Natural factors and human modifications contributing to marsh loss in Louisiana's Mississippi River Deltaic Plain

Natural factors and human modifications contribute to the estimated annual loss of 10,200 ha of coastal land in the Mississippi River Deltaic Plain Region of south Louisiana. This paper combines information on regional geology and human-induced habitat alterations to evaluate the relative importance of human and natural factors to marsh loss. Data on marsh area and habitat type for 139 7.5-min quadrangles were calculated from maps based on aerial photographs from 1955/56 and 1978, and data on regional geology obtained from published maps were used to construct multivariate model relating initial marsh area, change in urban and agricultural area, change in canal and spoil area, canal area in 1978, depth of sediment overlying the Prairie terrace, and subdelta age to marsh loss. The model indicated that between 25.0% and 39.0% of the marsh loss that occurred during the 23-year period was related to canal and spoil construction, and between 9.5% and 12.7% was related to urban and agricultural development. These are minimal estimates of loss because they do not include many secondary effects (for example, canal orientation, saltwater intrusion, and eutrophication) that can also result in indirect loss. Depth of sediment, initial marsh area, delta lobe age by 1978 canal and spoil area interaction, and indirect effects not included in the model accounted for remaining marsh loss.     

Plant Community Composition and Biomass in Gulf Coast Chenier Plain Marshes: Responses to Winter Burning and Structural Marsh Management

Many marshes in the Gulf Coast Chenier Plain, USA, are managed through a combination of fall or winter burning and structural marsh management (i.e., levees and water control structures; hereafter SMM). The goals of winter burning and SMM include improvement of waterfowl and furbearer habitat, maintenance of historic isohaline lines, and creation and maintenance of emergent wetlands. Although management practices are intended to influence the plant community, effects of these practices on primary productivity have not been investigated. Marsh processes, such as vertical accretion and nutrient cycles, which depend on primary productivity may be affected directly or indirectly by winter burning or SMM. We compared Chenier Plain plant community characteristics (species composition and above- and belowground biomass) in experimentally burned and unburned control plots within impounded and unimpounded marshes at 7 months (1996), 19 months (1997), and 31 months (1998) after burning. Burning and SMM did not affect number of plant species or species composition in our experiment. For all three years combined, burned plots had higher live above-ground biomass than did unburned plots. Total above-ground and dead above-ground biomasses were reduced in burned plots for two and three years, respectively, compared to those in unburned control plots. During all three years, belowground biomass was lower in impounded than in unimpounded marshes but did not differ between burn treatments. Our results clearly indicate that current marsh management practices influence marsh primary productivity and may impact other marsh processes, such as vertical accretion, that are dependent on organic matter accumulation and decay.     

Spatial and temporal changes in Louisiana's Barataria Basin marshes, 1945–1980

A computerized geographic information system with site-specific change-detection capabilities was developed to document amounts, rates, locations, and sequences of loss of coastal marsh to open water in Barataria Basin, Louisiana, USA. Land-water interpretations based on 1945, 1956, 1969, and 1980 aerial photographs were used as input, and a modified version of the Earth Resources Laboratory Applications Software developed by the National Aeronautics and Space Administration was used as a spatial data base management system. Analysis of these data sets indicates that rates of marsh loss have increased from 0.36% per year in the 1945–56 period, to 1.03% per year in 1956–69, and to 1.96% per year in 1969–80. The patterns of marsh loss indicate that the combination of processes causing degradation of the marsh surface does not affect all areas uniformly. Marsh loss rates have been highest where freshwater marshes have been subject to saltwater intrusion. The increase in the wetland loss rates corresponds to accelerated rates of subsidence and canal dredging and to a cumulative increase in the area of canals and spoil deposits.     

Salt marsh vegetation change in response to tidal restriction

Vegetation change in response to restriction of the normal tidal prism of six Connecticut salt marshes is documented. Tidal flow at the study sites was restricted with tide gates and associated causeways and dikes for purposes of flood protection, mosquito control, and/or salt hay farming. One study site has been under a regime of reduced tidal flow since colonial times, while the duration of restriction at the other sites ranges from less than ten years to several decades. The data indicate that with tidal restriction there is a substantial reduction in soil water salinity, lowering of the water table level, as well as a relative drop in the marsh surface elevation. These factors are considered to favor the establishment and spread ofPhragmites australis (common reed grass) and other less salt-tolerant species, with an attendant loss ofSpartina-dominated marsh. Based on detailed vegetation mapping of the study sites, a generalized scheme is presented to describe the sequence of vegetation change from typicalSpartina- toPhragmites-dominated marshes. The restoration of thesePhragmites systems is feasible following the reintroduction of tidal flow. At several sites dominated byPhragmites, tidal flow was reintroduced after two decades of continuous restriction, resulting in a marked reduction inPhragmites height and the reestablishment of typical salt marsh vegetation along creekbanks. It is suggested that large-scale restoration efforts be initiated in order that these degraded systems once again assume their roles within the salt marsh-estuarine ecosystem.     

Coastal zone development: Mitigation,marsh creation,and decision-making

Marsh creation is currently receiving wide attention in the United States as an important tool for mitigating the impacts of development in coastal wetlands. The perception that there is no net loss in valuable coastal wetlands when development is mitigated by the creation of man-made marshes can have a substantial impact on the permitting and decision-making processes. The effective result may be the trading of natural salt marshes for man-made marshes.Techniques for marsh creation were developed by the US Army Corps of Engineers to enhance and stabilize dredge spoil materials. Most research sponsored by the Corps has been directed at determining whether these goals have been accomplished. A survey of the research indicates that there is insufficient evidence to conclude that man-made marshes function like natural salt marshes or provide the important values of natural marshes. It is necessary, therefore, for decision-makers to understand the limitations of present knowledge about man-made marshes, realistically evaluate the trade-offs involved, and relegate mitigation to its proper role in the permitting process—post facto conditions imposed on developments that clearly meet state qualifications and policies.     

A rationale for coastal wetland restoration through spoil bank management in Louisiana,USA

The rationale and outline of an implementation plan for restoring coastal wetlands in Louisiana is presented. The rationale for the plan is based on reversing the consequences of documented cause-and-effect relationships between wetland loss and hydrologic change. The main feature is to modify the extensive interlocking network of dredged spoil deposits, or spoil banks, by reestablishing a more natural water flow at moderate flow velocity (<5 cm/sec). Guidelines for site selection from thousands of potential sites are proposed. Examples of suitable sites are given for intermediate marshes. These sites exhibit rapid deterioration following partial or complete hydrologic impoundment, implying a strong hydrologic, rather than sedimentological, cause of wetland deterioration. We used an exploratory hydrologic model to guide determination of the amount of spoil bank to be removed. The results from an economic model indicated a very effective cost-benefit ratio. Both models and practical experience with other types of restoration plans, in Louisiana and elsewhere, exhibit an economy of scale, wherein larger projects are more cost effective than smaller projects. However, in contrast to these other projects, spoil bank management may be 100 to 1000 times more cost effective and useful in wetland tracts <1000 ha in size. Modest spoil bank management at numerous small wetland sites appears to offer substantial positive attributes compared to alternative and more intensive management at a few larger wetland sites.     

Bird Use of Restoration and Reference Marshes Within the Barn Island Wildlife Management Area, Stonington, Connecticut, USA

/ Tidal marshes have been actively restored in Connecticut for nearly 20 years, but evaluations of these projects are typically based solely on observations of vegetation change. A formerly impounded valley marsh at the Barn Island Wildlife Management Area is a notable exception; previous research at this site has also included assessments of primary productivity, macroinvertebrates, and use by fishes. To determine the effects of marsh restoration on higher trophic levels, we monitored bird use at five sites within the Barn Island complex, including both restoration and reference marshes. Use by summer bird populations within fixed plots was monitored over two years at all sites. Our principal focus was Impoundment One, a previously impounded valley marsh reopened to full tidal exchange in 1982. This restoration site supported a greater abundance of wetland birds than our other sites, indicating that it is at least equivalent to reference marshes within the same system for this ecological function. Moreover, the species richness of birds and their frequency of occurrence at Impoundment One was greater than at 11 other estuarine marshes in southeastern Connecticut surveyed in a related investigation. A second marsh, under restoration for approximately ten years, appears to be developing in a similar fashion. These results complement previous studies on vegetation, macroinvertebrates, and fish use in this system to show that, over time, the reintroduction of tidal flooding can effectively restore important ecological functions to previously impounded tidal marshes.KEY WORDS: Estuarine; Tidal marsh; Wetland birds; Restoration     

Effects of Marsh Pond Terracing on Coastal Wintering Waterbirds Before and After Hurricane Rita

From February to March 2005–2006, we surveyed wintering waterbirds to test effects of terracing on coastal pond use before and after Hurricane Rita. Marsh terracing is intended to slow coastal marsh loss in the Chenier Plain by slowing marsh erosion and encouraging vegetation expansion. Terraces also increase marsh edge in ponds, possibly benefiting waterbirds. We monitored paired terraced and unterraced ponds in three sites within southwestern Louisiana’s Chenier Plain. Waterbirds were 75% more numerous in terraced than unterraced ponds. Waterbird richness was similar among ponds when corrected for number of individuals, suggesting terracing increased bird density but did not provide habitat unique from unterraced ponds. Birds were 93% more numerous following Hurricane Rita, mostly due to an influx of migrating waterfowl. Year round residents were similar in number before and after Hurricane Rita. Resident richness did not differ among years after correcting for number of observed individuals. Wading and dabbling foragers were more abundant in terraced ponds and these two guilds represented 74% of birds observed. We detected no difference among ponds for other guilds, i.e., probing, aerial, and diving foragers. Increasing proportion of mash edge increased bird density disproportionately: On average ponds with 10% edge had 6 birds observed and ponds with 30% edge had 16 birds observed. Terraces increased habitat interspersion and were an effective tool for increasing numbers of wintering waterfowl and wading birds. The extent to which terraces were sustainable following hurricane forces is unknown.     

Reducing the Effects of Dredged Material Levees on Coastal Marsh Function: Sediment Deposition and Nekton Utilization

Dredged material levees in coastal Louisiana are normally associated with pipeline canals or, more frequently, canals dredged through the wetlands to allow access to drilling locations for mineral extraction. The hydrologic impact on marshes behind the levee is of concern to coastal resource managers because of the potential impact on sediment transport and deposition, and the effect on estuarine organism access to valuable nursery habitat. This study examined the effects of gaps in dredged material levees, compared to continuous levees and natural channel banks, on these two aspects of marsh function. Field studies for sediment deposition were conducted biweekly for a year, and nekton samples were collected in spring and fall. Variation in nekton density among study arears and landscape types was great in part because of the inherent sampling gear issues and in part because of differences in characteristics among areas. Nekton densities were generally greater in natural compared to leveed and gapped landscapes. Differences in landscape type did not explain patterns in sediment deposition. The gaps examined appear to be too restrictive of marsh flooding to provide efficient movements of floodwaters onto the marsh during moderate flooding events. The “trapping” effect of the levees increases sediment deposition during extreme events. Gapping material levees may be an effective method of partially restoring upper marsh connection to nekton, but this method may work best in lower elevation marshes where nekton use is greater.     

Effects of Disturbance Associated With Seismic Exploration for Oil and Gas Reserves in Coastal Marshes

Anthropogenic disturbances in wetland ecosystems can alter the composition and structure of plant assemblages and affect system functions. Extensive oil and gas extraction has occurred in wetland habitats along the northern Gulf of Mexico coast since the early 1900s. Activities involved with three-dimensional (3D) seismic exploration for these resources cause various disturbances to vegetation and soils. We documented the impact of a 3D seismic survey in coastal marshes in Louisiana, USA, along transects established before exploration began. Two semi-impounded marshes dominated by Spartina patens were in the area surveyed. Vegetation, soil, and water physicochemical data were collected before the survey, about 6 weeks following its completion, and every 3 months thereafter for 2 years. Soil cores for seed bank emergence experiments were also collected. Maximum vegetation height at impact sites was reduced in both marshes 6 weeks following the survey. In one marsh, total vegetation cover was also reduced, and dead vegetation cover increased, at impact sites 6 weeks after the survey. These effects, however, did not persist 3 months later. No effects on soil or water properties were identified. The total number of seeds that germinated during greenhouse studies increased at impact sites 5 months following the survey in both marshes. Although some seed bank effects persisted 1 year, these effects were not reflected in standing vegetation. The marshes studied were therefore resilient to the impacts resulting from 3D seismic exploration because vegetation responses were short term in that they could not be identified a few months following survey completion.     

Calculating resource restoration for an oil discharge in Lake Barre,Louisiana, USA

Under the United States Oil Pollution Act of 1990, natural resource trustees are charged with assessing natural resource impacts due to an oil spill and determining the type and amount of natural resource restoration that will compensate the public for the impacts. Habitat equivalency analysis is a technique through which the impacts due to the spill and the benefits of restoration are quantified; both are quantified as habitat resources and associated ecological services. The goal of the analysis is to determine the amount of restoration such that the services lost are offset by services provided by restoration. In this paper, we first describe the habitat equivalency analysis framework. We then present an oil spill case from coastal Louisiana, USA, where the framework was applied to quantify resource impacts and determine the scale of restoration. In the Louisiana case, the trustees assessed impacts for oiled salt marsh and direct mortality to finfish, shellfish, and birds. The restoration project required planting salt-marsh vegetation in dredge material that was deposited on a barrier island. Using the habitat equivalency analysis framework, it was determined that 7.5 ha of the dredge platform should be planted as salt marsh. The planted hectares will benefit another 15.9 ha through vegetative spreading resulting in a total of 23.4 ha that will be enhanced or restored as compensation for the natural resource impacts.     

An evaluation of the impact of persistent water level changes on the areal extent of Georgian Bay/North Channel marshlands

A conceptual mathematical model has recently been devised to assist environmental managers in predicting the impact on coastal marsh areas of long-term changes in water levels. The model considers such impact solely in terms of the geometry of the confining basin, the change in ambient water level, and the maximum depth for which bottom-rooted emergent vegetation is present. This model is applied to 17 shoreline marshes of various shapes in the Georgian Bay/North Channel region of the Great Lakes.Model outputs of predicted maximum and minimum marsh area subsequent to changes in long-term levels are compared to marsh areas measured from available historical air photos dating from 1935 to 1985. The results of such comparisons indicate that such a geometric model, despite its neglect of the biological complexities of marsh ecology, can serve as a valuable tool for assessing the range of impacts of both natural and man-made changes in long-term ambient water levels on shoreline marshes.     

Land loss in coastal Louisiana (U.S.A.)

This paper examines causes and consequences of wetland losses in coastal Louisiana. Land loss is a cumulative impact, the result of many impacts both natural and artificial. Natural losses are caused by subsidence, decay of abandoned river deltas, waves, and storms. Artificial losses result from flood-control practices, impoundments, and dredging and subsequent erosion of artificial channels. Wetland loss also results from spoil disposal upon wetlands and land reclamation projects.Total land loss in Louisiana's coastal zone is at least 4,300 ha/year. Some wetlands are converted to spoil banks and other eco-systems so that wetland losses are probably two to three times higher. Annual wetland losses in the Barataria Bay basin are 2.6% of the wetland area. Human activities are the principal determinants of land loss. The present total wetland area directly lost because of canals may be close to 10% if spoil area is included. The interrelationship between hydrology, land, vegetation, substrate, subsidence, and sediment supply are complicated; however, hydrologic units with high canal density are generally associated with higher rates of land loss and the rate may be accelerating.Some cumulative impacts of land loss are increased saltwater intrusion, loss of capacity to buffer the impact of storms, and large additions of nutrients. One measure of the impact is that roughly $8–17 × 10⁶ (U.S.A.) of fisheries products and services are lost annually in Louisiana.Viewed at the level of the hydrologic unit, land loss transcends differences in local vegetation, substrate, geology, and hydrology. Land management should therefore focus at that level of organization. Proper guideline recommendations require an appreciation of the long-term interrelations of the wetland estuarine system.     

Effects of weir management on marsh loss,Marsh Island,Louisiana, USA

Weirs are low-level dams traditionally used in Louisiana's coastal marshes to improve habitat for ducks and furbearers. Currently, some workers hope that weirs may reduce marsh loss, whereas others fear that weirs may accelerate marsh loss. Parts of Marsh Island, Louisiana, have been weir-managed since 1958 to improve duck and furbearer habitat. Using aerial photographs, marsh loss that occurred between 1957 and 1983 in a 2922-ha weir-managed area was compared to that in a 2365-ha unmanaged area. Marsh loss was 0.38%/yr in the weir-managed area, and 0.35%/yr in the unmanaged area. Because marsh loss in the two areas differed less than 0.19%/yr, it was concluded that weirs did not affect marsh loss. The increase in open water between 1957 and 1983 did not result from the expansion of lakes or bayous. Rather, solid marsh converted to broken marsh, and the amount of vegetation within previously existing broken marsh decreased. Solid marsh farthest from large lakes and bayous, and adjacent to existing broken marsh, seemed more likely to break up. Marsh Island has few canals; therefore, marsh loss resulted primarily from natural processes. Weirs may have different effects under different hydrological conditions; additional studies are needed before generalizations regarding weirs and marsh loss can be made.     

Comparison of vesicular-arbuscular mycorrhizae in plants from disturbed and adjacent undisturbed regions of a coastal salt marsh in Clinton,Connecticut, USA

Roots of salt marsh plant speciesSpartina alterniflora, S. patens, Distichlis spicata, and others were examined for the presence of vesicular-arbuscular mycorrhizal (VAM) fungi. Samples were taken from introduced planted material in a salt marsh restoration project and from native material in adjacent marsh areas along the Indian River, Clinton, Connecticut, USA. After ten years the replanted area still has sites devoid of vegetation. The salt marsh plants introduced there were devoid of VAM fungi, while high marsh species from the adjacent undisturbed region showed consistent infection, leading the authors to suggest that VAM fungal infection of planting stocks may be a factor in the success of marsh restoration.     

Evaluation of tidal marsh restoration: Comparison of selected macroinvertebrate populations on a restored impounded valley marsh and an unimpounded valley marsh within the same salt marsh system in Connecticut,USA

Macroinvertebrates were examined on an impounded valley marsh in Stonington, Connecticut, that has changed from aTypha-dominated system to one with typical salt-marsh vegetation during 13 years following the reintroduction of tidal exchange. Animal populations on this restored impounded marsh were evaluated by comparing them with populations on a nearby unimpounded valley marsh of roughly the same size. Populations of the high marsh snail,Melampus bidentatus Say, were quantitatively sampled along transects that extended from the water-marsh edge to the upland; those of the ribbed mussel,Geukensia demissa Dillwyn, were sampled in low marsh areas on transects along the banks of creeks and mosquito ditches. The occurrence of other marsh invertebrates also was documented, but their abundance was not measured. The mean density ofMelampus was 332±39.6 SE/m² on the restored impounded marsh and 712±56.0 SE/m² on the unimpounded marsh. However, since snails were larger on the restored impounded marsh, the difference in snail biomass was less pronounced than the difference in snail density. MeanMelampus biomass was 4.96±0.52 SE g dry wt/m² on the restored impounded marsh and 6.96±0.52 SE g dry wt/m² on the unimpounded marsh. On the two marshes, snail density and biomass varied in relation to plant cover and other factors. The density and biomass ofGeukensia at the edge of the marsh were comparable on the restored impounded and unimpounded marshes. Mean mussel densities ranged from 80 to 240/m² and mean mussel biomass varied from 24.8–64.8 g dry wt/m² in different low marsh areas. In contrast, below the impoundment dike, meanGeukensia density was 1100±96.4 SE/m² and meanGeukensia biomass was 303.6±33.28 SE g dry wt/m². A consideration of all available evidence leads to the conclusion that the impounded marsh is in an advanced phase of restoration.     

Seed bank composition of open and overgrown calcareous grassland soils--a case study from Southern Belgium

The success of calcareous grassland recreation following abandonment depends to a large extent on the composition of the soil seed bank. We studied the species richness and composition of the seed bank along a chronosequence from well-developed calcareous grassland to scrub and forest vegetation, which had developed on calcareous grassland over the last 225 years. The seed bank density was highest in the calcareous grassland soils (930 seeds/m(2)). However, the seed bank was mainly composed of rather common species of nutrient poor grassland, which were poorly represented in the surface vegetation. There were no significant differences in soil seed bank density and species composition between the calcareous grassland and scrub vegetation that was less than 15 years old, largely because several grassland species had persisted in the scrub vegetation and were therefore able to replenish the soil seed bank. In contrast seed density and species richness declined significantly after 40 years of grassland abandonment. Indeed, forest soils had the lowest seed densities (214 seeds/m(2)) with only a few grassland species represented. This reflects the lack of grassland species in the field layer of the forest, and, therefore, the lack of seed production and seed bank replenishment. It is clear that recreation of calcareous grassland on long abandoned sites cannot rely on germination of target species from the seed bank alone. Even in the calcareous grassland soil, seeds of target species are not abundant as the majority produces transient seeds, which decay rapidly if they do not germinate immediately. Successful grassland recreation on such sites therefore may require seeds of target species to be introduced artificially (e.g. as seed mixtures, green hay, etc.). Alternatively, reinstatement of traditional practices, such as grazing or mowing, will increase the natural dispersal potential of these species, allowing population reestablishment in the long term.     

Recovery Trajectories After In Situ Burning of an Oiled Wetland in Coastal Louisiana,USA

The high degree of physical disturbance associated with conventional response options to oil spills in wetlands is driving the investigation of alternative cleanup methodologies. In March 1995, a spill of gas condensate in a brackish marsh at Rockefeller Wildlife Refuge in southwestern Louisiana was remediated through the use of in situ burning. An assessment of vegetation recovery was initiated in three treatment marshes: (1) oil-impacted and burned, (2) oil impacted and unburned, and (3) a nonoiled unburned reference. We compared percent cover, stem density, and biomass in the treatment marshes to define ecological recovery of the marsh vegetation and soil hydrocarbon content to determine the efficacy of in situ burning as a cleanup technique. Burning led to a rapid decrease in soil hydrocarbon concentrations in the impacted-and-burned marsh to background levels by the end of the first growing season. Although a management fire accidentally burned the oil-impacted-and-unburned and reference marshes in December 1995, stem density, live biomass, and total percent cover values in the oil-impacted-and-burned marsh were equivalent to those in the other treatment marshes after three years. In addition, plant community composition within the oil-impacted-and-burned marsh was similar to the codominant mix of the grasses Distichlis spicata (salt grass) and Spartina patens (wire grass) characteristic of the surrounding marsh after the same time period. Rapid recovery of the oil-impacted-and-unburned marsh was likely due to lower initial hydrocarbon exposure. Water levels inundating the soil surface of this grass-dominated marsh and the timing of the in situ burn early in the growing season were important factors contributing to the rapid recovery of this wetland. The results of this in situ burn evaluation support the conclusion that burning, under the proper conditions, can be relied upon as an effective cleanup response to hydrocarbon spills in herbaceous wetlands.     

Modeling hydrology and eutrophication in a Louisiana swamp forest ecosystem

The EPA Storm Water Management Model (1971) was used to model hydrodynamics, nutrient dynamics, and eutrophication in a Louisiana swamp forest ecosystem. The present system of canals and spoil banks in the swamp causes impoundment of swamp areas and does not optimize discharge from the swamp forest. Simulations showed that hydraulics could be managed to increase discharge rates to the lower estuary (22 percent), to increase productivity of the swamp forest (100 percent), and to decrease lake eutrophication (43 percent). This could be done by removing spoil banks in the swamp and allowing upland runoff to pass through the backswamp.     

Lead transport into Bayou Trepagnier wetlands in Louisiana, USA

Establishment of a petroleum refinery in 1916 near the headwaters of Bayou Trepagnier with subsequent dredging of the bayou resulted in spoil banks containing high levels of Pb. A large swamp abuts the eastern bank of the bayou. Cores were taken from 15 baldcypress [Taxodium distichum (L.) Richard] trees growing in the swamp along a 610-m transect (nine trees) and a 183-m transect (six trees) running perpendicular from the spoil bank. The cores were crossdated, annual rings were measured, and 5-yr segments of the cores were prepared and analyzed for heavy metals. Soil samples were collected along one transect and analyzed for metals. Levels of Pb in Bayou Trepagnier swamp trees were compared to levels in nine baldcypress trees growing along Stinking Bayou, a reference area. During the last 100 yr, Pb in growth rings of swamp baldcypress trees averaged 8.6 mg/kg (SD 4.88) along one transect and 7.9 mg/kg (SD 5.39) along the other. Lead in the soil along the first transect dropped from >2700 mg/kg (spoil bank) to 10 mg/kg at 420 m into the swamp. Baldcypress trees growing near the refinery on the spoil bank along Bayou Trepagnier (covered in an earlier study) averaged 4.5 mg/kg Pb, and trees along Stinking Bayou averaged 2.1 mg/kg. Trees in the swamp soil with 10 to 425 mg/kg Pb concentrated much more Pb than trees growing on the heavily polluted bank. Greater uptake of Pb by trees in the swamp is discussed in terms of soil dynamics and Pb sources.