DISTRIBUTIONS OF MERCURY AND PHOSPHOROUS IN EVERGLADES SOILS FROM WATER CONSERVATION AREA 3A, FLORIDA, U.S.A. CLEONE ARFSTROM1 , ANDREW W. MACFARLANE2 and RONALD D. JONES3 1 CH2M Hill Regional Office, 100 Inverness Terrace East, Eaglewood, CO 80112, U.S.A.; 2 Department of Geology, Florida International University, Miami, FL 33199, U.S.A.; 3 Southeast
Environmental Research Center and Department of Biological Sciences, Florida International University, Miami, FL 33199, U.S.A.
(Received 9 February 1998; accepted 1 September 1999)
Abstract. Soils in the southern half of Water Conservation Area 3A are mostly peats with some organic-rich marls. Mercury contents of 64 surface samples over a 500 km2 area average 28.7 ng cc−1 (209 ppb dry sediment), which is typical of organic-rich soils. High Hg contents in Everglades fish are therefore not caused by anomalously high soil Hg. Hg contents show no systematic lateral variation, consistent with deposition from well-mixed atmospheric sources rather than nearby point sources or runoff from canals. Cores from 9 sites contain more Hg and P at or near the surface than at 20–30 cm depth. Hg and P contents of individual cores correlate well and define separate background and anomalous populations. The subsurface distribution of P is determined largely by uptake by sawgrass and other plants. The correlation between P and Hg suggests that, although atmospheric Hg deposition has undoubtedly increased in recent decades, postdepositional mobilization of Hg may be important in Everglades soils. This finding, together with recent direct measurements of atmospheric Hg deposition, indicates that previous estimates of Hg deposition rates based on Everglades peat cores, which assumed that Hg is immobile in peat after deposition, have yielded large overestimates. Keywords: Everglades, mercury, mobility, peat, phosphorus, soil
1. Introduction The Everglades are the largest freshwater marsh in the contiguous U.S. (Gunderson and Loftus, 1993), and once covered south-central Florida from Lake Okeechobee to Florida Bay (Figure 1). Since the late 1800’s, draining for farming, housing and flood control has reduced the area of the Everglades by about 60% and altered the natural hydrology of the area (Myers and Ewel, 1991). The present Everglades is divided by levees into hydrologic compartments which include the Everglades Agricultural Area (EAA), Everglades National Park (ENP) and Water Conservation Areas (WCAs) 1, 2, and 3 (Figure 2). The largest water conservation area (WCA3; 2370 km2 ) is located in the central Everglades; about 80% of that area forms WCA-3A. The discovery of up to 3ppm Hg in Everglades largemouth bass (Hand and Friedeman, 1990; Ware et al., 1990) led state agencies to issue advisories against Water, Air, and Soil Pollution 121: 133–159, 2000. © 2000 Kluwer Academic Publishers. Printed in the Netherlands.
134
C. ARFSTROM ET AL.
Figure 1. Principal canals in south Florida, showing the position of study area in WCA-3A and in relation to other WCAS. Coarse stippled pattern indicates areas of high P inputs and dense cattail stands (modified from Davis and Ogden, 1994).
DISTRIBUTIONS OF MERCURY AND PHOSPHOROUS IN EVERGLADES SOILS
135
Figure 2. Canals, levees and control structures bounding WCA-3A and -3B, sample sites, and topographic contours of WCA-3. Bullseye symbols indicate cored sites. Elevations in feet.
the eating of fish. Since the modern Everglades are used only for recreation, flood control and water supply, high Hg levels were initially surprising. Jones (1992) hypothesized that high Hg levels may be related to increased P inputs from agricultural runoff which could enhance anaerobic soil conditions that promote mercury methylation. Methylmercury constitutes 99% of the mercury in organisms and is one of the most toxic forms of mercury. Farming in the EAA produces nutrientrich runoff which is transported by canals to the Everglades (Reddy et al., 1991).
136
C. ARFSTROM ET AL.
This runoff causes displacement of native flora, which are adapted to a low-nutrient environment, by cattails adjacent to canals (Figure 1). We sought to compare Hg and P in central Everglades peats with other organic soils, to look for areal patterns that could indicate contamination from local Hg sources such as canals or incinerators, to evaluate possible correlations between soil P and Hg, and to examine soil Hg and P variations with depth. Southern WCA3A was chosen because it is centrally located and probably best represents average conditions in the Everglades Trough, and high Hg levels in fish tissues are found there.
2. Background Soils in the study area began to accumulate about 5000 yr ago on the Pleistocene Miami Formation, which consists of bryozoan limestone in the western Everglades and oolitic limestone in the east (Scott, 1992). At that time, the climate in South Florida became more subtropical and rising sea levels reduced freshwater drainage from depressions on the peninsula (Gleason and Stone, 1994). Everglades Peat developed on topographically higher areas and is derived primarily from degraded sawgrass with parts of other plants; this soil is typically brown to black, with mineral content usually less than 10% of the dry mass. Loxahatchee Peat developed on topographically low areas covering about 2950 km2 , and is chiefly composed of remains of roots and rhizomes of the white water lily (Nymphae sp.). Loxahatchee Peat occurs in WCAs 1 and 2 and also in a long, narrow patch extending from western WCA-3A into the ENP. Carbonate muds are also present in the western margin of WCA-3A and in the Everglades National Park. Carbonate muds are nonlaminated, fine-grained low-magnesium calcitic muds derived from and underlain by limestone (Brown et al., 1991), and usually overlain by periphyton, a spongy algal mat of blue-green algae and diatoms mixed with mostly silt-sized calcite crystals.
3. Previous Studies Rood et al. (1995) surveyed soil Hg in cores from 45 Everglades sites with different hydroperiods, soil types and degrees of anthropogenic modification. They found the highest surface values in WCA-1 and parts of WCA-2, with some high values in southern WCA-3A. Lower values in parts of the ENP were attributed to the presence of calcitic mud rather than peat soils. Rood et al. (1995) dated the peats using 210 Pb and 137 Cs and calculated accumulation rates of 1.2 to 1.8 mm yr−1 (Delfino et al., 1993). Craft and Richardson (1993), using the same dating methods, found accumulation rates of 1.6–2.0 mm yr−1 in areas of reduced hydroperiods and 2.8-3.2 mm yr−1 in areas of extended hydroperiods or phosphorous enrichment.
DISTRIBUTIONS OF MERCURY AND PHOSPHOROUS IN EVERGLADES SOILS
137
Rood et al. (1995) found elevated Hg contents in the upper 10 cm of several cores, and interpreting Hg variations with depth as a simple record of deposition rate, calculated Hg deposition of 23–141 µg m−2 yr−1 in the Everglades since 1985, with an average of 53 µg m−2 yr−1 . Rood et al. (1995) calculated that the average Hg deposition rate was about 5 times higher after 1985 than ca. 1900. Volk et al. (1975) and Craft and Richardson (1993) measured phosphorous contents of 354–752 and 432–764 ppm dry sediment, respectively, in uncontaminated Everglades peats. Peats near canals associated with agricultural runoff contain 1200 to 1600 ppm P and show displacement of native flora by cattails (Figure 1; Reddy et al., 1991). Phosphorous in sewage and agricultural runoff is up to 95% soluble phosphate which is rapidly incorporated into algae and then into sediments (Horne and Goldman, 1994). Mean surface water P in WCA-2A was measured biweekly for 6 yr near inflow structures by Urban et al. (1993) and was found to decrease from nearly 0.2 mg L−1 to less than 0.05 mg L−1 within 6 km of the structures. Downstream from these structures, canal water is forced to flow overland unlike most other areas of the Everglades, and dissolved phosphorus is carried further into the interior of WCA-2A. Elsewhere, excess phosphorus is probably entirely taken up within 1 km of the canals (Urban et al., 1993).
4. Methods Surface soil samples were collected at 64 sites on a one-mile grid, covering about 500 km2 in the southern half of WCA-3A (Figure 2). Samples traverses were labeled A to J from west to east and 6 to 14 from north to south. The northern edge of the study area had no distinct break in topography. Surface samples were collected by airboat between September 25 and December 17, 1993. Triplicates were collected at each site for a total of 192 samples. Sites were located by Global Positioning System to within a 30 m radius. Water depths varied from about 0.3 m in the west near the Big Cypress Swamp to about 1.2 to 1.4 m near the L-67A and L-29 canals, mainly due to the topography of the area (Figure 2). The central study area is covered by sawgrass marshes interspersed with wet prairies and tree islands; sawgrass forms dense marshes in shallower waters and sparse marshes in deeper waters. Tree islands become predominate westward near the Big Cypress Swamp. Nine sites were cored on December 17, 1993 and February 12, 1994 at locations chosen to represent a wide range of conditions within the study area (Figure 2). Cores were collected using 60 cm long 3 inch (7.6 cm) internal diameter polycarbonate tubes. The top of each tube was threaded to accept an aluminum handle, and the bottom edge of each tube was sharpened to cut the peat with minimal compression. Cores were extruded and sliced at two-centimeter intervals into polypropylene cups and were frozen until prepared for analysis. If peat accumulated at about 2.0 mm yr−1 , a compromise between values of Delfino et al. (1993) and Craft and Richardson (1993), a core 20 cm long should contain material deposited during the
138
C. ARFSTROM ET AL.
past century. Most of our cores were about 36 cm long. The A-14 and A-10 cores are shorter because bedrock was reached, and were mostly off-white carbonate mud with some gray bands. The remaining profiles comprise very porous, dark brown to black peat with low mineral content. The transition from bedrock to soil was composed of a very fine gray mud. The method of total Hg analysis is detailed in Jones et al. (1995); Hg values reported in Table I are the average of three separate analyses. Overall reproducibility is better than ±1 ppb. To determine bulk density, a 10 mL beaker was filled with lightly packing wet soil, which was then transferred to an Al boat, weighed, dried overnight at 80 ◦ C and reweighed. Density is reported as dry weight per cm3 . Hg contents are reported both as ng Hg g−1 dry sediment and as ng Hg cc−1 . Total organic carbon was determined by loss on ignition. One gram of dried soil was combusted at 550 ◦ C for 3.5 hr. Ashed contents were weighed and the percentage of total organic carbon reported as % TOC = [(dry wt – ashed wt) dry wt−1 ] 100. Total phosphorus was determined on duplicates of each sample. Twenty-five mg dry ground soil were weighed into a 20 mL glass vial and 0.20 mL of 0.17 M MgSO4 and 1 mL of distilled water were added. The sample was evaporated to dryness at 80 ◦ C, then ashed at 550 ◦ C for 3.5 hr Ten mL of 0.24 N HCl were added, and the sample was dried overnight at 80 ◦ C to oxidize and hydrolyze all P-containing compounds to soluble reactive phosphate. Total phosphorus was measured with a rapid flow phosphorus analyzer, RFPA (Alpkem RFA 300; Solorzano and Sharp, 1980). 5. Results 5.1. S URFACE
SAMPLES
Sample locations, bulk density, and organic carbon, Hg and P concentrations are given in Table I. Organic carbon comprises close to 90% of soils throughout the study area; samples in the southwestern corner have higher marl contents and carbon contents as low as 60% (Figure 3a). Figure 3b shows surface soil-Hg concentrations on a volumetric basis (presentation of the data on a dry sediment basis is not significantly different). Hg contents range from 10.5 to 59.4 ng cc−1 (117–300 ppb dry sediment) with an average of 28.7 ng cc−1 ; most values are between 20 and 40 ng cc−1 . Hg contents in surface samples show no systematic trends with proximity to canals, from north to south, or from east to west. Soil-Hg concentrations were statistically analyzed using the procedures of Lepeltier (1969) and Sinclair (1974). All of the surface samples define a single straight line, indicating a single lognormal population (Figure 3b) and probably a common Hg source or source process for all samples. Both the even lateral distribution of Hg and the lognormal distribution of the sample population argue against a significant input of Hg from nearby point sources such as incinerators or from canal waters, and favor a single source that is homogenous over the
DISTRIBUTIONS OF MERCURY AND PHOSPHOROUS IN EVERGLADES SOILS
139
TABLE I Mercury, phosphorous, carbon and bulk density data for surface samples from southern WCA-3A
Site Latitude
A6 A7 A8 A9 A10 A11 A12 A13 A14 B6 B7 B8 B9 B10 B11 B12 B13 B14 C6 C7 C8 C9 C10 C11 C12 C13 C14 D6 D7 D8 D9 D10 D11 D12 D13 D14
26◦ 00.5970 25◦ 58.8470 25◦ 57.0970 25◦ 55.3470 25◦ 53.5970 25◦ 51.8470 25◦ 50.0970 25◦ 48.3470 25◦ 46.5970 26◦ 00.5970 25◦ 58.8470 25◦ 57.0970 25◦ 55.3470 25◦ 53.5970 25◦ 51.8470 25◦ 50.0970 25◦ 48.3470 25◦ 46.5970 26◦ 00.5970 25◦ 58.8470 25◦ 57.0970 25◦ 55.3470 25◦ 53.5970 25◦ 51.8470 25◦ 50.0970 25◦ 48.3470 25◦ 46.5970 26◦ 00.5970 25◦ 58.8470 25◦ 57.0970 25◦ 55.3470 25◦ 53.5970 25◦ 51.8470 25◦ 50.0970 25◦ 48.3470 25◦ 46.5970
Longitude
80◦ 48.8600 80◦ 48.8600 80◦ 48.8600 80◦ 48.8600 80◦ 48.8600 80◦ 48.8600 80◦ 48.8600 80◦ 48.8600 80◦ 48.8600 80◦ 46.9300 80◦ 46.9300 80◦ 46.9300 80◦ 46.9300 80◦ 46.9300 80◦ 46.9300 80◦ 46.9300 80◦ 46.9300 80◦ 46.9300 80◦ 45.0000 80◦ 45.0000 80◦ 45.0000 80◦ 45.0000 80◦ 45.0000 80◦ 45.0000 80◦ 45.0000 80◦ 45.0000 80◦ 45.0000 80◦ 43.0700 80◦ 43.0700 80◦ 43.0700 80◦ 43.0700 80◦ 43.0700 80◦ 43.0700 80◦ 43.0700 80◦ 43.0700 80◦ 43.0700
Dry ρ
Organic Total Hg
Total P
(g cc−1 ) C (%)
(ng g−1 ) (ng cc−1 ) (µg g−1 ) (µg cc−1 )
0.108 0.110 0.151 0.167 0.180 0.159 0.220 0.185 0.208 0.112 0.116 0.150 0.150 0.160 0.130 0.202 0.171 0.200 0.107 0.177 0.140 0.133 0.129 0.134 0.142 0.142 0.119 0.107 0.106 0.130 0.134 0.152 0.148 0.181 0.127 0.105
215 225 200 230 170 220 270 200 145 220 300 265 295 180 195 170 245 130 240 225 190 210 190 200 210 205 200 280 240 230 185 225 180 210 250 145
89.0 89.0 88.0 81.0 76.5 81.5 66.0 66.0 58.0 91.0 76.0 90.0 90.0 90.0 91.0 71.0 88.0 54.0 83.0 90.0 90.0 88.5 92.0 91.0 91.0 91.0 84.0 90.0 89.0 89.0 91.0 90.0 90.0 89.0 93.0 90.0
23.2 24.8 30.2 38.4 30.6 35.0 59.4 37.0 30.2 24.6 34.8 39.8 44.3 28.8 25.4 34.3 41.9 26.0 25.7 39.8 26.6 27.9 24.5 26.8 29.8 29.1 23.8 30.0 25.4 29.9 24.8 34.2 26.6 38.0 31.8 15.2
605 721 500 496 478 517 522 516 569 686 744 591 670 440 447 446 660 522 726 430 521 463 714 490 550 555 504 613 601 806 497 505 437 478 618 549
65.3 79.3 75.5 82.8 86.0 82.2 114.8 95.5 118.4 76.8 86.3 88.7 100.5 70.4 58.1 90.1 112.9 104.4 77.7 76.1 72.9 61.6 92.1 65.7 78.1 78.8 60.0 65.6 63.7 104.8 66.6 76.8 64.7 86.5 78.5 57.7
140
C. ARFSTROM ET AL.
TABLE I (continued) Site Latitude
E6 E7 E8 E9 E10 E11 E12 E13 E14 F6 F7 F8 F9 F10 F11 F12 G6 G7 G8 G9 G10 H6 H7 H8 H9 I6 I7 J6
26◦ 00.5970 25◦ 58.8470 25◦ 57.0970 25◦ 55.3470 25◦ 53.5970 25◦ 51.8470 25◦ 50.0970 25◦ 48.3470 25◦ 46.5970 26◦ 00.5970 25◦ 58.8470 25◦ 57.0970 25◦ 55.3470 25◦ 53.5970 25◦ 51.8470 25◦ 50.0970 26◦ 00.5970 25◦ 58.8470 25◦ 57.0970 25◦ 55.3470 25◦ 53.5970 26◦ 00.5970 25◦ 58.8470 25◦ 57.0970 25◦ 55.3470 26◦ 00.5970 25◦ 58.8470 26◦ 00.5970
Longitude
80◦ 41.1400 80◦ 41.1400 80◦ 41.1400 80◦ 41.1400 80◦ 41.1400 80◦ 41.1400 80◦ 41.1400 80◦ 41.1400 80◦ 41.1400 80◦ 39.2100 80◦ 39.2100 80◦ 39.2100 80◦ 39.2100 80◦ 39.2100 80◦ 39.2100 80◦ 39.2100 80◦ 37.2800 80◦ 37.2800 80◦ 37.2800 80◦ 37.2800 80◦ 37.2800 80◦ 35.3500 80◦ 35.3500 80◦ 35.3500 80◦ 35.3500 80◦ 33.4200 80◦ 33.4200 80◦ 31.4900
Dry ρ
Organic Total Hg
Total P
(g cc−1 ) C (%)
(ng g−1 ) (ng cc−1 ) (µg g−1 ) (µg cc−1 )
0.126 0.098 0.143 0.158 0.114 0.145 0.123 0.118 0.087 0.084 0.092 0.158 0.142 0.139 0.107 0.172 0.124 0.081 0.126 0.170 0.135 0.111 0.087 0.090 0.128 0.074 0.138 0.107
260 300 270 170 230 290 250 200 180 290 185 275 230 260 180 190 210 180 205 250 190 180 150 117 190 150 120 150
90.0 90.0 90.0 88.0 91.5 91.0 90.0 90.0 90.0 91.0 92.0 90.0 88.0 90.0 93.0 89.0 88.0 92.0 91.0 88.0 91.0 92.0 91.0 91.0 84.0 90.0 58.0 89.0
32.8 29.4 38.6 26.9 26.2 42.1 30.8 23.6 15.7 24.4 17.0 43.5 32.7 36.1 19.3 32.7 26.0 14.6 25.8 42.5 25.7 20.0 13.1 10.5 24.3 11.1 16.6 16.1
635 705 710 341 648 554 657 657 606 803 615 775 681 605 625 683 456 931 871 625 608 648 449 640 501 703 410 860
80.0 69.1 101.5 53.9 73.9 80.3 80.8 77.5 52.7 67.5 56.6 122.3 96.7 84.1 66.9 117.5 56.5 75.4 109.8 106.3 82.1 71.9 39.1 57.6 64.1 52.0 56.6 92.0
study area. Organic carbon and total Hg concentrations correlated poorly in surface samples (r 2 = 0.04), indicating that the organic content of the soils does not control mercury content. Phosphorous values range from 39.1 to 122.3 µg cc−1 , with an average of 79.1 µg cc−1 (350 to 850 ppm dry sediment, with a mean of 592 ppm). This range is similar to those reported from the interior parts of WCAs by earlier workers (Volk
DISTRIBUTIONS OF MERCURY AND PHOSPHOROUS IN EVERGLADES SOILS
141
Figure 3. Contours of element concentrations in surface samples: (A) Carbon; (B) Mercury, with a graph of concentration vs. log probability; (C) Phosphorous, with a graph of concentration vs. log probability.
142
C. ARFSTROM ET AL.
et al., 1975; Craft and Richardson, 1992). Higher P contents have been found in canal sediments (to more than 0.2%; Craft and Richardson, 1992; Merkel, 1997). Based on the rapid removal of P from waters leaving canals (Urban et al., 1993), the lack of P concentration gradients in surface samples with distance from canals (Figure 3c), and the agreement of our data with those from the interiors of other WCAs, we suggest that the that input of P to the interior of the study area by overland flow from canals is minimal. Like Hg, P contents of surface samples form a single lognormal population (Figure 3c), indicating a single source or source process for P. 5.2. C ORE
SAMPLES
Data for core samples are shown in Table II. Hg concentrations range from 1.7 to 41.9 ng cc−1 . Hg concentrations are higher near the surface than at greater depth in every core but D-14 (Figure 4). Hg increases by a factor of 6 in core A-7. Profile D14 shows relatively high mercury values throughout, but the highest values are still near the surface. Core C-12 contains high Hg near the surface and also between 25 and 35 cm depth. Several profiles (A-7, D-7, A-10, F-11, and D-14) show a distinct decrease in total mercury at the surface (0–2 cm sample). Vertical profiles of total phosphorus also show substantially higher concentrations near the surface in all cores except A-10 (Figure 5). Core A-10 shows irregular P distribution with depth, and the highest concentrations near the bottom of the core. Core A-14 also shows irregular variations, but has the highest P concentrations at the top. High P content near the surface is expected since phosphorus is transferred to upper layers by plant roots and much decomposition takes place in the upper parts of soil profiles. Hg and P correlate much more strongly within individual cores than in surface samples (Figure 6). Correlation coefficients (r) range from 0.16 to 0.97, and are largest in the profiles with the most pronounced near-surface Hg enrichments (cores A-7 and D-7). Also, at sites A-7, A-10, D-7, D-14, F-11 and I-7, Hg contents are lower in the 0–2 cm layer than in the 2–4 and 4–6 cm layers, a feature also noted in some sites by Rood et al. (1995). The 2–8 cm depth in Everglades peat soils is the interval where the greatest decomposition takes place, marked by the lowest Eh and highest methane contents (Bachoon and Jones, 1992). Based on the correlations between Hg and P in the cores, and the common maxima of Hg and P just below the surface (also seen in the WCA-3 Hg profile of Rood et al., 1995), we suggest that, like phosphorous, mercury may be taken up by plants and concentrated near the surface when they die, accumulate and decay. The soil Hg concentrations therefore may well not directly reflect the Hg deposition history. Log probability plots for Hg and P reveal at least two lognormal populations in several core samples (Figures 7 and 8). Like the correlation between Hg and P, the presence of separate lognormal populations is most pronounced in profiles with the greatest near-surface Hg and P enrichments (A-7, D-7, D-10, F-11). Background
DISTRIBUTIONS OF MERCURY AND PHOSPHOROUS IN EVERGLADES SOILS
143
Figure 4. Hg concentrations in soil cores. Most profiles show elevated values near the surface, but lower values in the uppermost 2 cm interval.
144
C. ARFSTROM ET AL.
TABLE II Mercury and phosphorous data for cores from Everglades WCA-3A. Accumulation rate calculated assuming a peat accumulation rate of 2 mm yr−1 Depth
Density
Total Hg
(cm)
(gm cc−1 )
(ng g−1 )
(ng cc−1 )
Total P (ng g−1 )
236 293 322 279 298 162 120 87 67 50 42 41 43 58 51 45 62 47
23.6 32.2 41.9 36.3 29.8 17.8 12.0 8.7 6.7 6.0 4.2 4.1 4.3 6.4 6.1 4.5 6.2 7.1
716 603 623 569 505 463 358 329 327 266 249 247 259 229 194 207 209 205
71.6 66.3 81.0 74.0 50.5 50.9 35.8 32.9 32.7 31.9 24.9 24.7 25.9 25.2 23.3 20.7 20.9 30.8
165 171 90 75 75 58 45 55 44 52 44 48 36 23
19.8 29.1 9.0 7.5 8.3 11.6 9.5 12.7 7.9 8.3 14.5 13.9 14.8 10.8
515 601 539 479 393 360 351 346 311 305 330 314 311 235
61.8 102.2 53.9 47.9 43.2 72.0 73.7 79.6 56.0 48.8 108.9 91.1 127.5 110.5
(µg cc−1 )
6Hg
Hg deposition
(µg)
(µg m−2 yr−1 )
Core A-7 0–2 2–4 4–6 6–8 8–10 10–12 12–14 14–16 16–18 18–20 20–22 22–24 24–26 26–28 28–30 30–32 32–34 34–36
0.10 0.11 0.13 0.13 0.10 0.11 0.10 0.10 0.10 0.12 0.10 0.10 0.10 0.11 0.12 0.10 0.10 0.15
19.5
47.2 64.5 83.7 72.5 59.6 35.6 24.0 17.4 13.4 12.0 8.4 8.2 8.6 12.8 12.2 9.0 12.4 14.1
Core A-10 0–2 2–4 4–6 6–8 8–10 10–12 12–14 14–16 16–18 18–20 20–22 22–24 24–26 26–28
0.12 0.17 0.10 0.10 0.11 0.20 0.21 0.23 0.18 0.16 0.33 0.29 0.41 0.47
7.7
39.6 58.1 18.0 15.0 16.5 23.2 18.9 25.3 15.8 16.6 29.0 27.8 29.5 21.6
DISTRIBUTIONS OF MERCURY AND PHOSPHOROUS IN EVERGLADES SOILS
145
TABLE II (continued) Depth
Density
Total Hg
Total P
(cm)
(gm cc−1 )
(ng g−1 )
(ng cc−1 )
(ng g−1 )
(µg cc−1 )
137 125 112 104 89 87 82 88 76
35.6 25.0 16.8 18.7 17.8 20.0 19.7 19.4 16.7
653 562 501 458 453 460 538 476 284
169.8 112.4 75.2 82.4 90.6 105.8 129.1 104.7 62.5
212 215 148 157 147 134 65 19 21 18 23 41 99 108 109 83 14 10
21.2 21.5 17.8 15.7 14.7 13.4 6.5 1.9 2.1 2.0 3.0 4.9 12.9 15.1 12.0 8.3 1.8 1.7
538 513 454 457 454 415 270 225 251 214 202 193 174 198 176 173 55 46
53.8 51.3 54.5 45.7 45.4 41.5 27.0 22.5 25.1 23.5 26.3 23.2 22.6 27.7 19.4 17.3 7.2 7.8
235 272 268 230
23.5 27.2 32.2 25.3
627 587 465 491
62.7 58.7 55.8 54.0
6Hg
Hg deposition
(µg)
(µg m−2 yr−1 )
Core A-14 0–2 2–4 4–6 6–8 8–10 10–12 12–14 14–16 16–18
0.26 0.20 0.15 0.18 0.20 0.23 0.24 0.22 0.22
8.7
71.2 50.0 33.6 37.4 35.6 40.0 39.4 38.7 33.4
Core C-12 0–2 2–4 4–6 6–8 8–10 10–12 12–14 14–16 16–18 18–20 20–22 22–24 24–26 26–28 28–30 30–32 32–34 34–36
0.10 0.10 0.12 0.10 0.10 0.10 0.10 0.10 0.10 0.11 0.13 0.12 0.13 0.14 0.11 0.10 0.13 0.17
10.1
42.4 43.0 35.5 31.4 29.4 26.8 13.0 3.8 4.2 4.0 6.0 9.8 25.7 30.2 24.0 16.6 3.6 3.4
Core D-7 0–2 2–4 4–6 6–8
0.10 0.10 0.12 0.11
47.0 54.4 64.3 50.6
146
C. ARFSTROM ET AL.
TABLE II (continued) Depth
Density
Total Hg
Total P
(cm)
(gm cc−1 )
(ng g−1 )
(ng cc−1 )
(ng g−1 )
(µg cc−1 )
200 148 133 111 96 88 69 62 49 39 29 51 38 54
20.0 14.8 13.3 11.1 9.6 8.8 6.9 6.8 4.9 3.9 3.2 5.6 4.2 7.6
429 374 322 291 306 243 208 192 174 168 156 148 157 136
42.9 37.4 32.2 29.1 30.6 24.3 20.8 21.1 17.4 16.8 17.2 16.3 17.3 19.0
225 160 145 130 81 62 62 51 31 38 49 85 68 64 49 51 34 42
29.3 16.0 14.5 13.0 8.1 6.2 6.2 6.1 3.1 3.8 4.9 9.4 7.5 6.4 4.9 5.1 3.4 4.2
413 383 396 377 266 247 218 202 160 160 179 162 154 141 118 132 117 125
53.7 38.3 39.6 37.7 26.6 24.7 21.8 24.2 16.0 16.0 17.9 17.8 16.9 14.1 11.8 13.2 11.7 12.5
6Hg
Hg deposition
(µg)
(µg m−2 yr−1 )
Core D-7 8–10 10–12 12–14 14–16 16–18 18–20 20–22 22–24 24–26 26–28 28–30 30–32 32–34 34–36
0.10 0.10 0.10 0.10 0.10 0.10 0.10 0.11 0.10 0.10 0.11 0.11 0.11 0.14
18.1
40.0 29.6 26.6 22.2 19.2 17.6 13.8 13.6 9.8 7.8 6.4 11.2 8.4 15.1
Core D-10 0–2 2–4 4–6 6–8 8–10 10–12 12–14 14–16 16–18 18–20 20–22 22–24 24–26 26–28 28–30 30–32 32–34 34–36
0.13 0.10 0.10 0.10 0.10 0.10 0.10 0.12 0.10 0.10 0.10 0.11 0.11 0.10 0.10 0.10 0.10 0.10
7.9
58.5 32.0 29.0 26.0 16.2 12.4 12.4 12.2 6.2 7.6 9.8 18.7 15.0 12.8 9.8 10.2 6.8 8.4
DISTRIBUTIONS OF MERCURY AND PHOSPHOROUS IN EVERGLADES SOILS
147
TABLE II (continued) Depth
Density
Total Hg
Total P
(cm)
(gm cc−1 )
(ng g−1 )
(ng cc−1 )
(ng g−1 )
(µg cc−1 )
144 255 248 244 229 158 140 107 110 139 182 130 160 163 174 171 180
14.4 33.2 29.8 31.7 22.9 15.8 14.0 11.8 13.2 20.9 29.1 18.2 27.2 31.0 29.6 27.4 30.6
427 595 478 464 425 412 379 318 291 292 280 274 273 265 241 228 255
42.7 77.4 57.4 60.3 42.5 41.2 37.9 35.0 34.9 43.8 44.8 38.4 46.4 50.4 41.0 36.5 43.4
177 204 185 119 73 77 73 47 47 62 52 64 52 79 87
21.2 26.5 31.5 14.3 7.3 10.0 8.0 4.7 5.2 6.8 5.2 6.4 5.2 8.7 10.4
413 383 396 377 266 247 218 202 160 160 179 162 154 141 118
49.6 49.8 67.3 45.2 26.6 32.1 24.0 20.2 17.6 17.6 17.9 16.2 15.4 15.5 14.2
6Hg
Hg deposition
(µg)
(µg m−2 yr−1 )
Core D-14 0–2 2–4 4–6 6–8 8–10 10–12 12–14 14–16 16–18 18–20 20–22 22–24 24–26 26–28 28–30 30–32 32–34
0.10 0.13 0.12 0.13 0.10 0.10 0.10 0.11 0.12 0.15 0.16 0.14 0.17 0.19 0.17 0.16 0.17
13.4
28.8 66.3 59.5 63.4 45.8 31.6 28.0 23.5 26.4 41.7 58.2 36.4 54.4 61.9 59.2 54.7 61.2
Core F-11 0–2 2–4 4–6 6–8 8–10 10–12 12–14 14–16 16–18 18–20 20–22 22–24 24–26 26–28 28–30
0.12 0.13 0.17 0.12 0.10 0.13 0.11 0.10 0.11 0.11 0.10 0.10 0.10 0.11 0.12
11.2
42.5 53.0 62.9 28.6 14.6 20.0 16.1 9.4 10.3 13.6 10.4 12.8 10.4 17.4 20.9
148
C. ARFSTROM ET AL.
TABLE II (continued) Total P
6Hg
Hg deposition
(µg)
(µg m−2 yr−1 )
Depth
Density
Total Hg
(cm)
(gm cc−1 )
(ng g−1 )
(ng cc−1 )
(ng g−1 )
(µg cc−1 )
0.12 0.11 0.12
106 83 80
12.7 9.1 9.6
132 117 125
15.8 12.9 15.0
25.4 18.3 19.2
0.16 0.21 0.15 0.12 0.11 0.11 0.12 0.13 0.11 0.11 0.11 0.11 0.11 0.12 0.13
148 142 119 116 122 111 120 88 77 76 63 33 61 68 51
23.7 29.8 17.9 13.9 13.4 12.2 14.4 11.4 8.5 8.4 6.9 3.6 6.7 8.2 6.6
373 366 353 319 278 240 211 226 247 212 188 138 235 303 204
59.7 76.9 53.0 38.3 30.6 26.4 25.3 29.4 27.2 23.3 20.7 15.2 25.9 36.4 26.5
47.4 59.6 35.7 27.8 26.8 24.4 28.8 22.9 16.9 16.7 13.9 7.3 13.4 16.3 13.3
Core F-11 30–32 32–34 34–36 Core I-7 0–2 2–4 4–6 6–8 8–10 10–12 12–14 14–16 16–18 18–20 20–22 22–24 24–26 26–28 28–30
13.2
populations of both elements are relatively constant below a certain depth in all profiles except for C-12 and D-14. These low values may reflect preindustrial atmospheric deposition of Hg, derivation from the substrate or both. Dissolved Hg and P from the subjacent limestone could be carried in groundwater and taken up by the lower layers of peat.
DISTRIBUTIONS OF MERCURY AND PHOSPHOROUS IN EVERGLADES SOILS
Figure 5. Total P in soil cores; all but one have the highest P content near the surface.
149
150
C. ARFSTROM ET AL.
Figure 6. Mercury vs phosphorus in soil cores. The two elements correlate well in most cores, with slopes typically between 0.4 and 0.6.
DISTRIBUTIONS OF MERCURY AND PHOSPHOROUS IN EVERGLADES SOILS
151
Figure 7. Log probability plot of mercury concentrations in profiles. Separate background and anomalous values are clearly present in most profiles.
152
C. ARFSTROM ET AL.
Figure 8. Log probability plot of phosphorous concentrations in profiles. Separate background and anomalous values are clearly present in most profiles.
DISTRIBUTIONS OF MERCURY AND PHOSPHOROUS IN EVERGLADES SOILS
153
6. Discussion 6.1. S URFACE
CONCENTRATIONS AND DISTRIBUTIONS OF SOIL MERCURY AND PHOSPHORUS
Mercury contents vary widely in soils, largely because Hg has a high affinity for organic matter and fine-grained minerals. Kabata-Pendias and Pendias (1984) and Alloway (1990) cite the following ranges of soil Hg contents (in ppb dry soil): soils over calcareous rocks: 10–500; organic light soils: 10–4000; and forest soils: 20– 150. Total Hg in surface samples from the study area (117 to 300 ppb dry sediment) are not high compared to other organic-rich soils. High Hg contents in Everglades fish must therefore be related to the processes that make Hg available from the soil rather than high Hg content of the soil itself. Homogenous surface distributions of Hg and P seem to rule out nearby point sources or canals as the major sources of Hg in the study area. Mercury and phosphorous show a correlation in surface samples (r 2 = 0.45). A perfect correlation would not be expected, since P has no appreciable vapor phase and rainwater contains little phosphorus (Horne and Goldman, 1994); however, some part of the phosphorous and mercury in the surface samples may come from a common source. Possible potential sources of these elements likely to be evenly distributed over the study area include the limestone substrate of the Miami Formation and deposition from the atmosphere. Merkel (1997) reported 10–14 ppb Hg in limestone samples collected along the western margin of the study area, and a total range of 9–60 ppb for limestones (some weathered) in the Everglades. The limestone, and groundwater equilibrated with the substrate, could therefore supply some part of the Hg in the peat. P contents of Miami Formation limestones from the Everglades range from 16–74 ppm, and P/Hg values range from 1051 to 8409, with an average value of about 3800 (Merkel, 1997). Surface samples of peat have average P/Hg near 2900, somewhat lower than the limestone substrate. There are presently no precise data on P/Hg values for atmospheric deposition in the area available for comparison. The atmosphere contains ≈850 metric tons of Hg, about 95% as Hg◦ vapor and the remaining 5% associated with particulates and organomercury compounds. Mercury vapor resides for 0.7–2.0 yr in the atmosphere (Nater and Grigal, 1992), and may travel far from its source. Particulate Hg associated with ash and soot may be efficiently removed by precipitation and deposited relatively close to the source (Mitra, 1986; Alloway, 1990). Anthropogenic emissions, which now make up about 30% of atmospheric Hg (Krenkel, 1990) are thought to have roughly tripled since the turn of the century (Mitra, 1986). Recent data from sediment traps indicate that atmospheric Hg deposition in south Florida is strongly seasonal, with 80–90% occurring during the summer rainy season (Guentzel et al., 1995). Hg deposition is about 29% higher in south Florida than in north-central Florida, but this gradient is not apparent at the scale of the study area. No significant gradient in Hg deposition was found from east to west
154
C. ARFSTROM ET AL.
across the Florida peninsula. Guentzel et al. (1995) concluded that atmospheric Hg deposition in the Everglades is driven by large scale regional or hemispheric processes rather than local sources. Dispersed atmospheric mercury is also a likely source of the mercury. 6.2. S OIL
MERCURY AND PHOSPHOUS PROFILES
Eight of the nine cores have their highest Hg contents near the surface (Figure 4). Sites A-7 and D-7 show increases of 5–7 times near the surface compared to 20– 36 cm depth. Sites A-14 and I-7 show increases of 2–3 times. Similar patterns were found by Rood et al. (1995), who calculated Hg accumulation rates of 7–11 µg m−2 yr−1 (average of 10 µg m−2 yr−1 ) near the turn of the century and 28–55 µg m−2 yr−1 (average 39 µg m−2 yr−1 ) since 1985 in WCA 3. The same authors calculated an overall average deposition rate for the Everglades since 1985 of 53 µg m−2 yr−1 , with values as high as 141 µg m−2 yr−1 in individual cores. We did not date our core samples, but a peat accumulation rate of 2 mm yr−1 would yield Hg deposition rates of up to 83.7 µg m−2 yr−1 in shallow sediments from our data (Table II). Recent direct measurements from traps yield Hg deposition rates of 19–28 µg m−2 yr−1 in south Florida (Guentzel et al., 1995), which are substantially lower than rates calculated from cores by Rood et al. (1995), and those approximated from our data. While one could speculate that Hg deposition rates have dropped drastically in the past 5 yr, or the measurements of Guentzel et al. (1995) are not representative for some reason, at this point no independent support for either speculation is available. However, the Hg deposition rate estimates of Rood et al. (1995) depend on the assumption that Hg is completely immobile after deposition, and this may not be correct. Cocking et al. (1995) documented uptake of soil Hg by several type of plants, and found the largest amounts of uptake by grass species. Some species of plants are so efficient in taking metals from the environment that they may be used in exploration for ore deposits or used as biomonitors for ambient metals (Rusmussen, 1994). Everglades peats accumulate from and interact with active plant communities, predominantly sawgrass and spike rush. If plants take up mercury and concentrate it upward in the soil profile, mercury deposition rates estimated using peat cores as simple Hg deposition records will be overestimates. In contrast to Hg, the behavior of P in Everglades peat soils is fairly wellunderstood. P is a growth-limiting nutrient which is rapidly removed from natural waters by microbes and plants, particularly in the Everglades where plant communities are adapted to low P conditions (Stewart, 1984). P is taken up by roots from deeper levels of the soil, incorporated into living plants and concentrated where plant matter accumulates and decays (Horne and Goldman, 1994). In addition, bacteria may release phosphorus from minerals for uptake by plants (Fenchel and Blackburn, 1979). Higher near-surface P contents of peats far from canals in the study area suggest that the vertical P distribution is largely controlled by plant uptake.
DISTRIBUTIONS OF MERCURY AND PHOSPHOROUS IN EVERGLADES SOILS
155
There are presently no direct data on bioaccumulation of Hg by sawgrass or other Everglades species, but the close correlation between P and Hg in cores suggests that Hg may also be at least partly remobilized by sawgrass and other plants. Such a process could concentrate it in the upper part of the soil profile, accentuating the likely effect of increasing secular atmospheric Hg deposition in the profiles. Uptake of Hg by sawgrass would also explain the elevated Hg just below the surface where most accumulation and decay of sawgrass takes place. The lowest Hg concentrations in all but two of the cores occur somewhat above their bottoms, which could indicate the depth at which Hg is being taken up by plants. In the lowest levels of all cores except F-11 and D-14 (which were both taken in thick peat in inundated areas and probably did not approach the bottom of the soil profile), P/Hg values are fairly constant near 4000, which is about the average value for the Miami Formation limestone substrate (Merkel, 1997). In the upper 10 cm of most cores, P/Hg values tend to decrease toward the surface and to lie between 2000 and 3000. These observations could be explained by derivation of P and Hg from the substrate with little chemical fractionation in the bottom of the cores, and fractionation in the upper cores either due to mobilization by plants, increased atmospheric deposition, or both. 6.3. D ILUTION
AND POSTDEPOSITIONAL MOBILITY OF
Hg IN
PEAT
Core I-7 in the northeastern corner of the study area showed high organic matter content, a deep soil profile and a comparatively small increase in Hg content from deep to shallow samples (3 times). Rood et al. (1995) stated that rapid peat accumulation dilutes mercury concentrations in sites near canals. Craft and Richardson (1993) found that peat accumulated 2 to 3 times as fast in areas subject to long or permanent inundation and/or nutrient enriched waters compared to areas not subject to these conditions. We used the dilution hypothesis to test the post-depositional mobility of mercury with a simple model. If deposition is primarily of dispersed atmospheric mercury, and if mercury remains immobile in peat after deposition as is commonly assumed (e.g. Rood et al., 1995), each sample site should receive roughly the same amount of mercury over time, and each core should contain about the same total amount of mercury. These conditions imply that the increase in Hg content in peat due to increasing atmospheric deposition should take place over a thicker interval of peat than in a profile with slow accumulation, and the maximum Hg concentration should be lower (Figure 9); this is the dilution effect referred to by Rood et al. (1995). In short, the maximum Hg concentration and accumulation interval should be inversely proportional regardless of the peat accumulation rate in any given core. If we assume that atmospheric deposition of Hg increased above background levels in each core at about the same time, total Hg deposition at each site may be compared by adding up the amount of mercury in each core above this level.
156
C. ARFSTROM ET AL.
Figure 9. Model of relationship between peat accumulation rate and maximum Hg content in cores if deposition is homogenous and no postdepositional mobilization of Hg occurs.
Computed Hg contents for the enriched part of each core range from 7.7 to 19.5 µg (Table II); either Hg deposition was not as even across the study area as the surface data would suggest, or Hg was mobile after deposition, or both. The homogeneity of Hg deposition in south Florida observed by Guentzel et al. (1995) indicates that postdepositional mobility is the more important factor. Maximum Hg contents in each core plotted against accumulation interval show no correlation (Figure 10); rapid peat accumulation may dilute Hg in some samples, but the Hg contents in our cores are not simple records of dispersed atmospheric Hg deposition. Cores C-12 and D-14 differ from the others, having high Hg values both near the surface and at about 30 cm depth. It seems unlikely that Hg deposition in those sites increased and decreased twice over the last few decades while nearby sites experienced a nearly continuous increase in mercury deposition. If Hg is immobile after deposition, such high values could be caused by episodes of unusually low peat accumulation rate (perhaps caused by drought) at those sites. It seems unlikely, however, that drought could affect site D-14, which is near the L-29 canal and subject to near-permanent inundation, without also affecting all of the other sites that are higher and further from the canals (Figure 2). It seems more likely that fire could have affected cores C-12 and D-14 without affecting the other cores, or that the variations in cores C-12 and D-14 are related to some other type of post-depositional Hg mobility.
DISTRIBUTIONS OF MERCURY AND PHOSPHOROUS IN EVERGLADES SOILS
157
Figure 10. Maximum Hg content vs accumulation interval for soil cores, omitting cores C-12 and D-14. No correlation is apparent, indicating that Hg deposition is either less homogenous than indicated by surface data, or that Hg is mobilized after deposition.
7. Conclusions The range of total phosphorus concentrations in surface samples from WCA-3A (350–850 ppm dry sediment) is typical of previous data from the interior regions of WCAs. Surface P contents do not show pronounced gradients from north to south or with distance from canals. Surface soil-Hg concentrations range from 117 to 300 ppb, which are typical for organic-rich soils, and do not correlate with soil bulk density or the percentage of total organic carbon. A modest correlation of mercury with total phosphorus is seen. Mercury and phosphorous contents of surface samples each define single lognormal populations. Like phosphorous, surface mercury concentrations do not show pronounced lateral gradients. The source of both elements in surface peat is probably dominated by single, laterally-homogenous sources. Regional atmospheric mercury is probably the main soil Hg source in the Everglades. Core samples show elevated values for both P and Hg near, usually 4–8 cm below, the surface. Phosphorous profiles in cores are determined by uptake of phosphorous by plants and concentration of that element in the upper layer of decaying plant matter in this nutrient-poor environment. In contrast to surface samples, Hg and P correlate well in the profiles, and define more than one lognormal population
158
C. ARFSTROM ET AL.
in most cores. The correlation with P suggests that Hg may also have experienced post-depositional mobilization and concentration near the surface by plants and microorganisms. Hg contents of near-surface samples are too high to be explained by atmospheric deposition alone, according to recent sediment trap measurements by Guentzel et al. (1995) and peat deposition rate estimates of Rood et al. (1995) and Craft and Richardson (1993). Our results suggest that previous attempts to use the Hg content of peat cores in the Everglades as simple depositional records of Hg may have produced significant overestimates of Hg deposition rates. Anthropogenic Hg emissions have increased with the industrialization of modern societies over the past 200 to 300 yr (Nriagu, 1990). Since 1900, anthropogenic emissions increased largely due to combustion of fossil fuel and burning of industrial and municipal waste. Soil profiles of organic-rich sediments must reflect these increases in some way; however, more complete knowledge of atmospheric Hg depositional rates and studies of mercury uptake by Everglades grass species are needed before a more accurate picture of mercury cycling in the Everglades can emerge.
Acknowledgements This study formed part of the Master’s thesis of the senior author in the Department of Geology at FIU. The research was supported by funding from the National Park Service (Everglades National Park) and the United States Environmental Protection Agency through cooperative agreement CA 5280-1-9016. This paper is contribution 117 of the Southeast Environmental Research Program at Florida International University.
References Alloway, B. J.: 1990, Heavy Metals in Soils, Glasgow, Blackie, New York, Halsted Press, 339 p. Bachoon, D. and Jones, R. D.: 1992, Soil Biol. Biochem. 24, 21. Brown, R. B., Stone, E. L. and Carlisle, V. W.: 1991, ‘Soils’, in Myers, R. L. and Ewel, J. J. (eds.), Ecosystems of Florida, University of Central Florida, Press-Orlando, Ch. 3. Cocking, D., Rohrer, M., Thomas, R., Walker, J. and Ward, D.: 1995, Water, Air, and Soil Pollut. 80, 1113. Craft, C. B. and Richardson, C. J.: 1993, Ecological Applications 3, 446. Davis, S. M. and Ogden, J. C.: 1994, Everglades: The Ecosystem and Its Restoration, St. Lucie Press, 826 pp. Delfino, J. J., Crisman, T. L., Gottgens, J. F., Rood, B. E. and Earle, C. D. A.: 1993, Spatial and Temporal Distribution of Mercury in Everglades and Okefenokee Wetland Sediments, Dept. of Envir. Eng. Sci. University of Florida, 119 pp. Fenchel, T. and Blackburn, T. H.: 1979, Bacteria and Mineral Cycling, Academic Press, 225 pp. Gleason, P. J. and Stone, P.: 1994, ‘Age, Origin and Landscape Evolution of the Everglades Peatland’, in Davis, S. M. and Ogden, J. C. (eds.), Everglades: The Ecosystem and its Restoration, pp. 149– 197.
DISTRIBUTIONS OF MERCURY AND PHOSPHOROUS IN EVERGLADES SOILS
159
Guentzel, J. L., Landing, W. M., Gill, G. A. and Pollman, C. D.: 1995, Water, Air, and Soil Pollut. 80, 393. Gunderson, L. H. and Loftus, W. F.: 1993, ‘The Everglades’, W. H. Martin, S. G. Boyce and A. C. Echternacht (eds.), Biodiversity of the Southeastern United States Lowland Terrestrial Communities, John Wiley & Sons, Inc., pp. 199–255. Hand, J. and Friedeman, M.: 1990, Mercury, Largemouth Bass and Water Quality: A Preliminary Report, Florida Dept. of Environ. Reg. Tallahassee, 112 pp. Horne, A. J. and Goldman, C. H.: 1994, Limnology, 2nd ed., McGraw-Hill, Inc., 576 pp. Jones, R. D.: 1992, Environmental Parameters Affecting the Release of Mercury from Soils of the Florida Everglades, Project proposal. Jones, R. D., Jacobson, M. E., Jaffe, R., West-Thomas, J., Arfstrom, C. and Alli, A.: 1995, Water, Air, and Soil Pollut. 80, 1285. Kabata-Pendias, A. and Pendias, H.: 1984, Trace Elements in Soils and Plants, CRC Press, 315 p. Krenkel, P.: 1990, ‘Mercury in the Environment: Methylation and Sediment Reactions’, Proc. of Workshop on Mercury Contamination in Florida: Impacts and Solutions, 140 pp. Lepeltier, C.: 1969, Economic Geology 64, 538. Merkel, R. S.: 1997, Sediment and Elemental Accumulations in Everglades Excavations, Unpublished M.S. thesis, Florida International University, 203 p. Mitra, S.: 1986, Mercury in the Ecosystem, Switzerland, Germany, U.K., and U.S.A., Trans. Tech. Publications, 327 pp. Myers, R. L. and Ewel, J. J.: 1991, Ecosystems of Florida, University of Central Florida Press, 765 pp. Nater, E. A. and Grigal, D. F.: 1992, Nature 358, 139. Nriagu, J.: 1990, Environment 32(7), 7. Reddy, K. R., DeBusk, W. F., Wang, Y., DeLaune, R. and Koch, M.: 1991, Physico-Chemical Properties of Soils in the Water Conservation Area 2 of the Everglades – Final Report, Soils Science Dept., Inst. of Food and Agricultural Sci., University of Florida, Gainsville, 214 pp. Rood, B. E., Gottgens, J. F., Delfino, J. J., Earle, C. D. and Crisman, T. L.: 1995, Water, Air and Soil Pollut. 80, 981. Rusmussen, P. E.: 1994, ‘Mercury in Vegetation of the Precambrian Shield’, in Watras, C. J. and Huckabee, J. W. (eds.), Mercury Pollution: Integration and Synthesis, CRC Press, 727 pp. Sinclair, A. J.: 1974, Journal of Geochemical Exploration 3, 129. Scott, T. M.: 1992, A Geological Overview of Florida, Florida Geological Survey, open file report No. 50. 78 pp. Solorzano, L. and Sharp, J. H.: 1980, Limnology Oceanography 25, 754. Stewart, K. K.: 1984, ‘Physiological, Edaphic and Environmental Characteristics of Everglades Sawgrass Communities’, in Gleason, P. J. (ed.), Environments of South Florida Present and Past II, Miami Geological Society, Coral Gables, Florida, 551 pp. Urban, N. H., Davis, S. M. and Aumen, N. G.: 1993, Aquat. Bot. 46, 203. Volk, B. G., Schemnitz, S. D., Gamble, J. F. and Sartain, J. B.: 1975, ‘Baseline Data on Everglades Soil-Plant Systems: Elemental Composition, Biomass and Soil Depth’, in F. G. Howell, J. B. Gentry, and M. H. Smith (eds.), Mineral Cycling in Southeastern Ecosystems, National Technical Information Service, U.S. Depart. of Commerce, Springfield, VA, pp. 658–672. Ware, F., Royals, H. and Lange, T.: 1990 Proc. Am. Conf. SEAFWA 44, 5.