Restoration Potential: Recovery is directly linked with the ability to preserve existing habitat and restore degraded habitat. Given the drastic decline in the extent of longleaf pine-dominated communities (Ware et al. 1993), elevation of flatwoods salamander populations above present levels is unlikely. Restoration of degraded mesic, seasonally inundated longleaf pine flatwoods and savannas has not been attempted, and may only be feasible in cases where soil disturbance is minimal. The effectiveness of reintroduction into areas where extirpated is unknown.
Preserve Selection and Design Considerations: High quality occurrences include several wetlands within a matrix of pine flatwoods and savanna. Based on the maximum distance adults are known to travel between reproductive and nonreproductive habitat (1.7 km), each breeding site should be surrounded by at least 10 sq km of terrestrial habitat. Longterm perpetuation of a viable population of flatwoods salamanders will presumably require protection of a larger area of terrestrial habitat encompassing a suite of alternative breeding sites (Travis 1994). A suite of wetlands guards against extirpation at any one breeding site, since animals can immigrate from nearby wetlands. The minimum viable population size needed to sustain a population longterm is not known. Preliminary drift fence data at Eglin Air Force Base, Florida, suggests that breeding population sizes are low relative to other Ambystoma (Palis, unpubl. data). However, this may be a site specific observation as larger breeding migrations have been observed elsewhere in the range (R. Moulis, pers. comm.). Presently, there is no method of assessing an occurrence based on the number of animals captured at a drift fence or the number of larvae inhabiting a breeding site.
Management Requirements: Maintenance of intact mesic longleaf pine-wiregrass flatwoods and ephemeral wetlands by mimicking natural forces, such as lightning-season fire, is the most appropriate form of management. On sites where timber extraction is practiced, several precautions should be taken to limit the impact to flatwoods salamanders. Tree harvest should be restricted to dry periods to prevent soil compaction and rutting. Clearcutting should be replaced with selective timber harvest and natural regeneration enhanced by fire, particularly lightning-season fire. If off-site species such as slash pine have been planted, they should be removed and replaced with longleaf pine at densities found in nature. Mechanical preparation of the soil should be avoided. If a site supports mature, closed-canopy pine plantations, they should be thinned with as little disturbance to the soil and remaining groundcover as possible. The natural hydrology and fire regime of terrestrial and aquatic habitats should be restored on sites where altered.
The wetland/upland ecotone appears to be critical to successful flatwoods salamander reproduction. Some areas are in need of periodic burning to clear encroaching shrubby vegetation that shades out herbaceous ground cover (Palis and Jensen 1995). Maintenance of a graminaceous ecotone and breeding site will require burning in the lightning-season when wetlands are dry or nearly dry (Huffman and Blanchard 1990) . Bury et al. (1980) recommended that wiregrass not be burned in winter (destructive to wiregrass [used for egg attachment] and possibly to salamanders directly). Palis and Jensen (1995) stated that winter burns may be needed to avoid catastrophic fires when warm-season burning is initiated.
Mechanical disturbance of the wetland-upland ecotone should be avoided. The practice of "protecting" wetlands by encircling them with plow line should be abandoned. Where present, berms should be removed and drainage ditches filled.
Breeding ponds should not be dredged or stocked with fishes.
Management Research Needs: Development of a quantitative means of using surveys of larvae to indicate the size of the adult breeding population is needed.
Demographic data are needed to better understand the natural history and, in particular, factors that limit population size (e.g., egg, larval, and metamorph survivorship; competition with other species).
Longterm drift-fence studies are needed at several nearby sites to examine inter-pond salamander movement and to delineate the range of natural population fluctuations.
More information is needed on the extent of upland habitat required to support a population breeding in a particular pond. Radiotelemetry or radioactive tagging of adults could be used to address this need.
Effects on salamander populations of different forms of resource management and of anthropogenic habitat disturbance need to be examined (Palis and Jensen 1995).