Ecological Studies of the Sunken Forest,
Fire Island National Seashore, New York

NPS Scientific Monograph No. 7
NPS Logo


In the Sunken Forest there is a close association between the cycling of nutrients within the ecosystem and the cycling of nutrients into and out of the system. In the maritime forest ecosystem with a salt-laden atmosphere, the inputs into the ecosystem in the form of aerosols impacted in vegetative surfaces are transferred to the ground in the intrasystem leaching circulation.

Of the three types of nutrient inputs into terrestrial ecosystems in the model developed by Bormann and Likens (1967), meteorologic inputs appear to be the most important in the Sunken Forest ecosystem. The geologic inputs by the diffusion of cations into ground-water systems must be minimal in light of the barrier-island hydrology previously discussed. Little geologic input in the form of sand and organic matter movement was noted during the period of this study. Very little sand was found in the collectors within the forest, the secondary dune system leeward of the Sunken Forest being generally stabilized by vegetation. Biologic inputs may be significant on the bayside margins of the Sunken Forest where sea-feeding birds nest; however, no bird nests of any kind were located in the ecosystem analysis plot.

The main output from the Sunken Forest ecosystem, although not measured in this study, is undoubtedly the transport of nutrients in the ground-water system. The discharge of the barrier-island ground water into the sea completes the cycle started with the generation of salt-spray aerosols which serve as condensation nuclei for precipitation or which were impacted on vegetative surfaces. The biologic and meteorologic outputs from the Sunken Forest appear to be relatively small.


The meteorologic inputs into the Sunken Forest consist of both nutrients carried in the precipitation and salt-spray aerosols carried by the wind and impacted on the vegetation. Precipitation collections actually consist of three components: (l) rain, which is precipitation that falls as liquid water (snow is included in this component); (2) dry fallout, which is water-soluble matter that settles on the collectors between storms; and (3) bulk precipitation, which is a mixture of rain and dry fallout (Whitehead and Feth 1964).

The rain component of the precipitation was sampled in a Wong automatic rain collector (model Mark V). The Wong collector, which was installed on the secondary dune at Sailors Haven (0.5 km east of the Sunken Forest), automatically opened only during periods of active precipitation. Bulk precipitation was sampled in collectors identical in design to those used in the throughfall analysis. Two of these open collectors were used, one at Sailors Haven and the other on the secondary dune crest near the Sunken Forest plot. Dry fallout was estimated by subtracting the amounts of cations in the rain collection from the bulk precipitation.

In each of the precipitation sampling periods, which were concurrent with stemflow and throughfall collection periods, clean polyethylene receptacles were installed for the rain and bulk precipitation collectors. Bulk precipitation required filtration to remove sand and other material blown into the collectors. Sample preparation of this material followed those outlined for litter analysis.

The cation composition of impacted salt-spray aerosols was determined by washing off twigs exposed to the wind at the top of the canopy. The aerosols were sampled twice during periods immediately following rain storms. At the start of the period, a twig was coated with an acrylic paint and served as an inert aerosol impaction surface. After a period in which there was no precipitation, the twigs in the canopy were removed and rinsed in 100 ml of distilled water for 30 seconds. This leachate was retained for cation analysis. Direct measurement of the magnitude of the salt spray was beyond the scope of this study due to the complexity of the aerosol deposition patterns (Boyce 1954).

Ground water was collected at the end of each collection period from a well 3 m east of the Sunken Forest plot. This polyethylene-lined well was sunk to a depth of 30 cm and contained ground water at all times during the study. Samples were withdrawn from the well using a polyethylene syringe. Ground water, precipitation, and aerosol leachate samples were all analyzed for cations following the procedures outlined for the throughfall analysis.

Results and Discussion

The amounts of cations in the precipitation, which are determined from the cation concentrations and the amounts of precipitation, vary considerably between collection periods (Fig. 45). The input of cations in the bulk precipitation and rain appears to be related to the direction of the wind during periods of active precipitation (Fig. 46). During the winter when prevailing winds offshore from the north and east are prevalent, the precipitation inputs of cations are generally lower than in the remainder of the year when prevailing winds are onshore from the south and west. September's dip in precipitation cation inputs is also coincident with storm winds from the north and east. These patterns are most apparent in the bulk precipitation and dry fallout components, further supporting the suggestion that seasonal patterns of gross throughfall transfer within the ecosystem are in part due to inputs of salt-spray aerosols.

Fig. 45. Precipitation cation inputs (mg/m2/day). (click on image for an enlargement in a new window)

Fig. 46. Wind direction during rain. (click on image for an enlargement in a new window)

The amount of cations in the bulk precipitation is generally twice as great as in the rain component (Table 25). Undoubtedly, most of the cation enrichment in the bulk precipitation is from the dry fallout and impaction of salt-spray aerosols on the open collector funnels, although small amounts (0-23 mg/m2/day) of sand and organic debris were found in the samples (Appendix IV). The smooth surfaces of the bulk precipitation collectors are not as effective in trapping salt-spray aerosols as the complex and finely divided vegetative surfaces (Boyce 1954). Therefore, the bulk precipitation must be considered to be a minimum estimate of meteorologic cation inputs.

A maximum estimate of meteorologic inputs can be made by adding the inputs in rain and the estimated total impacted aerosol input, which includes both dry fallout and aerosols impacted on vegetative surfaces. The estimate of salt-spray impaction based on ion ratios to excess sodium in throughfall and stemflow relative to rain in the open leads to an overestimation since some of the sodium in the precipitation under the canopy has been leached out of the plant tissues. The maximum aerosol impaction inputs for the Sunken Forest were estimated by multiplying the excess sodium in the leaching transfer (Fig. 10) by the K/Na, Ca/Na, and Mg/Na ratios of aerosols washed off the acrylic-coated twigs (Table 26).

Table 25. Meteorologic inputs into the Sunken Forest ecosystem g/m2/yr.

K Na Ca Mg

(Wong collector)
Dry fallout 0.405.410.500.47
Bulk precipitation
(open collector)
Aerosol impaction
(maximum estimate)
Minimum input
(rain and dry fallout)
Maximum input
(rain and aerosol impaction)
0.8418.59 1.102.66
Average input
(minimum + maximum/2)
0.7314.15 0.981.91

Impaction inputs are highly correlated with the speed of onshore winds which prevail during the growing season. The relationship of kilograms of sodium impaction per hectare (Y) to thousands of kilometers of wind blowing onshore (between 79° and 238°) at speeds of <7 m/sec. (X1) and >7 m/sec. (X2) is given by the equation:

Y = 2.49 X1 + 3.57 X2 - 0.264

The multiple correlation coefficient is 0.84 and the regression has an F value of 39.4 with 2 and 34 degrees of freedom which is significant at P = <0.l%.

Averaging the maximum and minimum estimates, the annual meteorologic inputs in the Sunken Forest ecosystem are: 14.15 g sodium/m2; 1.96 g magnesium/m2; 0.98 g calcium/m2; and 0.73 g potassium/m2 (Table 25). The sodium input far exceeds the amount of sodium in the primary production. While the magnesium input is about equal to that in the primary production, the potassium and calcium inputs are only 1/7 and 1/5 of the amounts contained in the production. Therefore most of the potassium and calcium presently taken up by the vegetation must have been accumulated by the ecosystem in previous years and repeatedly circulated in the intrasystem cycles.

Table 26. Cation concentrationsa and ratios in the nutrient cycle of the Sunken Forest ecosystem.

K/Na Ca/Na Mg/Na

Sea water 387 10,769 408 1297 0.0363 0.0381 0.119
Rain 0.2 3.7 0.3 0.6 0.0541 0.0811 0.162
Bulk precipitation 0.5 8.3 0.7 1.0 0.0602 0.0843 0.121
0.0435 0.0519 0.139
Average meteorologic input
0.0505 0.0693 0.139
Organic compartment
1.70 2.35 0.877
Primary production
2.76 2.66 1.01
Gross leaching 3.7 17.1 2.6 2.6 0.217 0.152 0.152
  Gross throughfall 3.6 15.8 2.5 2.5 0.228 0.158 0.158
  Gross stemflow 4.3 23.5 3.3 3.0 0.183 0.140 0.128
Available nutrient compartment
1.01 4.27 1.45
Groundwater 1.4 27.2 3.9 4.1 0.0515 0.143 0.151

aWeighted averages

The differences in the circulation patterns of cations within the Sunken Forest are evident from the changes in cation concentrations and ratios that occur between the ecosystem compartments (Table 26). Although the ocean is undoubtedly the ultimate source of most cations entering the Sunken Forest, and order of cation abundance in the meteorologic input is the same as in sea water, the ratios of potassium, calcium, and magnesium to sodium in all compartments of the nutrient cycle are greater than in sea water. Compared to sea water, salt-spray aerosols leaving the ocean surface are enriched in potassium calcium, and magnesium relative to sodium and chlorine (Sugawara 1965). The alteration of ionic ratios in salt-spray formation is thought to be largely the result of physical and chemical processes at the sea surface and in the atmosphere (Sugawara et al. 1949; Koyaina and Sugawara 1953; Bloch et al. 1966). However, the possibility of bursting bubbles ejecting nutrient-rich material concentrated at the sea surface must also be taken into consideration (Goering and Menzel 1965; Garrett 1964; Blanchard 1964).

Compared to sea water, the average meteorological input for the Sunken Forest has nearly a twofold enrichment in potassium and calcium relative to sodium. The selective biological accumulation of potassium, calcium, and magnesium in the primary production and organic compartment increases the ratios of these ions to sodium far above those in the input. The concentrations of all cations in the gross leaching are greater in the bulk precipitation, with potassium and calcium having the greatest enrichment. The magnesium:sodium ratios in the leaching transfers are similar to those in the inputs since most of both the sodium and magnesium are probably aerosols washed off the vegetative surfaces.

The cation ratios in the available nutrient compartment are largely determined by strengths of cation adsorption on soil colloids. Since sodium is weakly held on colloids, the ratios of all cations to sodium are greater than 1.0. In the ground water the ratios of various cations to sodium are the lowest within the ecosystem, approaching those of the average meteorological input.

The ground-water system represents an integration of the biogeochemical relations and the major output pathway for the Sunken Forest. The cation concentrations in the ground water exhibit a greater stability between collection periods than do those in the precipitation (Figs. 45, 47). The buffering capacity exerted by the Sunken Forest on the ground-water concentrations is similar to those exhibited by other systems (Bormann et al. 1969). The concentrations of sodium, calcium, and magnesium in the ground water are higher than in the gross leaching while those of potassium are lower, suggesting a differential uptake of potassium by the vegetation (Table 26). The cation concentrations of the ground water sampled approximate those of the geologic output, but since the depth of the sample well was barely below the rooting depth of the dominant vegetation, ground-water samples from deeper depths might show a somewhat different chemistry.

Fig. 47. Ground-water cation concentrations (mg/l).

Although outputs from the Sunken Forest ecosystem were not measured, an indication of the retention of various cations in the system can be gained from their relative residence times. The relative residence time indicates the comparative speed at which an element is circulating through the ecosystem; the lower the residence time, the faster the element is circulated out of the system. The relative residence time is calculated by dividing the amounts of relatively available cations in the ecosystem (organic and available nutrient compartments) by the average meteorologic inputs, under the assumption that the system is approaching a steady state (Tables 22, 25).

The relative residence times (in years) are: sodium 2.0; magnesium 14.8; potassium 60.5; and calcium 81.9. Sodium, a nonessential plant nutrient, is transferred through the system at a much greater rate than the other cations. The relative retention of magnesium in the system is greater than sodium but not as great as potassium and calcium which are held tightly in the living biomass and available nutrient compartments.

<<< Previous <<< Contents >>> Next >>>

Last Updated: 21-Oct-2005