USGS - science for a changing world

South Florida Information Access (SOFIA)


publications > paper > analysis and simulation of propagule dispersal and salinity intrusion from storm surge on the movement of a marsh-mangrove ecotone in South Florida > results

Analysis and Simulation of Propagule Dispersal and Salinity Intrusion from Storm Surge on the Movement of a Marsh-Mangrove Ecotone in South Florida

Results

> Results

Analysis of 10-Year Time Series of Empirical Water Budget Data and Salinity

Precipitation data from 2000 to 2010 showed a seasonal pattern (Table 1), with high rainfall occurring in the wet season (May-November). Rainfall amounts on 24 October 2005, the date Hurricane Wilma made landfall, reach a monthly average of time period we investigated. Potential evapotranspiration for the study site was estimated by USGS and showed a seasonal pattern with a peak occurring between April and August (Table 1), corresponding to the solar cycle. Groundwater level data for SH4 (mangrove) and SH5 (marsh) showed a seasonal pattern, with highest levels around September and lowest around March of each year (Table 1). Groundwater level increased significantly on the day Wilma made landfall and the following few days. Groundwater level returned to its usual trajectory of seasonal variation soon after the storm surge.

The groundwater salinity patterns of SH4 and SH5 were different (Table 1). Groundwater salinity at SH4 increased during the dry season and declined when the relatively fresher Harney River high tides (diluted river salinities from rainfall and freshwater inflows) combined with precipitation to wash out salt embedded within the limestone and peat during wet season. However, groundwater salinity at SH5 was relatively stable due to its distance away from the riverbank (~300 m). At SH5, groundwater salinity increased by a slight amount after Hurricane Wilma, whereas, groundwater salinity decreased at SH4. Surface water salinity at both sites doubled after Hurricane Wilma due to the storm surge that pushed water upstream from the Gulf of Mexico but was still lower than dry season salinity. We did not find any long-lasting salinity increase followed by the storm surge for either SH4 or SH5.

Table 1 Climate data from 2000 to 2010 at the study site, ET  is evapotranspiration, GW  is groundwater, SW  is surface water
Month Rainfalla
(mm)
Potential ETa
(mm)
SH4 GW 1eve1
(cm)
SH5 GW 1evel
(cm)
SH4 GW
salinity
SH5 GW
salinity
SH4 SW
salinity
SH5 SW
salinity
January 22.9±6.8 65.0±1.4 -6.4±2.8 -7.8±20 14.9±1.7 9.5±0.5 13.3±20 8.6±0.8
February 29.6±7.2 80.8±1.8 -13.3±35 9.6±3.0 16.3±1.5 9.7±0.2 16.0±2.0 9.4±0.9
March 33.8±12.4 126.1±2.0 -10.7±2.1 -8.5±1.6 20.0±1.0 9.7±0.2 21.2±1.5 10.0±1.3
April 56.2±17.8 151.4±2.4 -63±24 -4.5±2.0 21.7±1.1 9.3±0.1 24.1±2.0 10.7±1.2
May 73.9±19.3 168.7±2.8 -0.2±3.1 -1.4±1.8 25.6±1.3 9.2±0.1 25.9±2.4 12.4±1.6
June 207.4±24.6 152.7±4.7 4.2±1.7 7.7±1.8 20.3±2.6 9.2±0.2 18.0±2.8 11.2±1.5
July 179.5±28.7 157.5±2.6 7.5±0.7 9.9±1.1 16.0±1.9 8.8±0.3 8.4±1.6 7.4±1.6
August 222.6±43.4 147.3±2.9 10.0±0.5 12.8±1.2 12.1±1.5 9.0±0.3 6.7±1.0 5.5±1.3
September 204.5±29.6 125.9±2.4 13.3±0.5 17.4±1.3 11.7±1.4 9.0±0.3 8.3±1.1 5.7±1.1
October 64.8±15.2 109.8±2.4 11.4±0.8 15.3±0.9 10.9±1.3 8.9±0.5 83±1.1 6.1±0.9
November 36.8±8.1 75.7±0.9 6.9±1.0 12.4±3.4 11.1±1.5 9.5±0.4 9.1±1.4 7.6±1.2
December 27.4±6.1 58.1±1.3 -4.6±2.3 -0.8±1.8 13.2±1.6 9.6±0.6 10.1±1.1 8.4±1.0
Mean 96.6 118.3 1.0 3.6 16.1 9.3 14.2 8.6
Hurricane Wilma 50.0 51.5 47.2 50.3 5.2 10.2 16.6 12.2
All the data are monthly average±standard error, except on the date Hurricane Wilma made a landfall on 24 October 2010, which is a 1-day amount. Water level or stage data are in NAVD 88 datum
a Monthly cumulative amounts

Simulations of Soil Pore Salinity

We compare simulations of soil porewater salinity to surface water, due to lack of field data of soil porewater salinity during that time period. Soil porewater salinity is close to surface water salinity during wet season but may differ during dry season with evaporation of surface water. Simulations of the soil porewater salinity captured the seasonal signal pattern of surface water from SH4, which is dominated mostly by mangroves (Fig. 2). While model freshwater marsh-dominated cells show a dampened seasonal pattern similar to mangrove sites, the model does not fit SH5 surface water data very well (Fig. 2). Soil porewater salinity from mangrove-dominated cells increased dramatically in the model during the late dry season due to strong evapotranspiration and little precipitation. Overall salinities of the freshwater marsh sites were lower than mangrove sites, especially during dry season. This occurs because evapotranspiration in the interior freshwater marsh slows down if soil porewater salinity starts to increase, thus dampening the infiltration of underlying saline water.

plot showing simulation of soil porewater salinities averaging mangrove sites and marsh sites, respectively, compared with monthly average with standard error of surface water salinities from Shark 4 and Shark 5, respectively plot showing spatial distributions of individual mangroves and freshwater marsh output from simulation model
Fig. 2 (left) Simulation of soil porewater salinities averaging mangrove sites and marsh sites, respectively, compared with monthly average with standard error of surface water salinities from SH4 and SH5, respectively [larger image] Fig. 3 (right) Spatial distributions of individual mangroves (dark circles) and freshwater marsh (gray cells) output from simulation model. Area of circles represents canopy area of individual mangrove tree [larger image]

Simulation of Vegetation Dynamics

The 10-year average hydrology data were first used in the simulation model to determine whether the sharp vegetation ecotone observed at SH5 would be produced. The vegetation distribution that is produced shows a clear boundary between mangroves and freshwater marsh, resembling field observations of the coastal Everglades (Fig. 3). Then we simulated the effect of the hurricane. As shown in Table 1, Hurricane Wilma did not cause significant changes in groundwater salinity. The water level increased for several days but then decreased again. Our simulations, which included pulse increases of water level and surface water salinity, without assuming any direct mortality to the original vegetation or input of mangrove seedlings due to the storm surge, did not show any long-term effects on freshwater marsh dynamics from Hurricane Wilma stemming from salinity intrusion. In this simulation, it was assumed that no wind damage occurred from the hurricane, which could have contributed to a regime shift. It has been reported that direct hurricane wind damage to the mangrove forest can be 30-80%, depending on the intensity of hurricane and the distance from the eye of the hurricane (Armentano et al. 1995; Harcombe et al. 2009; Milbrandt et al. 2006). We repeated the simulation with assumptions of non-zero storm damage. Figure 4 shows the recovery of the basal area of mangroves projected by the model following 30%, 50%, and 80% mortality, respectively, due to hurricane damage. The mangrove forest recovered at different rates, without any expansion of freshwater marsh into the mangrove zone.

plot showing model-projected basal areas of mangroves following 30 %, 50 %, and 80 % destruction by hurricane, at year 80
Fig. 4 Model-projected basal areas of mangroves following 30%, 50%, and 80% destruction by hurricane, respectively, at year 80 [larger image]

Effects of Salinity Intrusion and Mangrove Dispersal

As mentioned in the "Methods," because the climatic data from 2000 to 2010 produced no situations in which a regime shift was likely, we developed several artificial scenarios for simulation. We combined three levels of salinity intrusion duration and three levels of mangrove propagule density passively dispersed to freshwater marsh sites to further analyze storm surge effects on ecotone dynamics. Results show that, in the case of salinity intrusion being rapidly washed away after hurricane, the ecotone between mangroves and freshwater marsh is, as might be expected, relatively stable (Fig. 5 left panels). As salinity intrusion duration increased, however, the model predicted more marsh habitats changing to mangrove permanently. The results are consistent with Jiang et al. (2012b), in which a mathematical model showed that only a relatively long "press" disturbance of salinity for a long period of time might cause regime shift. In these cases, passive transport of mangrove propagules by a storm surge could facilitate a vegetation regime shift triggered by salinity intrusion (Fig. 5, middle and right panels).

plots showing spatial distribution between freshwater marsh and individual mangrove trees under nine different scenarios
Fig. 5 Spatial distribution between freshwater marsh (gray cells) and individual mangrove trees (black circles) under nine different scenarios, which are combinations of three levels of mangrove seedlings transported into the marsh (none transported, a moderate number transported with 1,000 propagules/ha, and a large number transported with 2,000 propagules/ha) and three levels of salinity intrusion duration (no saltwater retention, 1-year salinity retention, and 2- year salinity retention) [larger image]

< Methods | Discussion >


Accessibility FOIA Privacy Policies and Notices

Take Pride in America logo USA.gov logo U.S. Department of the Interior | U.S. Geological Survey
This page is: http://sofia.usgs.gov/publications/papers/propdisp_salint/results.html
Comments and suggestions? Contact: Heather Henkel - Webmaster or (727) 502-8028
Last updated: 28 May, 2014 @ 10:15 AM (KP)