Previous PageTable Of ContentsNext Page

5 Impacts and Management Recommendations

Management information is based on literature review, contacts with experts, and guidelines provided in installation reports. Information on habitats and manage ment for rare plant and animal species was gathered from USFWS Recovery Plans, TNC Element Stewardship Abstracts, Jordan et al. (1995), Godfrey and Wooten (1979, 1981), Kral (1983), Small (1972), and Ward (1979).

Fragmentation and Land-Use Conversion

Impacts

Across Florida, many scrub communities have been fragmented and converted for residential developments or citrus production (Christman and Judd 1990). Natural communities on DoD installations also are fragmented by systems of roads, firebreaks, sometimes by drop zones, and facilities construction. The general effects of fragmentation on TES populations include outright habitat loss, population isolation, changes in plant and animal assemblages through altered competition, and changes in predation, parasitism, and herbivory patterns (reviewed in general by Trame and Tazik 1995).

Fragmentation can also result in localized extinctions of flora and fauna more often than in contiguous habitat, if species' dispersal rates or mechanisms do not allow them to emigrate between habitat patches. The effects of patch size, patch isolation (distance to nearest neighbor patch), and presence of connecting xeric habitat ("corridors") on species richness for several taxonomic groups was studied for peninsular, inland sand pine scrub communities (McCoy and Mushinsky 1994). The size of a scrub fragment (ranging from 1.5 ha to 190 ha) was positively correlated with adjacency to certain types of upland habitats and more types of habitat in the adjacent landscape, and negatively correlated with adjacency to disturbed or converted land and decreases in patch size over the past 50 years. Any effects of patch size could have been influenced by these factors. Larger patches of scrub were strongly correlated with high species richness of nonavian taxa, "characteristic" (typical of scrub communities) nonavian taxa, breeding birds, and characteristicavian taxa. The presence of rosemary was also more likely in larger scrubs (McCoy and Mushinsky 1994). Another study found the occurrence of 16 Florida scrub plant species in inland rosemary scrub to be correlated with a combination of or the interaction between increasing patch size (range: 0.03 to 3.6 ha) and decreasing patch isolation (Quintana-Ascencio and Menges 1996).

Nonetheless, there is evidence that even small patches of inland scrub can support characteristic plant and vertebrate diversity. Qunitana-Ascencio and Menges (1996) found that 31 out of 62 species were no less likely to be found in smaller patches or more isolated patches, even when the interaction with fire history was considered. McCoy and Mushinsky's (1994) vertebrate study found that the three largest study patches, each close to 200 ha in size, did not contain more species than combined groups of small patches of equal total area. Conservation of a few large scrub remnants may not be more valuable than maintaining many relatively small patches on the landscape, at least for the measures studied to date. In fact, scrubs of less than 50 ha supported relatively more species than expected, while scrubs of 150 to 200 ha supported fewer species than expected. This trend was consistent for threatened and endangered fauna as well as rare amphibian and reptiles (McCoy and Mushinsky 1994).

The requirements of specific listed species are needed in order to manage with an emphasis on conserving TES. It is known that the state threatened short-tailed snake (Stilosoma extenuatum Brown) is able to survive in fragments as small as an acre, as long as the habitat is not too degraded (Campbell 1992). This snake appears to be able to co-exist with human activities as long as the invertebrates and small vertebrates upon which it preys persist. This type of information is needed for many more listed plants and animals.

Highly fragmented communities are a threat to species that rely on frequent colonization from source populations (e.g., Bachman's sparrows; Dunning 1993). Florida scrub lizards were shown by Hokit, Stith, and Branch (In Press) to be absent from 89 of the 132 patches of scrub they surveyed on Avon Park AFR and the nearby Arbuckle State Forest from 1994 through 1996. Through the use of multivariate habitat modeling techniques, they determined that patch size and patch isolation were the only two variables that explained the scrub lizards' distribution. These reptiles exhibit poor dispersal capabilities and do not move well through dense habitat. Because of the relatively high degree of patch isolation, they are unable to colonize suitable habitat.

Isolated habitat patches can be problematic for species with large home ranges, especially if they are slow-moving or conspicuous, or if individuals becomevulnerable during dispersal movements. Such is the case with the gopher tortoise (Lohoefener and Lohmeier 1981). These tortoises live in a naturally discontinuous habitat. Scrub and sandhills are often bisected and surrounded by wet boggy areas that may serve as impediments to adult male courtship movement. This problem has been exacerbated by the encroachment and further fragmentation of the habitat by human activities. Alford (1980) observed that tortoise colonies tend to be of the same age class, which implies that gopher tortoises do not have high rates of inter-colony migration. Diemer (1992) studied home range movements of gopher tortoises in northern Florida and determined the mean home range for adult males was 0.88 ha. For females, the range was 0.31 ha. She noted the longest recorded movement as 0.74 km. The mean home range for gopher tortoises was reported by McRea, Landers, and Garner (1981) as 0.45 ha for males and 0.21 ha for females. These studies looked at movements over a relatively short period of time. Douglass (1976) monitored two male tortoises for 5 to 6 years and reported home ranges in these individuals as 4.2 and 6.3 ha, respectively. Home ranges of this size indicate that tortoises will travel sizable distances for mating purposes. Since many fragments of suitable tortoise habitat are far smaller in area than the potential home range for male tortoises, and because these patches may be considerable distances from one another, with no corridors connecting them, males residing in these small fragments may be restricted to these patches and therefore unable to attain access to new breeding females; i.e., the mature tortoises in these small patches are physically and reproductively isolated from others of breeding age. Such isolation over time can produce genetic distinction and the occurrence of some inbreeding can reasonably be assumed.

In recent years, conservation has emphasized increasing the connectivity of existing habitat patches to combat problems associated with fragmentation. Managers practicing stewardship in a variety of natural communities have sought to use corridors to augment existing tracts of high-quality habitat. The establishment of corridors will unite areas of suitable habitat and allow individuals of a species to emigrate between patches, promoting gene flow between colonies. This strategy may or may not be appropriate for occurrences of Florida scrub; and it may or may not be appropriate for all management objectives. Sites characterized by the harsh environmental conditions that define scrub have historically been disjunct and isolated from other such patches. Those species endemic to scrub may be adapted to isolated habitat. Indeed, the presence or absence of connecting habitats was not correlated with species richness for nonavian taxa, "typical scrub"nonavian taxa, breeding birds, and typical scrub avian species (McCoy and Mushinsky 1994). The value and potential risk of corridors to specific species needs to be evaluated. Researchers at the University of South Florida are currently looking at the spread of upper respiratory tract disease in gopher tortoises in relationship to habitatmanagement practices, including the presence of corridors (Earl McCoy, Professor, Department of Biology, University of South Florida, professional discussion, 29 June 1998). Populations can be decimated by disease spread via corridors when they would otherwise be protected from such epidemics by virtue of their segregation from other isolated populations.

In the absence of suitable corridors, some individuals will certainly move beyond the boundaries of their natural habitat and through developed areas in search of mates or new habitat. This behavior leads to increased mortality from a variety of sources, including human activities and predators.

Humans have been the principal predator of mature tortoises, using them as a food source during the Great Depression (USFWS 1990a). This practice remains a cultural ethos in many rural areas of the United States. Collecting for the pet trade also takes a toll on gopher tortoise populations. However, tortoise eggs and hatchlings suffer the most from habitat fragmentation. For example, in south-western Georgia, Landers, Garner, and McRae (1980) found high predation of gopher tortoise nests by skunks (Mephitis mephitis), raccoons (Procyon lotor), and crows. In Florida and Alabama, tortoises were depredated by feral dogs and cats. Young tortoises have been under predation pressure from gray fox, armadillos (Dasypus novemcintus), various species of snakes, and raptors (USFWS 1990b). Landers, Garner, and McRae (1980) also found several hatchlings destroyed by fire ants (Solenopsis spp.). Many predators follow roads, fire breaks, and ecotones rather than searching at random when they hunt for prey. The likelihood of finding a prey species using this strategy increases as patch size decreases.

Since the gopher tortoise functions as a "keystone modifier" species (sensu Mills, Soule and Doak 1993) in many of the xeric plant communities of Florida, other species are affected by its decline. Gopher tortoises dig a series of burrows through the dry, sandy soil supporting scrub, sandhill, and other xeric upland communities. Many other vertebrates use these burrows to conceal themselves from predators and to escape light, aridity, fire, and extremes in temperature, both in the summer and the winter. These include other TES, such as the gopher frog (Rana capito) and the Florida mouse, both species of special concern in Florida.

Federally threatened eastern indigo snakes (Drymarchon corais couperi) are close burrow associates of the gopher tortoise, and as such, their population has declined with that of the gopher tortoise. These snakes are not restricted to scrub but can be found there when tortoises are also present. Eastern indigo snakes have a much larger home range than do gopher tortoises. In Georgia, individual snakes use many tortoise burrows over a home range that can be several hundred hectares in size(USFWS 1982). These reptiles undergo long seasonal movements, sometimes several kilometers in extent. The ability to move such long distances enables indigos to establish themselves in new areas of suitable habitat. Early estimates suggested that areas need to be at least 200 ha in size (USFWS 1982); but more recently, experts suggest that indigo preserves be at least 4000 ha (USFWS 1998). An area this size could support a population of about 53 male and 210 female indigos, assuming there was total overlap of home ranges between males and females and that all of the preserve area was of suitable habitat. Because of the patchy nature of extant scrub relicts, few sites could sustain a population of indigo snakes exclusively. In most cases, indigos are forced to cross roads and highways while traveling between areas of suitable habitat. Indigos grow to be large (up to 8 ft long), relatively slow-moving snakes (USFWS 1982). Automobile traffic accounts for a significant amount of the mortality in eastern indigo snakes as well as other species like fox squirrels, southern hognose snakes (Moler 1992, Jordan, Wheaton, and Weiher 1995) and gopher tortoises. The docile nature and conspicu ousness of indigos have made them vulnerable to collectors for the pet trade and easy targets for people who kill snakes on sight.

Another TES that has suffered the effects of habitat fragmentation is the Florida scrub jay. Unlike the gopher tortoise and eastern indigo snake, scrub jays are scrub endemics, and are therefore severely limited in habitat and range. Their threatened status is a direct result of habitat loss. Much of the high value real estate in Florida now sits on former scrub jay habitat. Coastal scrub has been cleared for beachfront hotels, resorts, and housing developments. Inland scrub has been converted to citrus groves, pine plantations, and residential areas. Remaining scrub is frag-mented to the point that much of what remains is too small for a family of jays to colonize. Scrub jays use large, well-defined territories that range in size from 1 to 21 ha (2.5 to 52.5 acres) with a mean of 9.0 ha (22.5 acres) (Mitchell 1997). When patch size falls below a certain threshold, jays will not colonize that patch, even if it is of relatively high quality. Woolfenden and Fitzpatrick (1991) estimated the minimum threshold to be 5 ha (12.5 acres). Therefore, only large areas of scrub will sustain any significant scrub jay populations; such areas are becoming more and more scarce. Road hazards may also pose a significant threat to the Florida scrub jay. These birds frequently forage by roadsides, which offer the large openings jays require for foraging. Mortality from collisions with automobiles is common (Dreschel, Smith, and Breininger 1990) and may even factor into the extirpation of small populations.

Fragmentation can pose problems if barriers are created between upland habitats and the wetlands on which TES depend for breeding purposes. For example, amphibians like the gopher frog depend on gopher tortoise burrows in longleaf pine,xeric hammock and Florida scrub for shelter (Godley 1992). For mating purposes, however, these frogs will disperse a mile or more to find breeding ponds, only to return to the same tortoise burrow at the end of the breeding season (Godley 1992). If fragmentation creates a barrier between the xeric habitat and the breeding ponds, amphibians, like the gopher frog, will not be able to reproduce (A. Weakley, Southeast Regional Ecologist, Southeast Regional Office, The Nature Conservancy, professional discussion, 12 May 1995 [hereafter referred to as A. Weakley, 12 May 1995]). A recent study using species-area relationships to determine the effects of fragmentation on richness of vertebrates in Florida scrub found that when the influence of scrub patch size was removed from the analysis, distance (both to permanent water and to the nearest scrub patch) correlated strongly with non-avian vertebrate richness (McCoy and Mushinsky 1994).

When all other factors are equal, larger areas of habitat maintain the largest numbers of TES. A large population reduces the possibility of extinction. Reproductive success and mortality rates of some TES are inextricably linked to habitat quality and may be influenced by human activities and the structural features of communities neighboring scrub. Florida scrub jay populations, for example, are negatively influenced by the proximity of forests due to increased predation by woodland hawks (Breininger et al. 1995). The location and nature of operations for the space program and national defense can affect jays living on National Aeronautics and Space Administration (NASA) property at Kennedy Space Center and at Cape Canaveral Air Station (Breininger et al. 1996). Scrub jays and other TES populations may also be vulnerable to natural catastrophes, such as hurricanes or other climatic events, disease epidemics or heavy parasite infestations, especially when the colonies are small or closely packed. Therefore, populations in large, contiguous landscapes have the best chances of withstanding these pressures, ensuring their long term survival. If restoration efforts are successful, this will result in a healthier, more contiguous ecosystem and would allow for the continua tion of natural processes across the landscape. Such restoration would also permit more effective management at the landscape level (e.g., landscape-level burns).

Fragmentation of a plant community that requires fire can cause problems with fire management. For example, if natural communities requiring fire management are within city limits or are surrounded by housing, prescribed burning may not be feasible because of smoke management problems (A. Weakley, 12 May 1995). The smoke caused by prescribed fire may drift across nearby highways or landing strips, causing visibility problems for drivers and pilots, or may blow into residential areas, where it may pose health problems for the elderly and others with respiratory difficulties. The possibility of prescribed fire jumping fire breaks is also a dilemma for managers due to the high-intensity nature of fire in Florida scrub. Such fires areextremely difficult to control. The danger of fire escaping when the burn area is in close proximity to residential areas, munitions depots, or other developed real estate must be considered when using fire to manage small patches of scrub.

Management Recommendations

The creation of large patches of scrub habitat should be the focus of Florida scrub restoration activities. Florida managers should attempt to restore those degraded areas that are adjacent to areas of high-quality habitat even, if these areas cannot connect two larger patches. Depending on the species being managed, corridors may or may not be desirable. Regardless of whether tracts are being connected or not, blocks of habitat should be as round as possible. This will further minimize edge effects on TES populations.

At the same time, further fragmentation of natural communities should be avoided whenever possible. Scrub communities have already suffered fragmentation by manmade fire breaks and roads. Myers (1990) states that many of the rarer scrub endemics seem to proliferate in fire plow-lines, permanent fire breaks, and other areas that have been mechanically disturbed. Fire lines that divide a larger burn unit into several smaller ones for purposes of fire management could be left to mimic the natural discontinuity of scrub habitat (see mechanical disturbances section under recommendations regarding fire). However, ecotones should never be disturbed if natural systems management is a goal. Fire breaks constructed through these transitional zones only exacerbate fragmentation issues. Existing roads and plow-lines that run along natural ecotones should be abandoned and the ecotones allowed to recover with natural vegetation. Fire ditches should be filled with the appropriate soil and restored to the original grade (Christman 1995).

If natural areas must be cleared or developed, these activities should be concen trated in one area, preferably adjacent to areas that have already been developed, and not spread throughout natural communities. This will minimize edge effects caused by fragmentation. Activities that will interrupt TES population processes, or ecosystem processes, should be avoided, or an alternative location for the activity should be sought. For example, avoid fragmentation activities such as road building, that will create barriers between connected habitats used by a species.

Fire and Fire Suppression

Historically, fires resulting from lightning strikes occurred frequently in the southeast, burning longleaf pine communities during the growing season(Abrahamson and Hartnett 1990). When conditions were right, fires in longleaf pine communities were able to ignite adjacent Florida scrubs. Before current fire suppression policies were enacted, and scrub habitats became fragmented by human activities, fires spread naturally, sometimes burning areas the size of several counties (Means and Grow 1985).

Around 1920, the United States Forest Service began promoting active suppression of wildfires; this practice continued until recently (Frost 1993). This had drastic effects on vegetation and community structure in longleaf pine communities. Fire suppression in most inland Florida scrub communities has not been as detrimental as it has been in adjacent sandhills, because scrub has a naturally long fire return interval (estimates of 15 to 100 years) and fires are difficult to ignite (Myers 1990). But many inland scrubs are in need of burning. Human encroachment has limited the possibility of returning a completely natural fire regime to this ecosystem. The potential consequences of allowing naturally ignited wildfires to burn unchecked and unattended limit a "let-burn" response to wildfires in these areas. Long term persistence of Florida scrub will require controlled burning and/or an acceptable mechanical alternative to fire.

Impacts

Inland Florida sand pine-dominated scrubs generally are dominated by serotinous sand pine and are fire-structured communities. Coastal Florida sand pine-dominated scrubs are dominated by non-serotinous sand pine (Myers 1990); these communities may not be fire-structured. The following discussion relates to fire-structured scrub communities, with or without a canopy of sand pine.

Both long-term fire suppression, and fires that occur too frequently, can change the structure and composition of the community. Long-term fire suppression may result in succession to xeric hammock, as oaks dominate over sand pine. Fires that occur too frequently will prevent the regeneration of sand pine. Some scrubby flatwoods sites occurring in Florida today are a result of clearcutting sand pine and burning so frequently that sand pine was unable to reseed. Though the species is termed "fire resistant," fire in sand pine communities will kill individual trees. Sand pine regenerates profusely following fire when the serotinous cones open and release their seeds. New individuals can produce cones in as early as 5 years and may live to be 100 years old. In the absence of fire, stands typically begin to break up after 50 to 70 years (Myers 1990); old trees develop heartrot, which makes them susceptible to windthrow. Fire in rosemary-dominated scrubs occurs very rarely because these shrubs are slow to build up enough fuel to ignite. Rosemary is also killed by fire, and depends on seeding for reestablishment. Other scrub species thatrecover through establishment of new seedlings include scrub balm species, Ash's calamint, and short-leaved rosemary (Conradina brevifolia; federally listed as endangered). However, many shrubs characteristic of Florida scrub undergo vegetative reproduction rather than seeding following fire, including oaks, saw palmetto, and scrub palmetto (Stout and Marion 1993).

Many rarer scrub plant species seem to require the open patches of bare sand left after a fire to establish themselves. Hawkes and Menges (1996) looked at the role that open spaces in scrub play in the development and maintenance of species assemblages. They found the rosemary phase, which dominates many of the higher elevation sites, had many endemic or endangered herbs growing in light gaps and that the densities of most obligate-seeders or mixed strategy herbs increased as the amount of open spaced increased. The lower elevation sites had a more closed structure, and showed lower herb densities, with more resprouting species evident (Hawkes and Menges 1996). Gaps in lower elevation sites began to close after 10 years since the last fire, whereas gaps in the rosemary phase persisted for decades. This may explain why seeders and herbs prevail in rosemary-dominated scrub communities.

The processes that determine gap dynamics and abundance may influence the distribution and abundance of many scrub species (Menges and Hawkes, in press). The presence of these gaps may even be a more important predictor of species abundance than fire. These open microsites are created when large fuel loads burn in a fire. The long period of intense heat that results from burning slash piles, wind blown snags, or other large, concentrated fuel sources, kills the rhizomes and other parts of plants that would otherwise resprout after a fast-moving fire. This leaves an open area that can then be exploited by species reproducing via seed. Alterna tively, these gaps are created by road building and other mechanical disturbances, such as blowouts in coastal scrubs, pits and mounds created by treefalls, or animal burrows and mounds (e.g., gopher tortoise burrows; reviewed in Myers 1990). Wireweed (Polygonella basiramia; federally listed as endangered) grows in bare sand patches within sand pine or Florida rosemary scrub, and management information states that "an adequate supply of open areas are needed between large shrubs" for habitat (Howie 1994). Similarly, large-leaved jointweed (Polygonella macrophylla; a federal SAR) grows in sandy soil in areas where there are natural openings in the sand pine canopy, or along paths and powerline rights-of-way. Florida perforate cladonia (Cladonia perforata; federally listed as endangered) also inhabits sunny, bare sand between scrub vegetation (USFWS 1995).

Many studies have been conducted on the effects of fire on scrub and the post-fire recovery of scrub vegetation, particularly in inland, peninsular scrubs. Scrub isconsidered highly resilient after fire. Many scrub species, including oaks, palmettos, and ericads resprout rapidly after such an event and return to preburn levels within a few years. One study by Schmalzer and Hinkle (1992) looked at a coastal peninsula oak-saw palmetto on Merritt Island. They sampled a total of 24 transects in 4 study sites before the burn and at 6, 12, 18, 24, and 36 months after the fire and determined that the changes in the oak-palmetto stands were minimal. While the oak dominated transects showed the greatest resistance to change, the saw palmetto-dominated transects showed the fastest rate of recovery. Many other communities show marked changes as a result of fire, but no species were lost due to the fire in this scrub, nor were there any new species present that were not there before the burn. Fires in scrub revitalizes the populations of the plants and shrubs found in this community; it does not initiate sucession in the classical sense. This illustrates how scrub plants are well-adapted to fire.

Whether or not many scrub plants depend on fire as the only mechanism to maintain the structural features of this communtiy is not clear. A study comparing effects of fire followed by salvage logging, with effects of mechanical treatments (combinations of clearcutting, rollerchopping, broadcast seeding, and brack seeding) in the absence of fire, has shown that many scrub species (though not all) respond similarly to both treatments (Greenberg et al. 1995), but a burn-only treatment was not included in this study. For example, Greenberg et al. (1995) found that Florida rosemary germinated in sites that were both burned and salvage logged, and in sites that were mechanically disturbed but not burned, suggesting that mechanical disturbances, or the consequent conditions, may stimulate seed germination. Similarly, many rare scrub plants proliferate after mechanical disturbances to the soil such as plowed fire breaks and sand roads. Some of the largest populations occur in these areas and where scrub was cleared in the recent past (Myers 1990). However, Greenberg et al. (1995) observed some response differences between treatments, which suggested that effects of fire differ from those of mechanical treatments alone: saw palmetto recovered significantly better in sites that were both burned and salvage logged than it did in mechanically treated sites; similarly, rusty lyonia densities were higher in sites that were both burned and salvage logged compared to only mechanically treated (Greenberg et al. 1995). Increased post-fire vigor has been documented in 14 scrub plants (Berry and Menges 1995).

Though there are few studies of the effects of fire on scrub soils, elevated levels of phosphorus have been found in these soils following fire. Berry and Menges (1995) believe that in the absence of fire, the growth and reproduction of scrub plants is limited by phosphorous availability. They are using a fertilization experiment to determine which resources limit the vigor of the dominant scrub plants. Schmalzerand Hinkle (1991) found few changes after fire but did note a delayed increase in amonium-nitrogen and nitrate nitrogen.

Little information is available regarding how various scrub plant TES will react to fire. If scrub species resprout following fire, populations should be able to recover quickly. Species that persist as a seed bank between fires should easily recover as well, and fire may stimulate their germination and growth. However, species that must seed into scrub from refugia are likely to be at the greatest risk from fire, because scrub habitat is very limited in area, and many rare scrub plant species are confined to specific types of scrubs (e.g., only rosemary scrubs) or have restricted geographic ranges. Most scrubs recover quickly following a fire.

Florida Rosemary scrub is one phase of Florida scrub that does not recover quickly after fire, but some of the endangered plants that are associated with rosemary scrub, such as wireweed, show greater density and seed production when open areas between rosemary shrubs are abundant (Hawkes and Menges 1995; Menges and Kimmich 1996). Wedge-leaved snakeroot, another endangered herbaceous endemic scrub perennial, also shows the greatest survival when the densities of Florida rosemary are low following a fire. These plants take advantage of the open areas left after a fire and repopulate burned areas through both reseeding and resprouting reproductive strategies (Menges and Kimmich 1996). Their survival decreases as the distance to the nearest Florida rosemary shrub decreases. This may be due to some allelopathic properties of the shrubs or other below-ground competition.

Fire may pose a severe threat to the Florida perforate lichen (USFWS 1995), which is destroyed by fire, and must recolonize a site by spores and fragments (Johnson and Abrahamson 1990). It takes decades for lichens to reestablish following death from fire (Johnson and Abrahamson 1990). Studies have shown that other non-vascular taxa were more abundant in mature forest than in mechanically disturbed or burned sites (Greenberg et al. 1995). Several vascular plants are listed in the Recovery Plan for Nineteen Central Florida Scrub and High Pineland Plants (revised) (USFWS 1995) as not likely to survive fire (e.g., scrub lupine [Lupinus aridorum; federally listed as endangered], paper-like nailwort, short-leaved rosemary, Highland's scrub hypericum [Hypericum cumulicola; federally listed as endangered], and Lewton's polygala [Polygala lewtonii; federally listed as endangered]), but none of these species are known to occur on installations. Wireweed may or may not survive fires (USFWS 1995). Long-lived herbs that presumably resprout following fire include: pigeon-wings (Clitoria fragrans; federally listed as threatened), and scrub buckwheat (Eriogonum longifolium var. gnaphalifolium; federally listed as threatened; USFWS 1995), and Eulophia(Pteroglossaspis ecristata; a former federal candidate species), a forb that benefits from prescribed burning (Russo et al. 1993).

Scrub fauna also depend on areas created by fire and other, localized disturbances in scrub. Gopher tortoises prefer and attain their highest densities in grassy, sparsely canopied sites (Auffenberg and Franz 1982), and even though they typically forage in nearby sandhills that support more grasses and ground cover, gopher tortoises often burrow in the light gaps within scrub. In the absence of fire, young fire-intolerant scrub oaks will become established in the bare patches and may shade out any ground cover used by the tortoises here. Fire frequency is the single most important factor influencing habitat structure and food sources for the gopher tortoise in Florida scrub.

Similarly, open patches are also important as foraging sites for Florida scrub jays. Scrub jays forage as familial groups and require low shrubs and scattered open areas to do this. They use these clearings between scrub thickets for safety purposes while searching for food. One member of the group perches in a conspicuous place and acts as a sentinel as the remaining group members forage. If the sentinel detects a predator, this individual alerts the others with a distinctive warning call and the entire group flees to denser cover (McGowen and Woolfenden 1989). Without an open understory, this warning system does not work, and the jays are more vulnerable to predators stalking them from concealed vantage points. Therefore, jays avoid communities with a closed understory and will abandon open scrub when a heavy pine canopy develops, or when the height of the shrub layer reaches critical levels (Myers 1990). These birds use these open areas to store food as well. Scrub jays are omnivorous, but acorns serve as their principle plant food (USFWS 1990b). They harvest the acorns right off the oak trees in the fall and bury them by the thousands in these bare, sandy patches between the shrubs (DeGange et al. 1989). These cached acorns then serve as a food source through out the year.

Because prime jay habitat is maintained by fire, scrub jay populations will begin to decline when their territories show signs of succession. Data collected on the Merritt Island populations at Kennedy Space Center, Cape Canaveral Air Station and the Archbold Biological Station indicate that even large resident populations are declining due to habitat degradation (Mitchell 1997). Fire suppression in remnant areas of scrub reduces the jays' ability to survive and reproduce (Woolfenden and Fitzpatrick 1991). Tracts experiencing infrequent fires or fire suppression become overgrown. Tall, dense oak understories and a closed canopy of sand pines result. Preferred scrub jay habitat has scrub with oaks 1 to 3 m tall covering 50 to 75 percent of the area, and no more than 15 percent canopy cover by trees. Prescribed fire will prohibit scrub from undergoing succession to a closed canopied communityand will prevent shrubs from reaching a height that is unsuitable for jays. Woolfenden and Fitzpatrick (1991) estimate that about 50 years post-fire is the longest scrub jays would stay in an area before the habitat became unusable.

One mammalian TES, the Florida mouse, is listed as a species of special concern by the state of Florida and is under review for listing by the U.S. Fish and Wildlife Service (Humphrey 1992). Its decline is associated with habitat loss and fire suppression. This mouse is restricted to the fire-maintained xeric communities of peninsular Florida and has one of the smallest geographic ranges of any mammal. The Florida mouse is another burrow-dwelling species that takes advantage of pre-existing gopher tortoise burrows (Layne and Jackson 1994). Florida mouse populations are highest in the early successional stages of scrub following a catastrophic fire (Humphrey 1992). Like the scrub jay, their numbers begin to decline as the scrub becomes more shady and dense. The number of years post-fire the habitat remains suitable for the Florida mouse is variable. These mice have persisted in one scrub located in Levy County Florida for 44 years since the last fire, but population density suffered a sharp decline about 10 years after the fire (Humphrey 1992). Declines in acorn yields from the scrub oaks is probably one factor that accounts for the reduced suitability of long unburned, overgrown scrub habitat for both the Florida mouse and Florida scrub jay (Humphrey 1992).

Management Recommendations

Available information suggests that fire is essential for the maintenance of most Florida scrubs. Natural fires are catastrophic in scrub (meaning they kill off the canopy trees), but without fire, pyrogenic inland scrubs will succeed to a different community and will no longer support scrub-dependent TES populations. Therefore, instituting a fire management plan that mimics natural processes is the best way for managers to restore and maintain Florida scrub. There is no specific fire return interval applicable to all scrubs, and there is no definitive way a natural scrub should always appear. Management guidelines should reflect the desire to maintain patches of scrub in different stages, each providing important structural attributes, across the entire landscape. Thus, managers will need to apply fire in a way that promotes all the various stages of development in large fragments of scrub. This practice will allow managers to maintain populations of all the TES species having special habitat requirements (reviewed in Myers 1990, Christman 1995). For example, Christman (1995) suggested that Camp Blanding scrubs are in need of management (prescribed fire) when more than 50 percent of the scrub area that could theoretically support scrub jays has become unable to do so because of vegetative growth since last fire (this figure is subject to adjustment).

Prescribed fires must mimic natural fires in that the burned areas should not be contiguous. Some patches must remain unburned so all species will have refugia, forage, and cover. According to Woolfenden and Fitzpatrick (1991), optimal fire frequency to maintain scrub jay habitat is every 8 to 20 years. They claim that intervals less than 8 years may keep young oaks below acorn-bearing height and could favor the spread of palmettos, which could eventually replace the oaks. However on Merrit Island, scrub oaks frequently bear acorns only 2 to 3 years post-fire (P. Schmalzer 12 May 1998). Therefore, more frequent fire intervals may be appropriate under some circumstances. Christman (1995) suggested that Camp Blanding scrubs should have bare sand covering more than 10 percent of the ground. This recommendation would be achieved with the fire interval used to maintain jay habitat. Florida mice also favor the conditions that support Florida scrub jays and a fire interval that promotes these conditions. Habitat maintained in this way is ideal for other scrub-adapted vertebrates as well. Most scrub endemic shrubs and forbs also prefer these conditions (reviewed in Christman 1995). Therefore we also recommend that scrub be burned at the 8- to 20-year interval suggested by Woolfenden and Fitzpatrick (1991) to manage scrub for TES on military installa tions, unless the site is one of particularly low productivity (e.g., rosemary scrub), in which case fires may be of an even longer interval.

These recommendations however, are not universal. It is critical that the possible effects on other scrub endemics be considered when using the scrub jay (or any other organism) as an indicator species for scrub management. Those natural resource personnel managing a site for two or more scrub endemic species with incompatible habitat requirements will need to incorporate these differences into the design of their management plans and fire schedule. In this instance, managers must devise a plan that will promote the survival of all TES species living in the area. For example, consider the negative correlation between abundances of Florida scrub jays and sand skinks reported by Mushinsky and McCoy (1995). This phenomenon could pose a real problem for managers because, although both species seem to have similar habitat requirements, not enough is known about the relationship between the two to allow managers to make compromises that will be constructive on a bi-lateral basis. For now, the best solution is to look for creative ways in which to manage an area for both species. Since the sand skink and scrub jay are able to coexist on Archblod Biological Station, we should look there for possible alternatives. Sand skinks have been able to persist on Archbold in scrub stands greater than 60 years old. These animals are found here in the vicinity of firelanes and other artificial clearings (USFWS 1998). This suggests these artificial clearings may be important to sand skinks. Managers charged with maintaining populations of both species at one installation could leave some old growth stands of scrub unburned. Because these areas are useless to jays anyway, augmenting them with selectiveartificial clearings and designating them as sand skink habitat is one way to promote both species at one site. This solution shows the importance of incorporat ing an adaptive management strategy into any management plan.

A shorter fire return interval than the recommended 8- to 20-year period may be necessary if the management goal is to restore scrub in areas that have been left unburned for decades and are severely overgrown. Restoration efforts should include initial frequent fires (every 2 to 4 years) for the first 10 to 15 years (Breininger et al. 1996). This increased fire frequency is needed to reduce the accumulation of underground carbohydrate reserves in enough patches for the openings to return. The ratio of dead to live fuels is important in igniting and sustaining fires in scrub. Breininger et al (1996) believe that 25 percent of the above-ground biomass needs to be dead for effective fires. Thus at least 2 years between burns are necessary when restoring areas with abundant palmetto. At least 3 to 4 years are required between burns for areas of oak scrub with sparse palmettos.

The authors agree that a long term experimental approach will be needed to reestablish openings in densely overgrown scrub, but recommend a variety of ignition techniques and prescriptions to be used to help facilitate early restoration fires (Breininger et al. 1996):

1. Use small burn units in a few experimental landscapes.

2. Use ignition strips of mechanically chopped fine fuels.

3. Use narrow stripped, flank fires.

4. Use aerial ignition techniques.

These techniques should be used when land managers need to restore severely degraded scrub. Excepts from Christman (1995), Management of Florida Scrub at Camp Blanding, Clay County, Florida, provide specific techniques for maintaining relatively high-quality Florida scrub. Christman (1995) also explained the advantages and disadvantages of using head fires and backing fires to burn scrub and provided information on season of burn and creation and use of firebreaks:

If the management goal...is to maintain and restore natural communi ties, prescribed fires in scrub should be set in the growing season in adjacent upland communities and allowed to burn with the wind through the scrub and into natural wetland firebreaks.

Head fires die out variably as they enter the ecotone with wetlands and this maintains the natural variability of the ecotone. Head fires leave some areas intensely burned and others unburned, creating the habitat mosaic that insures survival of all scrub species. Backing fires, in contrast, tend to burn the groundcover more completely and homoge neously, and to burn hotter at ground level, possibly killing animals and plant seeds near the soil surface. Furthermore, backing fires are difficult to maintain in scrub unless fuels are especially dry.

On the other hand, backing fires are easier to control. In cases where maximum control of fire is imperative, backing fires may have to be used.

The use of head fires set in adjacent areas and allowed to burn uncontrolled into scrub is a theoretical technique used to mimic the way natural fires move through these communities. Managers may not be able to apply prescribed fire in this way because scrub is among the most difficult and dangerous plant communities to burn. When igniting scrub, prescribed burners are often fortunate to be able to accomplish their objectives without such fires escaping.

Low-intensity backing fires can be used for safety, but they have met with limited success. Traditionally, scrub will not burn in a backing fire because these fires are difficult to maintain. Backing fires are useless to managers unless they can get the fire into the shrub layer. Fire practitioners have many variables to consider in writing a prescription for a burn, including wind speed, direction, and variability; fuel moisture in both live and dead fuels; air temperature and relative humidity; atmospheric stability and instability; phenology or life stage of the plant and the timing of the burn; fuel size, continuity, load and fuel chemistry as well as a host of possible ignition techniques. Researchers and fire practitioners are beginning to figure out the relationships between these variables, but often the conditions that can sustain a successful backing fire do not exist when the fire is necessary to meet the objectives of the burn.

Managers must have a clear idea of why and when they want to burn and what they want to accomplish with the fire. They should have a specific set of objectives for a fire to meet, and should monitor the site after the burn to see if those objectives have been met (Mary Huffman, Program Director for the Lake Wales Ridge Program, The Nature Conservancy, Lake Wales Ridge, FL, professional discussion, 12 March 1998). Under certain conditions, managers at TNC's Saddle Blanket Lakes Scrub Preserve have been successful in their use of backing fires to maintain Florida scrub (M. Huffman, 12 March 1998), but in general, these fires are not effective in scrub management.

Again from Christman (1995):

Scrubs should be burned during the growing season because that is the period during which most lightning fires occur, and as a result the scrub plants and animals have become adapted to that regime. The best time to burn Florida scrub is in March, April or May...Long unburned scrub cannot be managed with winter fires...such fires, if they can be main tained at all, appear to hasten the degradation of scrub and its conversion to pioneer hammock by eliminating much of the ground layer but little of the larger woody vegetation. Whereas growing season fire in sand pine forest or pioneer hammock will kill the sand pines and above-ground parts of the oaks, thus favoring scrub, fire in the dormant season will burn little but the ground litter.

Abrahamson and Abrahamson (1996) conducted a study on the effects of such a low-intensity, winter burn in a long-unburned scrub on the Archbold Biological Station. Their 7-year examination showed there was little change in the floristic composition of post-burn stands compared to pre-burn stands, despite a wide variety of recovery strategies to such a burn. The populations of endemic herbaceous plants in this scrub were unable to be restored. Sand pine was likewise unable to regenerate itself as the dominant species in the canopy.

These results suggest that only fires occurring during the growing season will effectively promote scrub regeneration and halt succession to other cover types. But according to the most recent management practices and data collected on Merrit Island, it is the intensity of the fire, not the season in which it occurs that dictates the success of prescribed burns at restoring and maintaining scrub. Under the proper burning conditions (i.e., appropriate fuel moisture, humidity, temperature, and wind speed), managers at Kennedy Space Center have successfully restored long-unburned scrub by igniting high-intensity fires in November, February, April, and other times of the year (P. Schmalzer, 12 May 1998). Natural resourcepersonnel at Avon Park Air Force Range have also had success with non-growing season fires in maintaining scrub at this installation (Jim Orzell, Botanist, Avon Park Air Force Range, professional discussion, 26 June 1998). Again, adaptive management is the key to successfully maintaining scrub habitat.

Christman (1995) goes on to say:

The timing of subsequent prescribed fires in scrub should be variable, because no single fire-return interval could support the diversity of fire-recovery strategies and habitat preferences observed in native scrub species...

Prescribed fires are best ignited along existing roads [except those running through ecotones], and allowed to burn up to and through the scrub and into natural wetland firebreaks...If artificial firebreaks, either baselines (where the fire is started) or control lines (meant to stop the spread of fire), must be constructed, they should be temporary, created by various combinations of mowing, crushing, burning, or fire suppressant foam. Roller chopping with heavy drums should be avoided because it can adversely affect animal habitats and soil processes, damage plant roots and rhizomes upon which scrub regeneration is dependent, and create habitat for invasive species. Roller chopping with empty roller drums may be acceptable where mowing is impractical.

...Experience at The Nature Conservancy's Tiger Creek Preserve in central Florida has shown that when it is dry enough to ignite a scrub, it may be too dry to use natural wetlands as a firebreak because the duff and humus in the wetland/scrub ecotone may catch fire and smolder for weeks, causing unacceptable smoke problems on adjacent lands...This may not be a problem [where] scrubs...are relatively isolated on roads and residential areas. Certainly the humus and duff in wetlands, and especially in wetland ecotones, burned under natural conditions prior to management by modern man. Today's accumulation of surface organic matter...is probably greater than at most times in the past...Reduction of duff exposes bare mineral soil, which favors establishment of fire-dependent plant species. Some reduction of accumulated duff in wetland/scrub ecotones should be a goal of natural systems management.

On the other hand, if smoke from smoldering duff and humus is deemed unacceptable, managers may wish to pre-burn above ground vegetation in wetland firebreaks when it is possible to do so without igniting the humus, then burn the scrub toward the wetland on a later, drier date when the scrub will burn. By pre-burning the wetland firebreak when the Keech/Byram Drought Index (Keech and Byram 1968) is <350,managers can reduce fuel there without igniting the humus (Melton 1989).

In general, prescribed fires in Florida scrub should be ignited early in the growing season (March-May) while the vegetation is still relatively dry. If igniting the duff within the wetland firebreak must be avoided, the scrub should be burned when the wetlands are flooded and the duff is fireproof but the scrub itself is dry enough to burn. Prescribed fires in scrub that will not depend on available wetlands for firebreaks can be set anytime between March and July.

Managers should apply fire to maintain all of the various stages of scrub within the community. This is a challenging problem, because fire in scrub is high-intensity and often occurs under extreme weather conditions, thus it exhibits uncontrollable and unpredictable behavior (Cox and Roberts 1995). For this reason, wildfires have been the rule, and there is little literature available on prescribed burning in scrub (reviewed in Christman 1995; Doren, Richardson, and Roberts 1987; Kenner 1994). The best control of fire in scrub is proper planning (Christman 1995), as plow-lines are often ineffective in scrub, and attempts to control the fire in scrub by using people, equipment, and tools after the fire has begun often have been futile (Doren, Richardson, and Roberts 1987).

Managers are beginning to use fuel models to predict fire behavior in scrub (Doren, Richardson, and Roberts 1987; Cox and Roberts 1995). Environmental parameters, such as relative humidity, temperature, etc., are input into fuel models to predict the rate of spread and intensity of a fire (Doren, Richardson, and Roberts 1987). In addition, an available computer program (RXWINDOW) can use input from fuel models to determine the best environmental conditions (e.g., humidity, wind speed) for achieving desired results (e.g., tree mortality, flame length, rate of spread, and intensity) from prescribed burns (Christman 1995).

There is not a specific fuel model for Florida scrub. The National Forest Fire Laboratory (NFFL) fuel model for chaparral/high pocosin/mature scrub has been used with success for prescribing fire in the Yamato scrub, one of the last vestiges of this community along the Atlantic Coastal Ridge located near Boca Raton (Doren, Richardson, and Roberts 1987) and in sand pine forest of the Ocala National Forest, Marion County (reviewed in Christman 1995). Methods for burning the Yamato scrub (Doren, Richardson, and Roberts 1987) are presented:

Fire was prescribed to the Yamato Scrub, FL, based on predictions from the National Forest Fire Laboratory (NFFL) fuel model for chapar ral/high pocosin/mature scrub. This model was chosen because itpresupposed vegetation structure and qualities similar to those for the Yamato scrub. Based on model predictions for maximum spotting distance and probability of ignition, it was decided that some site preparation was necessary to ensure a successful burn. Strips of crushed vegetation were created around and/or across proposed burn sites using a small, empty roller drum chopper weeks before the burn. The number of strips used depended on the size of the site being burned. One pass over the vegetation, with drums set parallel to each other, produced a light slash, without appreciable soil disturbance or fuel compaction. Within the chopped areas, sand pines were dropped on site to reduce spotting potential and create hot spots to prepare a seed bed and open areas for vegetation and wildlife. About 2 weeks without rain were required to dry out the crushed fuels sufficiently and reduce fuel moisture to prescribed levels. When weather conditions were favorable, managers torched the downwind edge of the crushed area, then moved up each side until the first alternating chopped strip was reached, then fired across the strip, creating a head fire through the unchopped scrub area. This resulted in alternately blacklining (burning out fuels around the fire) and head firing the entire area, which aided in reducing overall intensity and direction of each headfire, and created manageable smoke and fire conditions. Managers were able to burn most of one site and all of a second site, because weather conditions were favorable. Develop ment of unfavorable weather conditions as the day progressed prevented complete burning of the first site, and all further ignition attempts were futile. The effects of roller chopping on the soil were minimal...The fire behavior predictions were remarkably accurate for the prescribed burns.

Christman (1995) also described the use of prescribed fire to burn several overgrown scrubs on Merritt Island:

On Merritt Island, managers used a Brown tree cutter, a D-6 Caterpillar with a V-blade, or a roller chopper to prepare strips and blocks within several long-unburned oak scrubs prior to burning (Schmalzer et al. 1994). After drying for a week or two the crushed strips were easily ignited with a drip torch. By the time the heading fire reached the uncut scrub it had built up sufficient intensity to carry into the standing vegetation. Managers at Merritt Island reported that the Brown tree cutter provided the best results, producing the best fuel bed with almost no soil disturbance.

Christman (1995) provided other examples of methods used to prescribe fires in Florida scrub. At Saddle Blanket Lakes Scrub Preserve, temporary firebreaks were created by mowing, and then burning the mowed strips. In sand pine scrub at Archbold biological station in Highlands County, and in sand pine forest in the Ocala National Forest in Marion County, linear backfiring was used to develop burned strips that would serve as baselines and control lines. Scrub at Oscar Scherer State Recreation Area in Sarasota County was mowed with a Brown tree cutter prior to burning (reviewed in Christman 1995).

An alternative method for restoring Florida scrub was used with some success at Blue Spring State Park (Kenner 1994). The use of a hot, fast-moving crown fire was a problem in this case, because managers didn't know how to keep these fires from damaging nearby housing and park facilities. Therefore, a combination of clearcutting sand pine to reduce fuel load was followed by a prescribed burn.

The 10-ha tract of scrub was clear-cut in the fall of 1989 and was burned in March 1990. Comparisons of pre-treatment and post-treatment data indicated a significant increase in plant diversity, especially grasses and legumes. A rise in animal diversity and abundance was also observed, including appearances of scrub jays on the tract in 1992, with two becoming permanent residents by summer of 1993. A Florida scrub lizard and an Eastern coach-whip, new species for the park, were also seen in 1992. Although no tortoise burrows were found in the park the day after the area was burned (they were rare in the park), the return of gopher tortoises has since been observed, and at least 18 burrows have been located. Some substrate disturbance occurred around the staging and loading area, but no exotics were observed to date.

Kenner (1994) provided management recommendations based on results from this study:

· Cut sand pine in February, March, or April. This allows the slash approximately three months to dry, so that it may be burned before excessive sprouting occurs.

· Cut the trees with a rotary feller-buncher. This will reduce the dense understory shrub layer to a fuel texture and arrangement that carries fire well under low to moderate wind conditions. List the use of a rotary feller-buncher as a requirement in the timber sale contract.

· Limb trees where they fall, to evenly distribute slash over the site. Do not use limb-removal devices that lead to the accumulation of large amounts of slash in a small area.

· Locate the loading area on previously disturbed sites whenever possible, to minimize the extent of ground disturbance caused by skidders and trucks.

Another study by Boyle, Schmalzer, and Adrian (1998) compared the use of fire and a combination of mechanical cutting and fire on the restoration of long-unburned scrub on Kennedy Space Center/Merrit Island National Wildlife Refuge. They used 15-m transects to sample the restoration sites pre-treatment and at 6-month intervals post-burn. Both the scrub oaks and saw palmettos resproughted after all treatments, but the palmetto recovery was less in all treatments involving mechanical cutting than with fire alone. Recovery of oaks was similar in all treatments. Persistent openings were produced only when cut brush was piled for burning and the prolonged heat killed the rhizomes and roots of the shrubs (Boyle, Schmalzer, and Adrian, 1998).

Examples have shown that prescribed fire can be used to restore scrub habitats if burns are carefully planned and applied. If fire is applied in conjunction with clearcutting or other forms of mechanical disturbance, methods causing the least amount of mechanical disturbance to the soil are preferred, as soil disturbances may lead to invasion by weedy species (Eric Menges, Senior Research Biologist, Archbold Biological Station, professional discussion, 5 February 1996).

In cases where use of fire in scrub is not feasible, managers should consider the use of mechanical disturbance as a tool to maintain scrub. The response of scrubs that have been chopped or mowed but not rootraked appears to be similar to the response from burning, but more study is needed. Significant differences between fire and mechanical disturbance without fire in Florida scrubs have not been found, but studies have had poor controls. Florida Park Service biologists at Jonathan Dickinson State Park are addressing this question (D. Roberts, Biologist, Florida Park Service, professional discussion, 11 March 1998). The U.S. Fish and Wildlife Service is funding their research for management of listed species in Florida scrub. The researchers have set up experimental plots in mature scrub to compare different treatments of mechanical disturbance, fire, and the two treatments combined. In one treatment, biologists at the park used a Brown tree cutter to remove the trees in a plot they subsequently burned with prescribed fire. They allowed that fire to burn into an adjacent plot of undisturbed sand pine to compare responses of scrub vegetation. In another plot they shredded 100 percent of the understory biomasswith an Alamo shredder, leaving the sand pine overstory and a mulch that will contribute to nutrient cycling. They will compare the results of this treatment with those of a fire. The data have not been analyzed yet to determine the responses, but managers should keep abreast of new information from this study and others.

The U.S. Forest Service has been able to maintain Florida scrub jay habitat by mechanical harvesting and regenerating sand pine in the Ocala National Forest (USFWS 1990b) without the use of fire. Managers used clearcutting techniques to remove the sand pine in blocks of 120 acres. New cuts are made adjacent to older cuts that have been mechanically reseeded to make large areas of suitable jay habitat. Jays are able to use these areas for about 10 years before the pines mature to the point where the habitat is no longer suitable. As much as 20 percent of forest is available to jays at any one time. The Ocala National Forest boasts more than 700 breeding pairs - the largest population in the state (L. Lowery, 25 February 1998). The USFS is able to maintain populations of other vertebrate TES at this location as well, including the eastern indigo snake, Florida mouse, Florida scrub lizard, short tailed snake, gopher frog, Florida pine snake, and sand skink. Problems with soil compaction and disturbance are minimal (L. Lowery, 25 February 1998). In areas where prescribed burning is not an option, this method of managing scrub could be used as an economically profitable alternative. When using these mechanical disturbance techniques to maintain scrub, care must be taken to minimize damage to plant root systems and rare ground lichens. (See Groundcover Disturbances.)

This type of disturbance may be helpful in managing rare plant species as well. Menges and Kimmich (1996) recommend that rosemary scrub patches with wedge-leaved snakeroot be burned every 10 to 25 years or by such time that the gaps between the shrubs become small enough so that most snakeroot plants are within 60 cm of the shrubs. However, they also state that because this plant has been found to grow vigorously along firelanes and in other areas where the soil has been subjected to disturbance, mechanical disturbance could well be used to manage this plant.

There is one final point on this subject: scrub has an evolutionary history involving adaptation to fire. At this time there is no scientific consensus that scrub can be maintained in the long term by mechanical disturbance alone.

Alteration of Hydrology

Impacts

Altered hydrology is not likely to be an issue for Florida scrub plants. Ground-level disturbances are not likely to interact significantly with local hydrology and these plants are adapted to live on the well-drained, droughty soils supporting scrub. Many scrub species have deep roots to access soil moisture at considerable depths, in addition to a shallow root system, so moisture stress is not a problem. Surpris ingly, flatwoods plants may be more likely to experience water stress than scrub plants, because flatwoods plants are shallow rooted (probably due to the high water table). Nutrient deficiency likely plays a greater role in ecosystem structure in scrub than water deficiency (reviewed in Myers 1990). Altered hydrology may indirectly affect fire frequency in scrub communities if the natural barriers between the scrub and the adjacent fire-maintained communities become dry enough to carry fire. If surrounding wetland areas are drained or their hydrology is otherwise changed to the point where these areas become more xeric, they will no longer function as a barrier to fire and adjacent scrub will burn with an increasing frequency.

Altered hydrology in surrounding wetland areas can negatively affect Florida scrub in another way. Many fresh water marshes in Florida burn readily and act as an ignition source for scrub in years when it has a fuel load high enough to burn. Fire frequency can be reduced in these wetlands when the hydrology is altered to the point where red maple (Acer rubrum), willow (Salix caroliniana), wax myrtle (Myrica cerifera), and groundsel (Baccharis spp.) become established. These and other trees and shrubs are much less prone to fire than those they may replace, and the overall fire frequency in these areas will be lower here, and in adjacent scrub (P. Schmalzer, 12 May 1998).

Management Recommendations

Avoid hydrological alterations that could dry out the surrounding wetland areas that serve as barriers protecting scrub from fires ignited in more pyrogenic communities. Engineers should refrain from hydrological alterations that promote the germina tion of plants that would decrease the fire frequency in these wetland areas.

Groundcover Disturbances

Impacts

Many scrub species show a similar response to burning and to mechanical disturbances, suggesting that the mechanical disturbances resulting from clear-cutting may mimic the natural disturbance process (Greenberg et al. 1995). Greenberg et al. (1995) note that scrub should be adapted to mechanical distur bance, because coastal scrub is disturbed by sand erosion, deposition, and movement by wind and water, and inland scrub species would have been exposed to similar selective pressures, as they occurred at one time on coastal dunes and barrier islands during higher sea levels. Although many scrub species responded similarly to different treatments, some response differences between treatments were noted by Greenberg et al. (1995). Plant community composition and structural character istics were measured in mature sand pine scrub and in sand pine scrub 5 to 7 years after different disturbance treatments: (1) high intensity burn, salvage logged, and naturally regenerated; (2) clear-cut, roller-chopped, and broadcast seeded; and (3) clear-cut and bracke-seeded. An increase in ruderals (bluestems, Andropogon spp.) and dog-fennel (Eupatorium compositifolium) occurred in all of the disturbance treatments, though this trend is not commonly reported following fire alone in scrub, suggesting that effects of mechanical disturbance (e.g., post-fire salvage logging) differ from effects of fire. Saw palmetto declined significantly following roller chopping, and increased following salvage logging combined with wildfire. Saw palmetto is important in carrying wildfire, even where sparse, and declines in populations could interfere with fire management (Greenberg et al. 1995). Results of this study should be viewed with caution, because it was not conducted at the time scale or spatial scale necessary to determine whether mechanical disturbances in scrub are sustainable. Furthermore, pretreatment data were absent from the study, and there was no control (burn-only treatment; Greenberg et al. 1995).

Another study (Breininger and Schmalzer 1990) documented long-term effects in an oak scrub site that was mechanically cleared more than 20 years earlier. In an adjacent non-disturbed oak/palmetto scrub site, the dominant species (in the over 0.5-m layer) were saw palmetto, myrtle oak, and sand live oak. The disturbed oak scrub, in contrast, was dominated by sand live oak with limited saw palmetto. Additional habitat differences noted in the disturbed oak scrub site included (1) more bare ground, (2) a taller shrub layer, and (3) more herbaceous species. It has been noted that weedy and non-native species (dogfennel, cogon grass) are over represented in scrub on the Ocala National Forest when compared to natural scrubs. This abundance may have resulted from site-preparation activities; scrubs arenaturally resistant to exotic species invasion, except when soils are disturbed (E. Menges, 5 February 1996).

Greenberg et al.'s (1995) study did not discuss effects of silvicultural treatments on rare species. Available information suggests that rare species usually do not recolonize a site by seed once the species is killed by root exposure, and if they do, recolonization is very slow. If adults occur in a site before disturbance and are only top-killed, they normally survive (D. Gordon, 5 February 1996).

Management Recommendations

In scrubs that have become degraded by fire-suppression, managers should devise ways to prescribe fires that will not cause intensive ground disturbance, as structural changes have been noted following intensive silviculture (Greenberg et al. 1995). Examples have been provided (see Fire and Fire Suppression).

If managers must use mechanical disturbances rather than prescribed fire to restore and maintain scrub, care must be taken to cause as little soil disturbance as possible. Bulldozers and other heavy equipment can be too intrusive in areas of scrub that support communities of rare plants and other terrestrial vertebrates. For example, even though adult gopher tortoises can dig burrows that can be 5 to 10 feet deep and 20 feet or more long (USFWS 1990), these burrows cannot protect the tortoises from the weight of a heavy dozer or tank. Lohoefener and Lohmeier (1984) believe that nests and hatchlings are often destroyed by heavy equipment and intensive site preparation activities within tortoise habitat. Sand skinks and blue-tailed mole skinks spend much of their time just beneath the surface of the sand. These TES can be crushed by vehicles weighing far less than a skidder or dozer.

Management activities are designed to improve TES habitat. Ideally, these areas should be cleared and removed by hand, but this recommendation is not realistic in all cases. Managers do not have the time or money to perform large-scale scrub management by hand-clearing trees or vegetation. In the process of performing these activities, losses of individual plants and animals will inevitably be incurred. The costs of using heavy equipment in sensitive areas must be weighed against potential gains to TES populations. Long term gains may be worth the price paid by a few individuals. On the other hand, large scale use of heavy equipment may reduce populations to the point where they cannot recover and the habitat created for them will remain unused. Again, managers should seek a compromise through creative and adaptive solutions to this problem. For example, volunteer labor could be used to hand-clear the most sensitive areas. Hack and squirt applications of herbicide can be used to kill large numbers of trees in a relatively small amount oftime. The trees will die and eventually succumb to windfall, creating an opening for other plants to exploit. Large trees may be girdled and left to die.

Natural scrubs should also be off-limits to off-road vehicles (including four-wheel drive trucks, motorcycles, and all-terrain vehicles [ATVs], unless they are needed during a prescribed burn) and heavy trampling. Many rare scrub species inhabit bare, open areas of sand within the scrub. Without protection, disturbance caused by vehicles and pedestrians is more likely to be concentrated in these areas because they are more accessible than the dense oak, rosemary, or palmetto thickets within the scrub. Finally, there is little information available regarding the effects of disturbance on slow-growing, ground-dwelling lichens in scrub. However, they are likely to be sensitive to disturbance by off-road vehicles and heavy trampling and may require 50 years or more to recover after a single disturbance event (FNAI and FDNR 1990).

Erosion

Topography is the main factor influencing erosion in Florida scrub. The terrain on Camp Blanding, for example, is fairly level and scrub sites here are protected from wind erosion by surrounding woodlands (L. Morris, Camp Blanding, FL, professional discussion, June 1998 ) Because the sandy soil on which scrub exists drains so rapidly, the erosional forces of rain do not seem to be an issue here. But many scrub communities inhabit fossilized dune systems that were historically formed by the erosional forces of the Florida landscape. The bare patches of loose sand that define Florida scrub today may be subject to wind and rain erosion, particularly when these forces follow ground cover disturbances (FNAI and FDNR 1990). Heavy tracked vehicles and motorcycles could cause problems by promoting erosion, but in the absence of regular vehicle use, erosion can be kept to a minimum. Camp Blanding has restricted vehicle traffic in scrublands (L. Morris, June 1998).

Exotic Species

Florida scrub may be less invasible than other vegetation types, because of its demanding physical environment and allelopathic properties associated with scrub dominants (USFWS 1995). However, activities that disturb soil can increase susceptibility of Florida scrub to invasion by species not natural to the community. Coastal scrubs on Cape Canaveral Air Station and elsewhere on the Atlantic Coast barrier islands appear to be more vulnerable than inland scrub to invasion byexotics including Brazilian pepper (Schinus terebinthifolius) and castor bean (Ricinus communis) (P. Schmalzer, 12 May 1998).

Impacts

Scrub habitat will be more susceptible to invasion by exotic plants whenever non-native fill dirt is brought into the community. Roadways built through Florida scrub facilitate the invasion by non-indigenous species (Greenberg, Crownover, and Gordon 1997) because conditions governing competition and survival become markedly altered. The sandy soil on which scrub vegetation grows will not compact well and is therefore inadequate for road building. Engineers often bring in lime rock, clay, and other soils foreign to scrub communities to build a road bed in these areas. Soil pH and levels of several nutrients in roadsides often differ significantly from native soils (Greenberg, Crownover, and Gordon 1997). These conditions allow the transport of source propagules to previously unattainable sites and promote the infiltration of invasive species into scrub habitat. Nutrient enrichment of the low-fertility soils in scrub not only promotes the invasion of non-indigenous species, but often corresponds with a decrease in native species richness (Greenberg, Crownover, and Gordon 1997).

The presence of clay and limestone in scrub can result in potentially higher soil moisture levels. Increased runoff from compacted road surfaces as well as increased water retention capacity due to a high clay content can contribute to the higher incident of weedy species. These species would be otherwise unable to establish themselves in the xeric soils surrounding the road. Improved roads also receive greater vehicle use than unimproved sand roads and the potential for propagules to be transported into these sites from distant seed sources is greatly increased. Some of the non-indigenous species of grasses Greenberg, Crownover, and Gordon (1997) noted as appearing along roadsides in scrub include Champagne (Rhynchelytrum repens [Willd.] C.E. Hubb.), smutt grass (Sporobolus indicus (L.) R. Br.), and centipede grass (Eremochloa ophiuroides [Munro] Hack.), a native of China. They also noted a number of species of herbs indigenous to Florida but uncharacteristic of Florida scrub growing in the study areas. Other exotic plant species either documented as invading or having the potential to invade Florida scrub include Bahia grass (Paspalum notatum), Brazilian pepper (Shinus terebinthifolius), and cogon grass (USFWS 1995). Table 4 lists some activities that may lead to invasion of exotic species.

Cogon grass. Cogon grass has been designated the worst perennial grass weed of southern and eastern Asia and one of the 10 worst weeds worldwide. It has been documented in Louisiana, Mississippi, Alabama, South Carolina, and Florida (reviewed in Coile and Shilling 1993). It becomes established in both pinelands and Florida scrub (USFWS 1995). This species can survive in dry, barren areas where other plants have difficulties, because it has a root system efficient at extracting water and minerals (Coile and Shilling 1993). The spread of cogon grass cannot be controlled using fire (Duever 1989).

Cogon grass is spread by wind-dispersed seed and by rhizomes, which can be transported on equipment (e.g., bulldozers; FNAI 1994b, USFWS 1995). Rhizoma tous spread and allelopathy aid cogon grass in the production of monotypic stands (Coile and Shilling 1993).

In Florida, cogon grass can be observed growing along roadsides, usually in full sun, forming dense stands of yellow-green grass. A quick identification feature is an off-center midrib, which is whitish. This feature is more apparent toward the tips of the leaves. Other features are translucent, dry, rough leaf margins (similar to cutgrass [Leersia sp.]), and the presence of many scale-like non-green leaves on the rhizomes (similar to Johnsongrass [Sorghum halepense]). Plants are usually about 1 m tall, but very rarely grow to 3 m tall, and are similar in appearance to Johnsongrass.

Management Recommendations

In general, the presence of exotics and pests in natural areas should be viewed as indicators of unnatural disturbances affecting the community. Thus, control should be primarily through preventing the conditions that allow for their establishment. Activities listed in Table 3 should be avoided (whenever possible), as these activities increase community susceptibility to invasion by exotics. However, communities may currently have problems with exotics/pests, and management recommendations for their control are outlined below.

Managers should obtain a copy of Langeland (1990), "Exotic Woody Plant Control" and Langeland and Stocker (1997), "Control of Non-Native Plants in Natural Areas of Florida," for information regarding control of exotic plants. These publications can be obtained by contacting the Publications Distribution Center, PO Box 110011, University of Florida, Gainesville, FL, 32611.

In general, manual removal of exotics should cause the least disturbance to the environment, if done carefully. However, manual removal can be labor intensive and may not work for large trees. When using manual methods, every effort must be made to remove the entire root system, because some species can resprout from only a 1/4-inch section of root (Langeland 1990). Only manual removal should occur in high-quality sites, unless it is determined that more intensive methods (e.g., chemical removal) are absolutely necessary to eliminate exotic or pest plants. Manual removal is also the preferred method in intermediate quality sites.

Mechanical removal (e.g., using bulldozers, specialized logging equipment) should be avoided in natural areas because it causes severe disturbance to soils and non-target vegetation. Mechanical removal should be used only when an area is being cleared for new land-use. Mechanical removal also requires follow-up treatment, as exotics will be quick to reinvade (Langeland 1990). Mechanical removal should only be allowed in moderately low quality and lowest quality sites. In these cases, the least disturbance methods should be used, and wetland protection and erosion measures should be taken.

Herbicides have been used successfully to remove woody exotics. Avoid herbicides within or immediately adjacent to TES or any permanent or seasonal wetlands. Herbicides can affect water quality and present a direct threat to rare species (Russo et al. 1993, USFWS 1983). Because of the risk of drift effects on TES and other non-target species, herbicides should not be used to control exotics in high-quality sites, and should not be used in intermediate-quality sites unless control is needed over large areas and manual removal is not feasible. However, because some exotics (e.g., Cogon grass) pose a serious threat to natural communities and are very difficult to remove manually, herbicides can be used.

If herbicides must be applied, the method and timing should be selected to minimize effects on non-target vegetation and the environment. The herbicide applicator must be well informed of the chemical properties of the herbicide, and under what circumstances it should be applied. Environmental precautions are stated on the herbicide label. In general, these guidelines should be followed:

The applicator also should be aware of potential weather conditions and should schedule applications accordingly (Langeland 1990):

Management Regarding Cogon grass. Specific recommendations for management of Cogon grass were provided by Coile and Shilling (1993):

Managers are advised to refer to the control measures in the IFAS publication "Cogon grass (Imparata cylindrica (L.) Beauv. Biology, Ecology and Control in Florida" by Colvin et al. 1994. Glyphosate (Accord or Roundup)** or imazapyr (Arsenal) are probably the best herbicides to control Cogon grass where they can be applied. Several treatments are necessary for effective control. The dead leaves of Cogon grass remain upright and do not decay easily, and these prevent herbicides from being effectively absorbed. For effective control, herbicide should be applied to living, green leaves, which will allow transport to rhizomes. Late fall is the best time to apply herbicides because plants are sending carbohy drates to roots and rhizomes for storage, and at this time the herbicide will also be translocated to rhizomes. Killing of rhizomes is necessary to control Cogon grass. It is essential to apply a herbicide after cultivation or burning.

Fertilization

Impacts

Because Florida scrub appears to be structured by nutrient stress (Myers 1990), the use of chemical fertilizers in and near Florida scrub may have drastic effects on the community. Native scrub vegetation does not need supplemental fertilizers to establish themselves and compete in scrub soils. Weedy species can out compete scrub vegetation following nutrient enrichment. These include the exotic species referred to in the previous section and species native to more fertile soils in Florida.

Management Recommendations

Use of fertilizers is not recommended and should be avoided in high-quality and intermediate-quality sites that support native scrub vegetation. In sites that are not restorable, fertilizers should be used with care, to assure that they will not enter wetlands.

* *Using glyphosate to control Cogon grass will also eliminate wiregrass and associated species (in Duever 1989). Since glyphosate and imazapyr are both broad spectrum herbicides, they will kill all or most plants that they come into contact with. Thus, care must be taken to avoid damage to nontarget, desirable vegetation (Langeland 1990).

Previous PageTable Of ContentsNext Page