Sediment transport and deposition along a stream in an agricultural basin (65 km²) in northern Missouri, USA were quantified as part of a long-term study to evaluate effects of silvicultural practices on the hydrology, sediment, vegetation, and wildlife characteristics of remaining forested riparian systems. Median cumulative sediment deposition, measured using feldspar clay pads, increased from August 1995 to August 1998 at a rate of about 1 cm/yr. Median deposition amounts from single floods ranged from 0.03 cm to 0.64 cm. Floodplain and riparian maintenance flows corresponded to monitored floods with calculated recurrence intervals as low as <2 years. Simple linear regression models, using flood event suspended-sediment load or streamflow characteristics, explained up to 82 percent of variability in median event sediment deposition on the floodplain clay pads. There was little apparent correlation between cumulative short-term deposition and site elevation, distance from channel, longitudinal distance, or fluvial landform type. This may be due to upstream channelization, floodplain complexity, short duration of events, or sediment-load characteristics of low-recurrence interval floods (<2 to 5 years) sampled in this study. Dendrogeomorphic measurements indicated a substantial increase in the mean rate of deposition on the Long Branch Creek floodplain from about 1950 through 1980. Eighty-nine percent of the clay pad monitoring sites and all dendrogeomorphic monitoring sites experienced net positive deposition emphasizing the role of this riparian area as a net sediment storage site.

Previous research shows an anomalous absence of plains cottonwood (Populus deltoides) woodlands along a 30 km sandy braid-channel reach of the Milk River upstream of the Fresno Dam in northern Montana, USA. This absence contrasts to well-wooded, meandering reaches upriver and downriver of the braid-plain. Our field measurements show that mechanical ice breakups and ice floes physically damage most plains cottonwood saplings and small trees along the braid-plain. Between 1911 and 1998, mechanical ice breakups occurred in 50% of the years of record. Of these, 12% were greater than bankfull discharge and followed dramatic temperature increases and subsequent snowpack melting in the headwaters. We suggest that the absence of cottonwood woodlands on the braid-plain is associated with low bank heights, wide channels, and low channel sinuosity in the braiding reach through which overbank river ice moves more easily compared to the meandering reaches. Ice blocks rafted onto the braid-plain have repeatedly scarred, bent, broken (trunks partially detached, roots not displaced), sheared (trunk detached, roots partially displaced), and toppled (trunks and roots displaced) cottonwoods, causing an absence of middle- to old-aged trees. Even though hundreds of new stem and stump shoots have sprouted since the last major ice damage (1996), very few trees are present on the braid-plain compared to the meandering reaches. Almost all (94–100%) cottonwoods on the braid-plain were damaged by ice in 1996 (but only 23% in the meandering reaches). All damaged cottonwoods were scarred; up to 42% of the ice-damaged trees on the braid-plain died from shearing and toppling. Wetlands in the U.S. and Canada are defined on the basis of frequency of flooding and presence of water-logged soils that support hydrophytic plants. Ice plays a critical role in limiting plant distribution in northern riparian wetlands and should be incorporated into revisions of existing classification systems.

Throughout the western states, riparian ecosystems have been affected by water diversions or spring and seep developments that decrease the quantity of instream flows and result in lowered water-tables. Water extraction is especially damaging in arid and semi-arid regions where the presence of instream and ground-water flows are crucial to riparian vegetation. We examined the temporal and spatial relationships between hydrologic gradients, vegetation, and soils in two central Nevada riparian meadows in order to identify plant species and environmental variables that can serve as indicators of water-table status. Species frequency and aerial cover, ground-cover composition, depth to water-table, and soil morphological and physical properties were measured along hydrologic gradients within two riparian meadow complexes. TWINSPAN, cluster, and multivariate discriminant analyses classified the vegetation into four ecosystem types. These occurred along the hydrologic gradient and included, from wettest to driest, wet meadow, mesic meadow, dry meadow, and basin big sagebrush meadow types. Canonical correspondence analysis (CCA) indicated that the variables most strongly related to plant species frequency within both meadows were those associated with depth to water-table. Integrative variables, including the number of days that depth to water-table was less than 30 cm and less than 70 cm, and degree-days of anaerobiosis were most closely related to the wet and mesic meadow vegetation types. The range in depth to water-table, elevation, and aerial cover of gravel and litter were all related to the dry and sagebrush meadow vegetation types. Indicator species associated with particular water-table regimes were identified for each vegetation type based on two-way ordered tables. Carex nebrascensis, an obligate riparian species, occurred at water-tables of 0–30 cm below the surface and was the most reliable indicator of shallow water-tables. Large temporal and spatial variability in water-table depths for the mesic and dry meadow types suggests that species associated with these types could be used only to indicate broad ranges in water-table depth. Integrative environmental variables that incorporated the temporal variation in water-tables (i.e., days that depth to water-table was less than 30 and 70 cm; degree days of anaerobiosis; range in water-table depth during the growing season) demonstrated closer relationships to the vegetation types than water-table alone. They were also more sensitive to the spatial and temporal differences in water-tables than individual plant species or vegetation types. Environmental and integrative environmental variables may respond more quickly to changes in local hydrology than plant species and are possibly more sensitive indicators of both current water-table status and potential vegetation.

Seven wetlands from two dune systems on the Swan Coastal Plain, Western Australia were studied for sediment properties associated with phosphorus retention and potential remobilization. Sediments from the Spearwood Dunes had higher contents of silt, clay, and organic material. Sediment total phosphorus was positively correlated with water content, organic matter and silt content but inversely related to the amount of coarse materials. Part of the organic matter was refractory humic substances, and the molecular weight of humic acids ranged from less than 10,000 to over 600,000. Significant portions of sediment total P (5–73%) were bound to humic substances. The fraction of “loosely-bound phosphorus” (NH₄Cl-P) accounted for less than 10% of total-P. NaOH-extractable P was high; some humic substances were extracted with NaOH. Correlations between phosphorus forms and other components suggest an affinity between ‘fine particles’ of similar hydrologic density during sediment transport and reworking, and this involves most humic-P in these wetlands.

The common reed, Phragmites australis, has spread throughout Gulf and Atlantic Coast marshes of the U.S. in the past thirty years. In the Chesapeake Bay area, natural resource managers are uncertain as to the current distribution or recent colonization rate of this low-salinity marsh grass. A geographic information system (GIS) was used to determine the distribution and expansion rate of P. australis within four brackish and three tidal freshwater marshes in the upper Chesapeake Bay region. Vegetation patterns were mapped by interpreting aerial photographs from the 1930s, 1970s, 1980s, and 1990s. The aerial extent of P. australis stands was measured by digitizing vegetation boundaries, correcting for distortion, and analyzing the data using a GIS. A geometric growth formula documented an intrinsic rate of increase for each P. australis stand. In addition, aboveground biomass was sampled from monotypic stands of P. australis. Results show that P. australis is present in all seven marshes; however, it is most pervasive in the three tidal freshwater marshes. Based on interpretations of photographs, P. australis was present prior to 1938 in these three marshes. In each successive time period, there was a net increase of P. australis in all seven marshes that were examined. In the past ten years, however, the rate of increase has declined or stabilized in each marsh in which P. australis was well-established prior to 1985. This slowdown is especially prevalent in the three tidal freshwater marshes. In contrast, the three brackish marshes most recently colonized by P. australis showed high annual intrinsic rates of increase (0.06 to 0.19 yr⁻¹ or more). Also, the highest biomass estimates were found in two of these three marshes.

Cogongrass (Imperata cylindrica), an invasive perennial introduced from Southeast Asia, is currently spreading throughout the southeastern United States from Florida to Louisiana. In the U.S., cogongrass is generally not considered a wetland species, although it's range is expanding in regions with high wetland abundance. The objective of this study was to determine if excessive soil moisture might prevent cogongrass from establishing in areas with seasonally flooded soils. In one greenhouse experiment, we examined cogongrass germination and seedling growth in soils that were freely drained, saturated, and inundated. We performed a second greenhouse experiment to evaluate growth and survival of cogongrass seedlings of four different size classes in five soil moisture treatments ranging from dry to inundated. Cogongrass germination was lowest when seeds were overtopped with water. There were no differences in germination between saturated and freely drained treatments; however, seedlings grew largest in freely drained soil and were smallest when immersed. In our second experiment, most cogongrass plants survived except when given no water, but growth differed by watering treatment depending on seedling size. Increasing moisture was more detrimental to the growth of small seedlings compared to the growth of larger cogongrass plants. Overall, cogongrass was most sensitive to soil inundation in the earliest stages of establishment; thus, excessive moisture conditions in the spring, during early seedling development, could restrict invasion of cogongrass by seed. Once cogongrass is established, however, its tolerance of flooding appears to increase.

We excluded predatory fish from a marsh weedbed to evaluate experimentally their impact on invertebrate prey. Gut analyses of wetland fish, including pumpkinseed sunfish (Lepomis gibbosus), brown bullhead (Ictalurus nebulosus), black crappie (Pomoxis nigromaculatus), and common carp (Cyprinus carpio), revealed that large numbers of midge larvae (Diptera: Chironomidae) were consumed. However, our exclusion of these predatory fish from study habitats did not result in midge population increases. On the contrary, fewer epiphytic midges occurred where predatory fish had been excluded (P = 0.0043). Populations of midge competitors (especially Planorbidae and Physidae) and invertebrate midge predators (especially Corixidae and Glossiphoniidae) were suppressed directly by fish, and midges that co-existed with fish apparently benefitted indirectly from those interactions. For epiphytic midge larvae, the negative direct influence of fish predation was strong, but positive indirect effects apparently were even more powerful.

We examined larval anuran assemblages at 12 temporary wetlands occurring on the MacArthur Agro-Ecology Research center (MAERC) in southcentral Florida. MAERC is an active cattle ranch, and the wetlands on the site are heavily influenced by an extensive series of ditches that drain the landscape. Ditching has resulted in a change from a historically extensive marsh system to a series of isolated wetlands surrounded by upland habitats. Because a majority of anurans in Florida breed exclusively or facultatively in wetlands whose drying regime excludes fish, we were interested in determining the value of these modified wetlands as breeding sites. We examined the effect of wetland size and hydrology on anuran use, and compared breeding activity across three summers that varied greatly in rainfall pattern. We sampled tadpoles from May 93 to August 93 and from May 94 to September 95. A total of 3678 tadpoles from 11 species was collected. Rana utricularia was the most abundant species and the only species found in every wetland. Species richness was related positively to wetland size (r = 0.65, p = 0.023) but not hydroperiod (r = 0.03, p = 0.93). Tadpole abundance was not related to wetland size (r = 0.35, p = 0.29) nor hydroperiod (r = 0.40, p = 0.22). Annual variation in rainfall resulted in significant changes in species composition. A drought during 1993 resulted in no breeding. A high water table in the spring of 1995 resulted in localized flooding in early summer on part of the ranch. Wetlands in these areas were exposed to spillover of water from ditches containing fishes. Wetlands so impacted showed significant changes in species composition from the previous year (x² = 1008, p < 0.0001), whereas wetlands that were not impacted did not differ in composition. The wetlands at MAERC provide dynamic habitats that offer varying breeding opportunities that are highly dependent on meteorological conditions.

Soil accretion, sediment deposition, and nutrient (N, P, organic C) accumulation were compared in floodplain and depressional freshwater wetlands of southwestern Georgia, USA to evaluate the role of riverine (2600 km² catchment) versus depressional (<10 km² catchment) wetlands as sinks for sediment and nutrients. Soil cores were collected from three floodplain (cypress-gum) and nine depressional (three each from cypress-gum forest, cypress-savannah, and herbaceous marsh) wetlands and analyzed for radionuclides (¹³⁷Cs, ²¹⁰Pb), bulk density, N, P, and organic C to quantify recent (30-year) and long-term (100-year) rates of sediment and nutrient accumulation. There was no significant difference in organic C, N, or sediment accumulation between depressional and floodplain wetlands. However, P accumulation was 1.5 to three times higher in the floodplain (0.12–0.75 g/m²/yr) than in the depressional wetlands (0.08–0.25 g/m²/yr). Sediment and nutrient accumulations were highly variable among depressional wetland types, more so than between depressional and floodplain wetlands. This variability likely is the result of differences in historical land use, hydrology, vegetation type, NPP, and perhaps fire frequency. Mean (n = 12) one-hundred-year rates of sediment deposition (1036 g/m²/yr), organic C (79 g/m²/yr), N (6.0 g/m²/yr), and P accumulation (0.38 g/m²/yr) were much higher than 30-year rates (sediment = 118 g/m²/yr, C = 20 g/m²/yr, N = 1.5 g/m²/yr, P = 0.09 g/m²/yr). Higher 100-year (²¹⁰Pb) sediment and nutrient accumulations likely reflect the greater numbers of farms, greater grazing by livestock, and the absence of environmentally sound agricultural practices in southwestern Georgia at the turn of the century. Our findings suggest that the degree of anthropogenic disturbance within the surrounding watershed regulates wetland sediment, organic C, and N accumulation. Phosphorus accumulation also is greater in floodplain wetlands that have large catchments containing fine textured (clay) sediments that are co-deposited with P.

Isotopic analyses of oxygen and hydrogen of water (δ¹⁸O and δD) and nitrogen and oxygen of nitrate (δ¹⁵N and δ¹⁸O) are used in conjunction with conventional water chemistry and hydrologic measurements to investigate water flow and nitrogen cycling mechanisms through two riparian zones adjacent to upland agricultural fields. Within the Kankakee watershed of northwestern Indiana, a native riverine wetland was compared to a constructed riverine wetland to assess the wetland restoration in terms of water flow and nitrate attenuation mechanisms and efficiency. Conditions in the constructed wetland are controlled by a system of individual cells separated by dikes and levees and into which water is periodically pumped, while the native wetland occupies an area of remnant river meanders or oxbows. Analyses of samples taken from well transects across both wetlands suggest that water flow across the constructed wetland has been greatly altered. Nitrate cycling characteristics show significant differences between the two wetlands and particularly, nitrate attenuation efficiency is greatly reduced in the constructed wetland.

We used paleoecological methods to infer environmental conditions in Water Conservation Area 2A (WCA-2A) of the Everglades before impoundment in 1961, and we compared pre-impoundment environmental conditions to present conditions. Abundances of siliceous microfossils (diatoms, chrysophyte cysts, sponge gemmoscleres, and plant phytoliths) were analyzed in two sediment cores from nutrient-enriched northern WCA-2A and in two sediment cores from unenriched southern WCA-2A. Nutrient enrichment in northern WCA-2A after 1961 was associated with an increase in relative abundance of eutrophic diatoms. A pH increase in much of WCA-2A after 1961 was indicated by a decrease in relative abundance of acidic indicators (Eunotia diatoms and the sponge Anheteromeyenia ryderi) and a decrease in richness of chrysophyte cysts. An increase in anoxia in nutrient-enriched northern WCA-2A, during about 1961–1980 when high water depth was maintained, was suggested by an increase in visible dissolution of siliceous microfossils. A decrease in palm phytolith abundance after 1961 and lack of recovery after water depth was lowered suggest either that water depth is still maintained high enough to adversely affect vegetation or that there has been insufficient time for recovery. Numbers of sponge gemmoscleres decreased to less than 1% of pre-impoundment numbers, suggesting that there has been a dramatic decrease in sponge abundance. The environmental changes discovered in this study suggest directions for future management of the Everglades.

For two years, we measured soil moisture, pH, salinity, and ion concentrations bimonthly from 55 vegetation plots in six Mediterranean salt marshes of SE Spain. Edaphic characteristics during dry and wet seasons were compared within six selected plant communities. The dominant species in each of these communities were Suaeda vera, Lygeum spartum, Limonium sp, Sarcocornia fruticosa, Halocnemum strobilaceum, and Arthrocnemum macrostachyum. Although soil salt concentrations were lower during the wet season, different patterns of temporal variation in total dissolved salts, sodium adsorption ratio, Cl⁻, Na , K , Ca² , and Mg² were observed in the soils occupied by different plant communities. The variation patterns for SO₄²⁻ were the same in all plant communities. The Ca² /Mg² , Ca² /Na , and SO₄²⁻/Cl⁻ ratios increased during the wet season because of the more pronounced decrease in Cl⁻, Na , and Mg² concentrations relative to Ca² , and SO₄²⁻. Canonical correspondence analysis (CCA) related the species distribution with certain soil conditions. The edaphic variables that best explained the data were maximum soil moisture, mean K /Na ratio, mean Ca² /Mg² ratio, maximum pH, and maximum K . Two different environments were identified: dry salt marsh and wet salt marsh. To examine soil ionic gradients in these two environments, two further CCA were applied. The variables that best explained soil-vegetation relationships in the dry salt marsh were mean K /Na ratio, mean Ca² /Mg² ratio, maximum K , and minimum Ca² /Na ratio. Some representative dry salt marsh species were Lygeum spartum, Atriplex glauca, Suaeda vera, and Frankenia corymbosa. The variables explaining the data set in the wet salt marsh were maximum sodium adsorption ratio, maximum Mg² content, mean Ca² content, and mean Ca² /Mg² ratio. Species such as Arthrocnemum macrostachyum, Sarcocornia fruticosa, Juncus maritimus, and Tamarix boveana were representative of this environment.

When wetland restoration includes re-establishing native plant taxa as an objective, an understanding of the variables driving the development of plant communities is necessary. With this in mind, we examined soil and physiographic characteristics of depressional wetlands of three vegetation types (cypress-gum swamps, cypress savannas, and grass-sedge marshes) located in a fire-maintained longleaf pine ecosystem in southwestern Georgia, USA. Our objective was to establish whether plant community development in these wetlands is controlled primarily by hydrogeomorphic features or by different disturbance histories. We did not identify physical features that uniquely separate the wetland vegetation types. Instead, we observed a range of topo-edaphic conditions that likely drive variations in hydrologic regimes, which in turn, are probable influences on fire regime. We propose that several long-term successional trajectories may be initiated in the prolonged absence of fire, altered hydrology, or both, which link the distinctive vegetation types. Thus, a range of vegetation types may be suitable as potential restoration goals for these depressional wetlands. We suggest that the opportunities or constraints for use of prescribed fire in the long-term management of restored wetlands and adjacent uplands should be a significant consideration in the development of restoration strategies targeting specific plant communities.

Pollinator activity, flowering phenology, plant size, and seed set of an endangered annual plant, Pogogyne abramsii, were compared in natural and created vernal pools on Del Mar Mesa, San Diego County, California, USA. The purpose was to test the hypothesis that an artificial habitat would have fewer or less effective pollinator visits than natural vernal pools and that, as a result, the plants would have lower fecundity. The effect of differences in flowering phenology and plant size among pool types on pollinator visitation and seed set was also evaluated. The most frequent insect visitors to P. abramsii flowers were the Eurasian honey bee, Apis mellifera, two anthophorid bees, Exomalopsis nitens and E. torticornis, and three species of bee flies, of which Bombylius facialis was by far the most abundant. All species of insect visitors displayed density-dependent foraging on P. abramsii, with greater numbers of visits per unit area where flower density was greater. Nevertheless, visits per flower were negatively correlated with flower density. Flowering phenology of P. abramsii was delayed in most created compared to natural vernal pools, which affected the frequency by species of insect visitors to created compared with natural vernal pools. Peak flower densities were significantly lower in created than in natural vernal pools. The created pools had more visits per flower. Plants were smaller in created pools compared to natural pools, and seed production was positively correlated to plant size. Seed set in P. abramsii was greater than one per plant in both created and natural pools but was significantly lower in created vernal pools. Most of this difference could be attributed to the larger size of plants in the natural pools. We conclude that pollinator limitation should not preclude the sparser populations of smaller plants in the created pools from having a positive growth rate. Subsequent establishment and persistence of populations in a majority of created basins on this site confirms this assertion.

We determined the life history, secondary production, and trophic basis of production for Callibaetis fluctuans in a mitigated wetland at the Green Bottom Wildlife Management Area, Cabell County, West Virginia, USA. Quantitative benthic samples were collected from three sites covering a range of water depths and vegetation types from January to December 1994. Length-frequency histograms were constructed for each sample date, and secondary production was calculated using the size-frequency method. Foregut analyses were conducted seasonally to determine diet. Callibaetis fluctuans had a multivoltine life cycle with at least three cohorts. Fast larval growth occurred during the spring and fall, whereas an overwintering cohort had slower growth. Total annual production was 208.0 mg·m⁻²·yr⁻¹ (AFDM) with a production-to-biomass ratio of 12.6. Amorphous detritus comprised >80% of the diet in all seasons and accounted for 70% of total production. To account for this production, C. fluctuans consumed 3675 mg·m⁻²·yr⁻¹ of detritus. This is one of the first studies to provide production estimates for an aquatic insect in a permanently inundated wetland.

Invertebrate assemblages were studied in eight monoculture wetland mesocosms constructed for wastewater treatment. Low concentrations of dissolved oxygen (D.O.) were measured in bulrush mesocosms while higher concentrations of D.O. were measured in open water mesocosms containing submerged pondweeds. Invertebrate taxa richness was positively related to D.O. concentrations that were, in turn, related to vegetation communities. Reference wetland sites contained a variety of plant species along with extensive open water areas. Invertebrate taxa richness was greater at reference sites than in any wastewater mesocosm. Invertebrate samples from the wastewater mesocosms and reference sites were analyzed for five trace elements. While the concentrations of aluminum, arsenic, mercury, and silver were below values harmful to wildlife, the concentrations of selenium reached levels of moderate concern on one occasion. Data from this study suggest that selenium bioaccumulation by invertebrates may be related to the type of vegetation community or detrital habitat type. Wetlands designed for invertebrate production for waterfowl should take into account the potential for low D.O. concentrations and trace element bioaccumulation associated with vegetation community types.

We studied phenol oxidase (PO) activity in two Sphagnum-dominated peatlands and a Carex-dominated freshwater marsh during a strong summer drought in Central New York, USA to determine whether PO activity might respond to expected climatic changes. Peat was sampled at different depths and within distinct vegetation types within the marsh. Carex-derived peat supported substantially higher PO activity (average = 0.030, range = 0.011–0.051 μMOL diqc min⁻¹ mg dry peat⁻¹, at soil pH 5.5) than Sphagnum peat (average = 0.006, range = 0.001–0.015 μMOL diqc min ⁻¹ mg dry peat⁻¹, at soil pH 3.8). In both peat types, PO activity showed a strong exponential increase with increased solution pH. Phenol oxidase activity in Sphagnum peat did not vary significantly during the drought, suggesting that PO activity may be constrained by low pH and enzyme inhibitors. Conversely, PO activity in the marsh peat varied with peat type and sample date but not as a consistent function of water-table depth. As a result, PO activity in Sphagnum peat appears to be regulated less by aeration and more by pH and possibly enzyme inhibitors. When pH is favorable, PO activity depends more on wetland vegetation type and botanical composition of the peat than climatic factors.

Currently, no method exists to core large volumes of inland freshwater wetland soils that maintains stratigraphic integrity, minimizes unnecessary disturbance, and cores up to a depth of 50 cm. Our objective was to create a large-volume soil coring device that could be applied with consistency to a variety of wetland substrates. The result is a hand-operated soil corer that resembles the aluminum irrigation pipe corer that DeLaune et al. (1978) used to core soft marshy substrates. Instead of aluminum pipe, we used regular steel stovepipe and a variety of tools for insertion. After the sample is extracted from the sediment, the handle can be quickly removed for ease of transportation and storage of a core. The stovepipe can be cut open to expose the soil sample so that it can easily be sectioned for incremental analysis. The corer was used to take 130 samples in 18 different sites, spanning many different wetland substrate types. Our method has many applications, including radiochronologic dating, seed bank analysis, bulk density measurement, and soil contaminants analysis.