Everglades Fire Regime

Under the natural fire regime, fire in the Everglades is a fixture in the landscape. Early accounts of fire in the Everglades revealed a casual attitude towards fire (Robertson 1953). Under the historical fire regime, large uninhabited expanses and the low severity of fires, resulting in little change to most plant communities, likely interacted to produce this attitude (Robertson 1953). Few records or accounts of fire in the Everglades are available; however, several lines of evidence suggest the long-term role of recurring fire in the Everglades. Today, lightning is a source of ignition of many fires in the Everglades and is generally considered to have been so in the past. Many vegetation types succeed to another vegetation type in the absence of fire (Robertson 1953, Snyder 1991). Lenses of charcoal and ash have been found throughout the Everglades in certain types of peat layers (Cohen 1974).

Historically, lightning fires were thought to have been distinctly seasonal (Gunderson and Snyder 1994, Wade et al. 1980, Snyder 1991). Though numerous lightning ignited fires occurred from May to September, most area burned in May and June at the end of the dry season, beginning of the wet season (Gunderson and Snyder 1994, Snyder 1991). The extent to which Native Americans employed burning is unclear. The presence of archaeological sites on the highest elevation sites shows that Native Americans were present in the Everglades. It is thought that Native Americans used fire to modify their environment, but probably avoided fires that caused peat consumption (Robertson 1953 and references therein). (Note that at least some of the highest elevation sites utilized by Native Americans were peat soil based). Early conjecture stated that fires were set by Native Americans as early in the dry season as fire would carry (Robertson 1953), while later authors thought that fires were set all year with the largest fires occurring in April, May and June as they are today (Snyder 1991).

The frequency with which natural fires burned in any one community is less well understood. Typical methods for determining fire return intervals, such as studies of tree rings and fire scars, have not been successful in the Everglades (Taylor 1980 in Snyder 1991). Observations on plant recovery, fuel accumulation rates and present day observed fire return intervals have lead to speculation on relative fire return frequencies among communities.

· Sawgrass marsh community is thought to have burned more frequently than most other freshwater wetland communities. The only community that burned more frequently was the pine rockland community (See Slash pine). The historical fire return intervals for sawgrass marsh has been estimated as 3-25 years (Hendrickson 1972 in Wade et al. 1980). Other estimates range from1-5 years (FNAI 1990) and Wade et al. (1980) state that most sawgrass marshes can burn as frequently as every 3-5 years based on biomass accumulation. Records indicate that one sawgrass site burned three times in 4 months (Wade et al. 1980). There is considerable variation in the structure and density of sawgrass marshes, which may explain the range of fire return intervals.

· Long hydroperiod marshes were historically wet enough that they rarely burned and often functioned as fire breaks especially for wet season fires (Wade et al. 1980, Gunderson and Loftus 1993). Maidencane dominated long hydroperiod marshes are the exception. Maidencane will burn when flooded and burns almost as well as sawgrass (see freshwater marshes) (Wade et al. 1980). Under drought conditions, long hydroperiod marshes did burn with severe fires consuming organic soil (Wade et al. 1980).

· Marl prairies, which were drier than preceding communities, tend to be more sparsely vegetated than sawgrass marsh. Sparse vegetation is thought to have resulted in patchy and less frequent burns. The exception to this pattern is for muhly grass dominated marl prairies, which have thicker vegetation, and can carry fires more frequently and result in more complete burns (Wade et al. 1980, Gunderson and Loftus 1993). Frequency of fires in muhly grass prairies may be limited by biomass accumulation: Biomass in muhly grass prairies reaches a steady state in 3 years (Herndon and Taylor 1986). Vegetative changes in marl prairies, an increase in abundance of muhly grass, are thought to have resulted from changes in hydrologic and fire regimes (Atwater 1954 in Gunderson).

· Hardwood hammocks and tree islands burn least frequently of Everglades plant communities. Fires in tree islands tend to occur only in dry periods. Though exact fire return intervals are not known, most hardwood hammocks on Long Pine Key had burned in the last 40 years as of 1983 (Olmstead et al.1983). This frequency of fire may be artificially short due to human alteration to the natural fire regime through modifications to hydrology (Taylor 1981). Examination of fire records in the Big Cypress swamp (including many human ignitions) indicates that hammocks there burn with a frequency of once in 170 plus years (Snyder 1991).

Plant community changes depend on severity of fires. Severe fires that coincided with periodic widespread droughts, occurring on the order of decade intervals (Gunderson and Snyder 1994), affected plant communities more than frequent fires discussed above. Severe fires that resulted in organic soil consumption and thus soil elevation changes result in rapid changes in plant communities by affecting hydrology (Gunderson and Loftus 1993). Sawgrass marsh can become ponds, long hydroperiod marsh or willow heads. Long hydroperiod marsh can become pond communities. Tree islands can become willow heads, remnant rings of saw palmetto or barren limestone outcrops (Wade et al. 1980). Soil accumulation, a much slower process that occurs in the absence of severe fires, is thought to direct succession in the other direction (Gunderson and Loftus 1993).

The extent of natural wet season fires was thought to be relatively small (Taylor 1981) based on current day fire behavior (Wade et al. 1980, Klukas 1973) and the discontinuity of charcoal deposits in peat (Cohen 1974). However, periods of drought spaced at relatively infrequent intervals (once a decade) were associated with large fires (Gunderson and Snyder 1994). Current day climatological data indicate that because rainfall is dependent on convective storms the nature of rainfall can be patchy, which results in localized droughts. Localized moisture patterns could have affected fire sizes (Duever et al. 1994). The idea that fire size is related to ambient moisture conditions was not born out by early examination of patterns of moisture and fire size or fire number. Neither number of fires or area burned was well correlated with either water levels or precipitation (Taylor 1981). Taylor (1981) suggested that fire size was likely more related to short-term weather patterns such as wind and relative humidity and was related to the location of firebreaks in relation to fire ignition points. Conversely, recent analysis has shown that in the years when the most extensive non-prescription fires occurred in Everglades National Park were also characterized by low water levels in April and May one year proceding fires (Beckage and Platt 2003).

Alterations to the Historical Fire Regime

Several clear changes have occurred to the Everglades fire regime. Season of fire has shifted, with more fires and more acreage burning in the dry season. Where hydrologic regime has been altered, frequency of severe fires has changed.

The season when fires occur in the Everglades has been altered by humans. Prior to human habitation, lightning-caused fires would have burned in the Everglades throughout the wet season with most acres burning in May and June at the onset of the wet season. Humans ignite non-prescription fires accidentally or on purpose throughout the year, but the greatest number of fires are in the dry season. Human caused fires are shifted earlier in the year than lightning caused fires with most fires and most acreage burned in April and May. As of the early 90’s the area burned by human caused fires was almost twice that of lightning fires (Gunderson and Snyder 1994). Human caused fires are often viewed as management problems because they tend to occur under drier conditions than lightning fires and tend to become larger, and cause muck fires (Taylor 1981).

The change in modern human caused fire patterns is unknown, because we cannot know the season and extent to which aboriginals used fire. Several authors have speculated that non-aboriginal people burned for some of the same reasons as Native Americans and thus timing could have been similar. People burn to kill mosquitoes, kill snakes, clear brush, drive out game, create fresh pasture to attract game and to find gator holes (hide hunters) (Robertson 1953).

The second change to the fire regime has been indirectly caused by human changes to the hydrologic regime. In some areas water has been impounded, thus fire frequencies are decreased (Gunderson and Loftus 1993). But the larger management problem is where marshes have been drained either by stopping sheet flow or by increasing runoff in canal systems. Under these drainage scenarios, severe ground fires have become more common resulting in changes to plant communities (Gunderson and Loftus 1993, Robertson 1953, Snyder 1991). Wildfires under dry conditions in the 1990’s have caused at least top-kill of all trees on many thousands of tree islands within the water conservation areas (Hoffman et al. 1994). Statements by several authors and policies by managing agencies dating as far back as the 1950’s reflect concern over the perceived increase in frequency with which tree islands suffered severe fire due to changes in hydrology and fire ignition patterns (Robertson 1953, Taylor 1981). Marshes in urban interface areas often have shortened hydroperiods as waters are diverted for flood control. These same areas are often subject to increased fire ignition sources as accident or arson. Thus these areas often experience increased fire frequencies (Lockwood et al. 2003).

In other situations relationships are more complex. In some of the water conservation areas, long hydroperiod marsh communities are being invaded by monotypic stands of sawgrass. This change would indicate a lowering of water levels, however data do not support this hypothesis. One theory is that prescribed fire programs aimed at preventing muck fires have been so successful that the shifting mosaic nature of these communities has been disrupted. It has been postulated that severe fires have not been frequent enough to create new slough habitat by lowering soil elevations so the proportion of sawgrass increases (Davis et al. 1994).

The spread and proliferation of invasive exotic plants has also altered natural fire regimes. Melaleuca-dominated forests can carry crown fires, a previously non-existent fire regime in South Florida. Old world climbing fern acts as a ladder fuel carrying marsh surface fires into tree canopies of tree islands. Other exotic plants decrease the ability of plant communities to carry fire (See Melaleuca and Exotic plants for more information).