Conclusion

Lightning-initiated canopy gaps are an important but relatively understudied disturbance in mangrove forests around the world and are particularly important in the mangrove forests of Everglades National Park. I found that the environmental conditions within the gaps differed from the surrounding forest. The expanded gap size averaged 289 ± 20 m² (± SE). As gaps filled with saplings, light transmittance decreased exponentially. Overall, gaps had greater fine woody debris but less coarse woody debris than the surrounding forest. Soil torsion and soil compaction were lower in the gaps than the forest. The abundance of fiddler crab burrows decreased with distance upstream, additionally large and medium burrow abundance increased linearly with total seedling abundance. Newly formed lightning gaps had greater dead root biomass compared to the intact forest. I did not find a systematic reduction in the soil cohesiveness as new gaps aged. Soil surface elevation declined between 8.5 mm to 60.9 mm in newly formed lightning gaps; this loss was due to superficial erosion (8.5 mm) and subsidence (60.9 mm). Lightning apparently kills many of the shallow surface roots and leads to a decline in the soil surface elevation in new gaps. Subsidence occurring below the shallow soil zone generated the greatest overall soil elevation loss. Recovering (Growing) gaps had lower live root biomass but similar soil surface elevation patterns as the intact forest. This study suggests that if Rhizophora mangle can exploit the increased flooding from soil surface elevation loss in new lightning strike gaps, then it would have to occur in the first 7 to 10 years. After this time, the soil surface elevation patterns (recovering gap) reflect the intact forest eventually removing this opportunity. I was unable to find a relationship between the change in soil elevation and survivorship, mortality rate, or recruitment in the new gaps. However, because of the small sample size (n = 3 new gaps) and the short time period of the study, this is not a unexpected result.

I found that the saplings and seedlings in new gaps survived the lightning strike, and this is the first documented evidence that in a mangrove forest a large amount of the non-canopy trees present before the lightning strike survive. This evidence comes in the form of a distinct two-cohort regeneration apparent in the seedling and sapling height distribution. The first cohort consists of the propagules and seedlings present at the site pre-strike and individuals recruiting very soon after the canopy is removed. The second group recruits into the site some number of years post strike. The high densities of Rhizophora mangle as seedlings and saplings in the recruiting and growing gaps stages imply that lightning strike disturbances in these mangroves favors their recruitment and does not favor Avicennia germinans and Laguncularia racemosa. However, average A. germinans seedling height was found to increase in later gap stages, suggesting an increase in the transition probability from seedling to sapling stage perhaps related to gap successional development.

Enumerating survival, recruitment, and growth across life stages by species is of critical importance in understanding and predicting changes in forest structure, composition, and development especially in mangrove communities. This dissertation is the first report of recruitment/mortality rates for multiple mangrove species across life stages in gaps and closed canopy forest. Survival, recruitment, and growth varied across three successional stages of mangrove forest (newly initiated lightning gaps, closing gaps, and intact forest). The life stage parameters for the three dominate life phases (seedlings, saplings, and adult) of the three dominant mangroves (Avicennia germinans, Laguncularia racemosa, Rhizophora mangle) differed.

The seedling and sapling recruitment rates of A. germinans were 1.5 times greater than mortality in new lighting-initiated canopy gaps indicating an expanding population. New gaps also had 2.6 to 10.6 times greater rate of seedling mortality for R. mangle and L. racemosa compared to the recruitment rate, indicating decreases in these populations. Seedling stem elongation was greatest in the new gaps. Taken together at least, seedling recruitment rate during my study was twice as high in new gaps, as in the other forest stages. Presumably this recruitment rate will continue to increase as the conditions within the gaps favor propagules establishment. Additionally, I conclude that new light gaps may favor A. germinans seedling recruitment in this initial stage of gap succession. Finally, future studies of life stage population parameters (survival, recruitment, and growth) should include recruiting gaps. From the extremely high densities of R. mangle seedlings and saplings in the study it must concluded that R. mangle recruitment and survivorship increase greatly some time after the new gap stage of succession.

At the growing gap stage of development seedling mortality rate of R. mangle was 10 times greater and sapling mortality was 13 times greater than recruitment. The recruitment of R. mangle adults was 4 times greater than mortality. The gaps have developed to a phase in which there was reduced stem elongation, sapling and adult growth, and few individuals able to recruit into the adult life stage. Sapling populations are high (~ 1 sapling m²), and seedling populations are low (0.6 seedling m²). R. mangle dominates seedling and sapling stages of the growing gaps and eventually are making the transition to the adult life stage. The end results indicate that at the growing gap stage of succession of the lightning gaps R. mangle stems were being favored as adult trees.

In the intact forest, A. germinans seedlings and sapling recruitment was 3 times greater than the mortality rate. Additionally, L. racemosa and R. mangle seedling mortality was 2 times greater than the recruitment and sapling mortality was 28 times greater than recruitment. In general, growth within the forest was low across all life stages compared to the new gaps. These population parameter results suggest that A. germinans becomes a co-dominant to dominant in closed canopy mangrove forest of South Florida (Craighead 1971) by having higher recruitment than mortality for the seedling and sapling stages. Given enough time the population of A. germianans seedlings and saplings will continue to expand whereas the populations of L. racemosa and R. mangle were decreasing.

In conclusion, this study suggest that lightning strike disturbance in these mangroves favors R. mangle recruitment based on densities and does not favor A. germinans and L. racemosa. Overall, vegetative dynamics in lightning initiated canopy gaps indicate that this disturbance may maintain South Florida mangroves in a cyclical or arrested successional state of development. My results provide population parameters needed to understand and predict recruitment and survivorship for each of the three dominant species (A. germinans, L. racemosa, and R. mangle) during the gap-phase dynamics of the mangrove forest and within the intact closed canopy forest. Additionally, I determined growth estimates enabling better understanding of intact forest and development within the stages of gap-phase dynamics. The results of this study provide new insights into the regeneration process of lightning disturbed systems and into other mangroves systems experiencing gap dynamics mechanisms.

Potential impact of Everglades Restoration on lightning gap dynamics.

I was able to show that site hydrology (both as groundwater head pressure and river stage) does have a direct impact on the soil elevation both in the intact forest (both short-term (Chapter V) and long-term (Chapter IV)) and in disturbed sites. However, I was unable to make a strong link to seedling recruitment or survival to changes in soil elevation. I found that a number of the environmental variables that varied with river location: amount of coarse woody debris, soil bulk density, soil torsion and soil compaction and fiddler crab (Uca thayeri) burrow densities. Additionally, river location was significant for propagule and seedling densities in R. mangle, propagules of L. racemosa, and biomass of L. racemosa and R. mangle. However, I believe that the long-term gap successional trajectory will be minimally impacted by Everglades hydrological restoration. The gap dynamics appear to have a consistent pattern regardless of river position. R. mangle dominates seedlings in recruiting gaps and saplings in the growing gaps at all of the sites along the river. Additionally, R. mangle seedlings, saplings, and adults in this forest are highly resilient to hydrological conditions (Krauss 2004). The unknown variable is the influence of hydrological manipulation on A. germinans and L. racemosa recruitment. However, since it appears that R. mangle is favored by lightning gap disturbance on the whole river, the impact of manipulated hydrology will be minimal on the gap successional process. The rate at which the gap transitions between forest successional stages may be affected by changes in hydrological discharge, however, and determination of changes in transition rates will have to be the subject of future studies. Sea level rise is also impacting this forest. In the short-term ecological view, the Everglades restoration will clearly influence the hydrology the mangrove forests even affecting the soil elevation. However, sea level rise will continue to modify the mangrove community by driving transgression into the freshwater marshes beyond the ability of management to mitigate the long-term impacts.