Charles Andrew Cole

A comparison of created and natural wetlands in Pennsylvania, USA

Recent research suggests that created wetlands do not look, or function, like the natural systems they are intended to replace. Proper planning, construction, and the introduction of appropriate biotic material should initiate natural processes which continue indefinitely in a successful wetland creation project, with minimal human input. To determine if differences existed between created and natural wetlands, we compared soil matrix chroma, organic matter content, rock fragment content, bulk density, particle size distribution, vegetation species richness, total plant cover, and average wetland indicator status in created (n = 12) and natural (n = 14)wetlands in Pennsylvania (USA). Created wetlands ranged in age from two to 18 years. Soils in created wetlands had less organic matter content, greater bulk densities, higher matrix chroma, and more rock fragments than reference wetlands. Soils in reference wetlands had clay loam textures with high silt content, while sandy clay loam textures predominated in the created sites. Vegetation species richness and total cover were both greater in natural reference wetlands. Vegetation in created wetlands included a greater proportion of upland species than found in the reference wetlands. There were significant differences in soils and vegetation characteristics between younger and older created wetlands, though we could not say older created sites were trending towards the reference wetland condition. Updated site selection practices, more careful consideration of monitoring period lengths, and, especially, a stronger effort to recreate wetland types native to the region should result in increased similarity between created and natural wetlands.

Wetland hydrology as a function of hydrogeomorphic (HGM) subclass

Characterizing wetland hydrology is key to assessing relative function over a range of wetland types. However, hydrologic data are often lacking. To address this lack of information, we categorized a set of 24 reference wetlands by hydrogeomorphic (HGM) subclass from 1993 to 1995, installed monitoring wells and piezometers, and assessed local water-table levels, pH, and specific conductance by month. Four HGM wetland subclasses were common to central Pennsylvania (riparian depression (n=8), slope (n=7), mainstem floodplain (n=5), and headwater floodplain (n=4)) and formed the basis for our analysis. Median depth to water in the wells differed by HGM subclass. Riparian depressions had the shallowest depth to water (−8 cm) and headwater floodplain wetlands the greatest (−70 cm). Comparisons of the percent occurrence of a piezometric head (from comparisons between paired piezometer and slotted wells) indicated that riparian depressions and slopes had significant ground-water inputs (47 and 48%, respectively), whereas the mainstem floodplain (31%) and headwater floodplain wetlands (23%) were more surface-water-driven systems. Water occurred within the root zone (30 cm) most often for riparian depressions (80% of observations), intermediate for slopes (48%), and least for mainstem floodplains (17%) and headwater floodplains (6%). Headwater floodplain wetlands were never inundated by overbank flow during this study but instead received water from snowmelt and overland flow after rain events. Mainstem floodplain wetlands were inundated by floods during major storm events. The upper 30 cm of soil (i.e., the root zone of plants) was almost continually saturated in riparian depressions, but rarely for both floodplain systems. Slope wetlands were intermediate between riparian depressions and floodplain systems in the amount of time water was present within 30 cm of the ground surface. Riparian depressions and slopes had lower pH than floodplain systems, and pH did not vary significantly by month for any HGM subclass. Floodplain systems (both headwater and mainstem) had greater values of specific conductance than either riparian depressions or slopes; riparian depressions were the only HGM subclass to show seasonality in specific conductance. Factors other than HGM subclass that may have influenced the hydrologic pattern and water quality parameters included bedrock geology, disturbance levels, and watershed attributes.

A comparison of the hydrologic characteristics of natural and created mainstem floodplain wetlands in Pennsylvania

Abstract An understanding of wetland hydrology is critical for the assessment of the success of created wetlands relative to natural wetlands. Our objective was to determine if hydrologic characteristics differed between natural and created mainstem floodplain wetlands. We measured water depth every 6 h, between November 1, 1996 and August 31, 1998, and determined median depth to the water table, range of water table fluctuation, percent of time water was within the root zone (30 cm), and the number and duration of periods where water remained in the root zone. The natural wetlands were different from the created wetlands as median depth to water was generally deeper, there were shorter periods where soils were saturated or inundated, and there was a lower percentage of time where water was in the root zone. The created wetlands, in effect, were wetter and for longer periods. The created wetlands had a large component of open water at each site. Most naturally occurring mainstem floodplain wetlands in central Pennsylvania are vegetated with very little open water. We suggest that in the rush to make sure there is some water in mitigation wetlands we have gone too far in keeping sites inundated. In reality, many natural wetlands are merely saturated, or much drier.

Plants as indicators of wetland water source.

At the local scale, plant species distribution is determined primarily by the environmental characteristics of a site. In a wetland, water chemistry and hydroperiod are two of the most important of these environmental characteristics. Both are functions of water source. In central Pennsylvania, groundwater input tends to be continuous, while surface water may be permanent or seasonal. The chemistry of groundwater and surface water differs since groundwater is influenced by the substrate through which it flows. Because of these differences, and because of their effects on plant species distribution, it is possible to use vegetation as an indicator of the dominant water source of a site. Plots within 28 wetlands in central Pennsylvania were sampled, and the plots were classified by water source. The three hydrologic categories were groundwater, seasonal surface water, and permanent surface water. The core of the study was the analysis of half of the plots to identify species that were associated with a particular water source. Several groups of indicator species were identified. Some species, including Nyssa sylvatica, were strongly associated with the presence of groundwater. Others, such as Symplocarpus foetidus, were strongly associated with the presence of seasonal surface water. Several aquatic species were associated with permanent surface water. The remainder of the plots were used to test the predictive ability of the indicator species identified. The vegetation of a wetland plot predicted its hydrologic category with 72% accuracy. The identification of more indicator species could lead to the development of a useful tool for wetland research and management, since monitoring hydrology is often both expensive and time-consuming.

Assessing the relationship between biomass and soil organic matter in created wetlands of central Pennsylvania, USA

Abstract Created wetlands are frequently structurally different from the natural wetlands they are intended to replace. With differences in structure might come differences in function. Most created wetlands in central Pennsylvania have very low amounts of soil organic matter relative to levels found in natural wetlands. However, anecdotal evidence also suggests that plant production is equivalent in created wetlands to natural wetlands. There is little evidence to indicate that this plant biomass in created wetlands is finding its way into the soil as organic matter. This might translate into a lack of function in the mitigation wetlands. To address this issue, we studied plant biomass production in seven created wetlands in central Pennsylvania (USA). We measured above- and below-ground biomass and compared results with known values of soil organic matter and hydrology for the same wetlands. We found biomass to be approximately equivalent to that produced in natural freshwater marshes, although the below-ground component was somewhat higher. We found no relationship of biomass to soil organic matter, even though site conditions were wet enough to retard plant decomposition.

PATTERNS OF WETLAND HYDROLOGY IN THE RIDGE AND VALLEY PROVINCE, PENNSYLVANIA, USA

Developing an understanding of wetland hydrology that is free from site-specific constraints is difficult. Many hydrologic studies are focused upon a single site and the development of water-budget components. Our previous research (Cole et al. 1997) examined the hydrology of several wetlands based upon monthly sampling during the growing season. Those data did not provide adequate information on moisture regimes and did not tell us enough about year-round hydrodynamics. Our new objective was to expand hydrologic analyses to a larger proportion of our reference wetlands and extend them over a longer period of time. We continued to organize our wetlands and our analyses around hydrogeomorphic (HGM) principles. We found ground-water-dominated wetlands (riparian depressions and slopes) to be the wettest sites. Surface-water systems (headwater and mainstem floodplain wetlands) were drier. We found little difference between slopes and the floodplain wetlands in the amount of time water was within the root zone. Riparian depressions were wetter longer, as the average duration of water within the root zone was almost a year for riparian depressions and much less for all other wetland types. Disturbance seemed to play a large role in hydrologic behavior, even more than did HGM classification. We believe that knowledge of HGM subelass might serve as a useful surrogate for actual knowledge of site-specific hydrology. The level of uncertainty increases with surface-water systems, but we have shown a large degree of predictability by HGM subelass. Our data likely have applicability within the entire Ridge and Valley province of the Appalachian Mountains in the United States, although our conclusions have not been tested over that wide latitudinal range.

The assessment of herbaceous plant cover in wetlands as an indicator of function

Abstract In the United States, wetlands are often created (as compared with restored) as mitigation for damage done to natural wetlands by development or other activities. There is increasing concern that these created sites do not function as do natural wetlands, even after a period of years. Monitoring of these created wetlands often consists of an assessment of the percent herbaceous plant cover as some indicator of the functional success of the wetland. However, it is not at all clear that assessment of herbaceous cover translates into an accurate indicator of wetland function. In this paper I review several functions commonly ascribed to wetlands and assess the reported relationship of percent herbaceous cover to those functions (if any). Of six functions reviewed, only one has a probable (though indirect) positive relationship with the percent herbaceous plant cover on a site. More useful assessments of wetland function might be made with other structural indicators, such as basin morphometry, tree density, or basal area.

Transferability of an HGM wetland classification scheme to a longitudinal gradient of the central Appalachian Mountains: Initial hydrological results

The fundamental role of hydrology in determining HGM classification and function leads to the assumption that any test of regionalization might do well to begin with a comparison of hydrologic variation within regional subclasses across a geographical or landscape continuum. This paper deals with only the basic hydrologic comparisons between New York, Pennsylvania, and Virginia for similarly classified wetlands in all three regions. Water levels for headwater floodplain wetlands varied substantially between the New York region and the Pennsylvania and Virginia regions; the latter regions were very similar. The same pattern was evident for slope wetlands across the three regions, but there was no significant difference in water levels for riparian depressions. Based upon the hydrologic data alone, it seems that applying the classification outside of central Pennsylvania had mixed results; it worked well south to Virginia and less well north to New York. One substantial influence in New York was the presence of beaver (Castor canadensis) that greatly influenced almost every watershed we worked in. Furthermore, climate differences between the three regions may also have a large impact — the New York sites were subject to much more snow than sites further south.

Ecological Theory and Its Role in the Rehabilitation of Wetlands

This chapter was written based on four basic beliefs: 1) wetlands will continue to be destroyed worldwide, 2) knowledge of functions and values of natural wetlands will continue to increase, 3) wetland rehabilitation will increase in popularity as a means of increasing wetland resources, and 4) experience from the United States can contribute to wetland rehabilitation both within the United States and elsewhere in the world.

Aboveground decomposition dynamics in riparian depression and slope wetlands of central Pennsylvania

We examined two types of groundwater-fed wetlands (riparian depressions and slopes) classified using the hydrogeomorphic (HGM) system. These wetland types had previously been shown to differ hydrologically. Our first objective was to determine if HGM was a useful structuring variable when examining aboveground decomposition dynamics (rate of mass loss and rate of nitrogen loss). Our second objective was to determine what soil variables were related to any differences in aboveground decomposition dynamics we might find regardless of HGM subclass. We used the litterbag field bioassay technique, and employed a standard litter type (Phalaris arundinacea) across all wetlands. Our results indicated that HGM would not readily serve as an adequate structuring variable for aboveground decomposition in riparian depressions and slope wetlands of central Pennsylvania. Discriminant analysis and classification and regression tree (CART) modeling found soil cation exchange capacity, soil pH, soil organic matter, and soil % nitrogen to be potentially important soil variables related to mass loss, and soil % nitrogen and soil pH to be potentially important variables related to nitrogen loss rate.