• Department Home
• Forestry Home
• About Us: Forestry
• Forest Management
• Wildland Fire
•   • Fire Prevention
•   • Prescribed Fire
•   • Safety
•   • Tools & Downloads
• Services We Provide
• County Foresters
• Field Operations
• State Forest Recreation
• Fire Weather
• Training and Education
• Calendar of Events
• Publications
• Employment: Forestry
• Forestry Site Map
• Contact Us
• Search
• Select a Division

Ecological and Economic Consequences of the
1998 Florida Wildfires

Project Process Overview and Insight with Executive Summary (pdf, 204k)

I.    Economic Effects of Catastrophic Wildfires
II.   Haines and Lavadas Dispersion Index Relationship
III.  Geography and Florida Scrub-Jay Populations
IV.  Reducing Hazardous Fuel Conditions in Pine Flatwoods
V.   Wildfire Impacts on Pine Bark Beetles
VI.  Behave Model Performance
VII. Fuel's Management Reduces Tree Mortality Following Wildfire
VIII. Short-term Response of Plant Species Special Concern and Exotics to The 1998 Florida Wildfires
IX.   Wildfire Home Protection Strategies

The 1998 wildfires had unusually large economic effects, conservatively estimated at $620 million, with over half of the damages through timber market losses and distortions. A static analysis of 1998 wildfire risk based on fixed permanent inventory plots in this area identified as high risk: coniferous forests such as baldcypress, forests with substantial understory vegetation, fragmented forests, and forests in or near hydric sites and wetland forests. Stands at low risk of wildfire were on xeric sites or had experienced a wildfire during the previous survey interval, 4-13 years prior to 1998. Historical analysis of wildfires over nearly two decades and across the state found that wildfires in Florida can be modeled using the same log-log linear model used in studies of wildfire size distributions in other regions with or without fire suppression programs.

In general, large fires were responsible for most of the area burned although the extent that this was true varied by ecoregion and year. For example, the 1998 fires in northeastern Florida showed a much flatter distribution than observed in other serious fire years, indicating large fires played an unusually large role that year. Analyses of annual wildfire area by county over this long period and across the state found that wildfire area was negatively related to county housing density and negatively related to wildfire area in previous years but was not significantly related to prescribed fire permits issued in that county. Coupled with the results of the fixed FIA plot analysis, this finding implies that wildfires have the primary suppressive effect on wildfire risk. Efforts to quantify the role of prescribed burning in reducing wildfire risk were unsuccessful in all of the analyses, although it should be noted that these analyses only considered wildfire area and not wildfire intensity or damage.

This study focuses on evaluating the predictive ability of these two indices with regard to the potential for large wildfires. Using data from the summer of 1998, when Florida experienced one of its most severe wildfire seasons on record, the relationship between both the Haines Index and the Atmospheric Dispersion Index to the occurrence of large wildfires will be examined. The Haines Index performed well in predicting large wildfires during June of 1998. The ADI also performed well although the number of false alarms produced was much greater than for the Haines Index. Breaking out only the stability component of the ADI (i.e. neglecting transport winds) produced an index that performed comparably to the Haines Index.

Examination of the normalized ratio of the ADI stability component to the transport wind provided a further means of evaluating the atmosphere's ability to contribute to the rapid growth of wildfires. This ratio predicted all but two days of large wildfire growth and produced fewer false alarms than the other indices for 1998. For 1999, both the Haines index and the ADI performed poorly as indicators of large wildfire growth. The ratio of the stability component of the ADI to the transport wind again yielded better results. This ratio also indicated cases where the Haines index would work well (stability dominates) and when the ADI is most useful.

All but one of the study sites were in public ownership or were proposed for conservation acquisition. South Brevard has the most scrub sites in public ownership that have an abundance of Florida scrub-jays. More scrub conservation land acquisition has been directed towards South Brevard on the Florida mainland where large scrub-jay populations occur. The reduction in fire frequency is most severe within North Brevard so that the lower abundance of jays in North Brevard is the result of habitat degradation and not necessarily habitat potential.

Extensive surveys were conducted in 1993 to support the Statewide Surveys and the Brevard County Scrub Conservation and Development Plan which was not implemented. The study sites had 150 pairs during 1993 and declined to 64 pairs in 1999. This represented a population growth rate (R) of 0.865 or an average population decline of 13.5% per year. The rate of decline threatens the recovery potential because the population growth rate under optimal conditions is not likely to result in rapid population recovery.

Population declines would have been steeper if there had not been significant immigration into the study sites. Immigration into study sites was not unexpected because the sites included the densest aggregations of remaining Florida scrub-jays in the region. The study sites were locations where dispersing Florida scrub-jays had greater chances to breed than surrounding areas, partially because the study sites were of higher quality than their surroundings because of their habitat size, composition, and contiguity.

The above rate of population decline was slightly influenced by a presumed statewide epidemic (1997-1998) that resulted in a greater than normal mortality rate. Florida scrub-jay breeder survival is high during most years except during short, rare episodes of high mortality. Data from all years clearly indicated that the most important cause for decline was poor habitat quality that resulted from the disruption of natural fire regimes. Lower reproductive success and survival was comparable to other studies that have shown that Florida Scrub-Jays can not persist in habitats that are subject to fire regimes that are infrequent. All studies indicate that Atlantic coast populations need a more frequent fire regime than previously believed because vegetation grows faster on the coast. The effects of individual wildfires vary greatly depending on the characteristics of individual fires and the recent history of fire and population patterns at a sites where fires occur.

Recovery team efforts indicated that the metapopulation in South Brevard was the fourth largest metapopulation in the species range and would be expected to be the fourth largest if all habitat that is proposed for conservation is restored to optimal conditions. Although some of the habitat once proposed for conservation is likely to be destroyed, more habitat has been found than originally estimated. Ongoing studies indicate that metapopulation size was underestimated by 15-30 pairs in 1993 because private ranchland that is now under public ownership or soon to be acquired was not inventoried.

Organizations involved in management have proceeded from a primary acquisition mode to one that includes fire management. Much progress was made in the past year towards restoration and management. Previous wildfires in north and south Brevard have also improved habitat. However, nearly all territories have patches of tall scrub and these conditions have long-term mortality rates that exceed reproductive success rates, at least on Merritt Island. No studies of colorbanded Florida scrub- jays have yet demonstrated that Florida scrub-jays can persist in areas where there always is an abundance of tall scrub.

Florida scrub-jays had not recolonized some restored areas which were below their carrying capacity. The habitat quality of most occupied habitat remained poor so that there was not a large production of new potential breeders. Exchanges among subpopulations were infrequent for females and extremely rare for males so that local population dynamics were more relevant than metapopulation dynamics. Forests hindered dispersal and separated many restored areas from occupied territories that produced potential breeders. A gradual territorial budding process was hypothesized as one mechanism for reestablishing Florida scrub-jays in restored areas where jays have been extirpated. This process is predicted to be slow and relies on contiguity between restored areas and occupied areas. Florida scrub- jays have been demonstrated to initiate new territories by processes other than budding, but these are also uncommon. Therefore, forest barriers between occupied and restored habitat should be expeditiously eliminated where the forests are artifacts of human activities.

Expediting most recovery actions would improve the chances of successful recovery and decrease extinction probabilities. The reason for this is that managing to reduce the level of Florida scrub-jay decline is likely to have greater success than trying to increase populations that have become very few in number. Much restoration is needed for recovery in most landscapes. Restoration includes mechanical cutting of scrub that remained unburned for long periods, frequent fire, and tree thinning. Data from this study was routinely applied by organizations. Restoration progress increased in the past 12 months as management organizations acquired fire management staff.

Adaptive management must be coupled with scrub-jay demography data given variations in fire behavior and vegetation responses to management. Habitat mapping applications are generally to coarse to provide the information needed for management and predicting population responses. Many questions remain concerning the demographic success associated with the variation of habitat conditions that approach but do not achieve optimal habitat suitability. It is unlikely that every territory will be restored to optimal conditions so that reproductive success will exceed mortality in some areas so that it will be important to determine the distributions of optimal habitat necessary for recovery. These questions are the focus of future studies as the study expands into newly acquired areas and continues in areas that are now being subject to habitat management actions.

Three sites were selected, each representing one of the treatments, and at each site 15 stands covering a range of rough ages, time since the last silvicultural treatment, were chosen. The stands were inventoried for height, fuel load, and percent cover of the rough and these data were used to develop a custom fuel model for each treatment / rough age combination. These models were then used in the BEHAVE fire prediction system to estimate fire behavior and probability of overstory pine mortality under drought and normal weather conditions for each treatment / rough age combination. In all untreated stands (rough age > 10 years), BEHAVE predicted extreme fire behavior, making any fire control operation very difficult and probably ineffective, and 100% mortality of overstory pines.

Prescribed burning provided immediate protection for the pines by reducing fire behavior under both weather conditions. However, this protection quickly disappeared as the rough aged and the shrubs recovered following the prescribed fire. Partial harvest impacted fire behavior in much the same manner and for a similar length of time as prescribed fire. Understory herbicides showed little immediate reduction in fire danger. However, beginning in second year after treatment, fire behavior decreased, a condition that lasted at least until year 6, making survival of overstory pines and successful fire suppression more likely. Under some circumstances, partial harvests and understory herbicides can substitute for fire to reduce fire behavior in pine flatwood forests but forest managers must be aware of each alternative's strengths and weaknesses. Combining treatments may provide immediate and long- term fire protection but these combinations need more research.

In contrast, the wildfires of 1998 occurred in June and July during an extended drought when Keetch-Byram drought indices were above 700 and temperatures were in the 90's most days. At the conclusion of the fires it was clear that they killed large numbers of trees. What was uncertain was how the remaining trees that still had some live crown would fare in subsequent years, and how the bark beetle and reproduction weevils that inhabit Southern pine forests would respond to this mosaic of dead, dying and damaged trees.

The pine bark beetle complex (Coleoptera: Scolytidae) in the South includes five common species that at times can cause tree mortality: the Southern pine beetle (SPB), Dendroctonus frontalis (Zimmermann); the pine engravers, Ips grandicollis (Eichhoff), I. calligraphus (Germar), and I. avulsus (Eichhoff); and the black turpentine beetle (BTB), D. terebrans (Olivier). Of these, SPB is the most aggressive, routinely killing relatively healthy trees over large areas across the South during outbreaks (Price et. al. 1991). Ips beetles are less aggressive than SPB with tree-killing attacks generally restricted to stressed or damaged trees. Ips populations often build up in logging slash, windthrown, drought-stressed or lightning struck trees. From these population foci, subsequent generations can emerge to attack apparently healthy trees nearby (Drooz 1985). Thatcher (1960) estimated that pine engravers were responsible for 3.7 million m3 of timber loss annually in the South while Baker (1972) reported 1.1 million m3 was killed annually in Florida alone. BTB is attracted to stumps and injured trees.

Trees weakened by fire, old age, adverse weather, or damaged by storms and naval stores or harvesting operations are frequently attacked. Although BTB is primarily considered a secondary pest attacking weakened or damaged trees, outbreaks resulting in extensive tree mortality have occurred in all states along the Gulf coast (Drooz 1985).Pine reproduction weevils, Hylobius pales (Herbst) and Pachylobius picivorus (Germar) (Coleoptera: Curculionidae), differ from bark beetles in that they breed in the stumps and roots of recently cut, killed or severely damaged trees. Adults that emerge from the dead material feed on the phloem tissue of seedlings or small branches of larger trees. When feeding is concentrated on seedlings these weevils often cause severe reductions in pine reproduction in infested areas.

Reproduction weevils are capable of flying over 2 miles to reach breeding sites and adults can live for nine months or longer in the field (Bullard and Fox 1969). Attraction to and subsequent increases in weevil populations in the burned areas might affect pine regeneration in and around such areas, and these weevils have been implicated in the transmission of Leptographium procerum (Kendrick) Wingfield, a root disease reportedly capable of killing larger, mature pine trees (Klepzig et. al. 1991, Nevill and Alexander 1992a). The above insects may affect the long-term health of the forests in the vicinity of the burns.

One possibility is the number of dead trees is so large or they were so altered by the fire that these insect species will not be able to take full advantage of a seemingly abundant resource. Conversely, the large numbers of dead and damaged trees may result in a build up of bark beetles that could then kill weakened trees that might otherwise have survived the fires. In addition, the high populations could spill over into nearby unburned areas killing relatively healthy trees. The availability of severely weakened trees could allow engraver beetles and BTB to maintain high population levels over several years. Increases in regeneration weevil populations could substantially affect or delay natural or artificial forest regeneration.

The combination of these events may start a prolonged cycle of forest decline, especially through the transmission of L. procerum.We examined tree mortality, and relative abundance of bark beetles and wood borers along a gradient of fire intensity from no fire to high intensity using a variety of standard trapping and sampling techniques. Sampling devices included Lindgren funnel traps to determine the relative abundance of pine foraging species and their associated predators, and unbaited pitfall traps and crawl traps to sample reproduction weevil populations. In addition, we monitored tree mortality within study stands over a one year period and sampled roots of live trees to determine the prevalence of Leptographium species one year after the wildfire.

VIII. Short-term Response of Plant Species of Special Concern and Exotics to The 1998 Florida Wildfires
Susan L. Grace
The wildfires of 1998 were devastating to the state of Florida with over 500,000 acres burned and a cost of $390 million. While the wildfires were destructive economically, ecologically some plant species of concern may benefit from fire. Surveys of plant species were conducted on public lands extensively burned in the 1998 wildfires to investigate the response of species of concern and exotics following the wildfires. A list of species of concern was compiled and selected sites were surveyed intensively in 100m2 areas along 0.1 mile points within burned sites for plant species. Inventories were conducted during the fall immediately after the wildfires and the following spring and summer of 1999.

For most sites studied little to no previous plant surveys existed and new plant species lists were developed as a result of this study. Several new locations for the fall flowering ixia, Florida toothache grass, Florida threeawn grass, and Rugel's pawpaw were discovered during the fall survey. Additional locations of Florida beargrass, wild cocoa and rain lily were documented from the summer surveys. Logging activities were prominent following the wildfires and have long term impacts on invasive species that were not detected in the scope of this project. No significant establishments of invasive species were observed in this study, however, additional follow-up monitoring of invasive species is necessary in wildfire areas. Stable, reproducing populations of invasive species occur along right of ways that are on or adjacent to public land areas surveyed and can act as seed sources for future establishment of heavily logged or intense soil disturbed sites. A portion of the support for this project was also used to fund an independent study of the impact of the wildfires on the federally endangered Rugel's pawpaw, part of an ongoing study to document the demography of the species. Population size was estimated at about 200 individuals before the wildfires and following the wildfires over 2000 individuals and an increase in flowering was observed. This overall project of the impact of wildfires on plant species of concern has demonstrated that a suite of plants considered to be fire-adapted species has increased in both flowering and numbers of populations following the wildfires of 1998.