit/ramet) for each site, fruit yields over all fruiting ramets encountered in the transects was averaged. A one-time average ramet density (number of ramets/ha) for each site was also calculated by averaging the number of ramets counted in 10 quadrats, and multiplying by 1000 (to obtain a 'per ha' estimate). Average fruit yield for each site (kg fruit/ha) was also calculated.
Flowering patterns: For both experiments, growing season (April–July) burns had high flowering levels during the first year after burning (average proportion flowering = 0.45 in the first experiment and 0.50 in the second, n = 8 sites in each experiment). This decreased dramatically in the second year after burning (average proportion flowering = 0.09 in the first experiment and 0.20 in the second, n = 16 sites). In contrast, palmettos in winter (November-February) burns flowered at intermediate levels during both the first and second years after burning (average proportion flowering = 0.21 and 0.28 for first and second years after burning, n = 16 sites). Palmettos burned in different seasons also had differences in speed of initial onset of flowering after burning. First flowering after winter fires occurred after 2–5 months, but did not occur until 9–12 months after growing season burns.
Observations of cabbage palm caterpillar Litoprosopus futilis (which consume palmetto buds) from 1997 to 1999 indicated occasional heavy mortality of saw palmetto inflorescences in local areas.
Fruiting patterns: Fruiting failure occurred in 1997 (only 2 of 16 sites had any fruit at all) due to fruit infection by the fungus Colletotrichum gloeosporioides. This outbreak was induced by prolonged periods of heavy rain in April and May 1997, during flowering and early fruit formation. This fruit failure masked any differences in fruiting due to other treatment factors in 1997. Of the two sites that had fruiting in 1997, the highest fruit yield was 11 kg/ha. Subsequently, most sites had fruiting in 1998 and 1999.
In 1998, fruit yield averaged around 200 kg/ha across all sites established in 1997.
For sites established in 1998, growing season burns had higher proportions of palmetto fruiting than winter burns in both 1998 and 1999. Fruit yields were higher in growing season burns than in winter burns only in infrequently burned sites, and only in 1998, the first year post-burn (average yields = 325 kg/ha and 0.25 kg/ha for growing season, infrequently burned sites and winter, infrequently burned sites; n = 4 sites in each group).
In 1999, the second year post-burn, fruit yields in infrequently burned sites were slightly higher in winter burns (289 kg/ha and 195 kg/ha for winter, infrequently burned sites and growing season, infrequently burned sites; n = 4 sites in each group). Frequently burned sites generally had low fruit yields ((163 kg/ha and 242 kg/ha for frequently burned sites in 1998 and 1999; n = 8 sites).
Conclusions: This study demonstrates that flowering and fruiting in saw palmetto can be influenced by controlled burning. Infrequent growing season burns were associated with higher fruit yields in saw palmettos. An optimal burning frequency for saw palmetto flowering is probably every 5–8 years, but this burn frequency might not always be compatible with other land uses (e.g. cattle grazing) and effects on other fauna and flora must be taken into consideration.
Note: If using or referring to this published study, please read and quote the original paper. Please do not quote as a conservationevidence.com case as this is for previously unpublished work only.