Home Archived October 29, 2018

South Florida Information Access (SOFIA)

publications > paper > groundwater control of mangrove surface elevation: shrink and swell varies with soil depth > discussion


Materials & Methods
Literature Cited
Figures, Tables & Equations
PDF Version
The soil surface elevation changed substantially during the year; the deep-RSETs recorded the greatest average elevation (15.14 mm) at the end of the wet season (November 9, 2002). The patterns of cumulative change in soil surface elevation were very similar for both the deep-RSET and original-SET, but the pattern of the shallow-RSET was distinctly different (Fig. 3). The overall annual accretion rate of 6.6 mm yr-1 was similar to the 4.4-7.8 mm yr-1 reported in another mangrove study in southwest Florida (Cahoon and Lynch 1997). The influence of accretion and erosion on the change in soil elevation was minimal over the duration of this study, as it was not a significant factor in any of the regression models. Elevation for all three SETs had changed substantially before accretion at the site was even measurable, indicating the importance of subsurface processes. In addition to accretion and soil swelling, shallow and deep subsidence have been reported to be significant factors for the interpretation of soil elevation change (Cahoon et al. 1995). Here we were able to account for the opposing influences of subsidence and soil swelling by sampling the entire soil profile while including the processes of deposition and erosion in multiple regression models.


The entire mangrove peat-dominated soil profile was strongly influenced by groundwater. The rate of change in groundwater head pressure had a strong positive linear relationship to the rate of change in soil surface elevation for the deep-RSET (Adjusted R2 = 0.90), suggesting that the entire soil profile is swelling in response to hydrological recharge. In this area, change in the daily groundwater piezometric pressure reflects freshwater recharging of the estuary and monthly tidal influences. Other mangrove SET researchers (Cahoon and Lynch 1997; Smith and Cahoon 2003) have reported seasonal response to soil elevation, but a direct relation to forcing by a hydrological parameter has not been previously shown. Because this particular peat has relatively low superficial hydrological conductivity and is typically continuously saturated, peat swelling may not be the only mechanism explaining this relationship. Nevertheless the tight coupling suggests this is the most likely mechanism driving changes in soil elevation.

Soil shrink and swell has been reported numerous times but almost exclusively in regards to soils with high clay compositions (Hillel 1971). As far as the authors are aware there are few reported shrink and swell observations in regards to wetland soils composed almost exclusively of peats driven by changes in groundwater head pressure. Those studies reported are confined to Sphagnum peatlands (Price and Schlotzhauer 1999) along with one reference to surface elevation changes in a salt marsh, but this was linked to semidiurnal surface tidal flooding (Nuttle et al. 1990). Our study indicates that changing groundwater head pressure was driving the monthly shrink and swell of the soil surface elevation in this peat matrix. Another study (Cahoon and Lynch 1997) suggested the importance of mangrove peat shrink and swell, in addition to growth, decomposition, and shallow subsidence as possible mechanisms for explaining annual elevation patterns. In our study, we were able to show that the peat matrix undergoes shrink and swell and that the majority of the expansion and contraction occurs in the bottom zone.


Soil elevation over the depth of the root zone had a moderate relationship with the DRC in Shark River daily stage (Adjusted R2 = 0.16). The first five sampling events recorded no deposition since marker horizons were not completely covered; yet we recorded substantial change in surface elevation influenced by the shallow soil zone suggesting belowground influences. It should be noted that the marker horizons showed progression towards complete coverage by having less of the marker horizon visible each of the five successive sampling events. We were able to remove the influence of deposition and erosion by determining the relationship between thickness of the shallow zone (0-35 cm) and river stage. As daily rate of change for the river stage increased, the thickness of the shallow active root zone decreased (R2 = 0.24, F1,34 = 10.57, p < 0.004). This analysis indicates that changing river stage has a stronger influence than previously noted for elevation change, but it is still only a moderate relationship. The lack of a strong hydrological link to the shallow soil profile is not wholly unexpected. Biological (root growth, crab burrow dynamics) processes rather than strictly hydrological influences dominate this shallow soil zone. Other possible explanations for the lack of a strong hydrological coupling are a shift in redox to more reducing conditions or a decline in root growth.

Erosion and deposition were not a great influence in explaining the change in surface elevation of the shallow-RSET over the short period of this study for the following reasons. The rate of deposition and erosion were not a significant parameter in the shallow-RSET model. The first five sampling events indicated substantial change in surface elevation influenced by the shallow soil zone when no deposition and erosion were measured. The model was rerun for only those periods with marker horizons measurements and no difference was found in the final model.


graph showing actual soil surface elevation of the original design surface elevation table versus calculated soil surface elevation (proportion of the deep-rod surface elevation table)
Fig. 6. Actual soil surface elevation of the original-SET (mm) versus calculated soil surface elevation (mm) (proportion of the deep-RSET). Dark solid line represents 1:1 ratio. n = 36. [larger image]
The response of the soil elevation change does not appear to be directly proportional to the depth of the soil profile encompassed by the SET device. The original-SET (0-4 m) followed the groundwater influence (R2 = 0.61), but not as strongly as the deep-RSET (0-6 m; R2 = 0.90). Compared to the deep-RSET, the original-SET encompassed 2 m less of the soil profile, which reduced the coupling between change in soil elevation and change in groundwater piezometric pressure (slope of the regression equation beta1 = 0.040 for the original-SET versus beta1 = 0.074 for the deep-RSET, Table 2).

We used the proportion of the soil profile sampled by the original-SET as compared to the deep-RSET to predict the average elevation of the original-SET based on the corresponding deep-RSET readings. Original-SET number one benchmark depth was 4.04 m and the deep-RSET number one benchmark depth was 5.47 m, resulting in a proportion of the entire soil profile sampled by original-SET number one of 0.74 (i.e., 4.04/5.47 m). If the relationship was linear with proportion of soil profile sampled then the actual values should fall near the calculated values along the one to one line (Fig. 6). The values predicted for original-SET based on this ratio were higher than the actual elevation values recorded (Fig. 6), suggesting that the deepest 2 m of peat not encompassed by the original-SET have a disproportionately larger influence on the absolute soil elevation.

To further corroborate the importance of the influence of the bottom zone on overall soil profile expansion and contraction, we determined the percent of variation explained by each component zone to overall soil column expansion and contraction. We determined that the largest constituent zones, the middle and bottom zones, drive the expansion and contraction of the entire profile. These two parts account for 94.2% of the soil profile and explain 85% of the variance in overall soil profile expansion and contraction. The bottom zone accounted for 63% of the variation in the absolute change in thickness but comprised only 31% of the profile. The middle zone accounted for only 22% of the variation but comprised 63% of the profile (Fig. 5, Table 3). These data suggest that the bottom zone has a greater influence on overall change in soil surface elevation than would be expected based on its relative proportion and that in this zone changing groundwater pressure would be the most influential.

Our results indicate that increases in groundwater flow should have a direct positive effect on absolute soil surface elevation for the entire soil profile by expanding the bottom soil zone. Since expansion and contraction affects the water storage potential of the peat matrix it is an important consideration for studies of water balance and nutrient fluxes (Nuttle et al. 1990). The current hydrological restoration of the Everglades and increases in sea level will directly affect this mangrove forest. Any modification to freshwater flows via the Everglades Restoration will affect the elevation of the mangrove forest by expansion and shrinkage. In order to determine how other processes (bioturbation, organic production, decomposition, disturbance, and subsidence) will affect long-term change in soil surface elevation, researchers must account for this shrink and swell signal and remove it from the analysis. The influence of these hydrological processes must be taken into account in the context of monitoring the effects of hydrological restoration or sea level rise. Understanding the factors influencing the change in soil elevation as it relates to different parts of the soil profile will be critical when trying to predict long-term mangrove sustainability in an increasing sea level environment.

< Results | Acknowledgments >

| Disclaimer | Privacy Statement | Accessibility |

U.S. Department of the Interior, U.S. Geological Survey
This page is: http://sofia.usgs.gov/publications/papers/mang_surf_elev/discuss.html
Comments and suggestions? Contact: Heather Henkel - Webmaster
Last updated: 04 September, 2013 @ 02:04 PM(KP)