Water: Monitoring & Assessment
Bioassessment Methodology for Mid-Atlantic Wetlands (Maryland Case Study)
Last Updated: March 2000
Don W. Sparling
Project and Amphibian Coordinator
U. S. Geological Survey (USGS)
Biological Resources Division
Patuxent Wildlife Research Center
11510 American Holly Drive
Laurel, MD 20708-4017
Phone: (301) 497-5723
E-mail: Don W. Sparling (email@example.com)
Coordinator for Macrophytes
U.S. Department of Agriculture
Wetland Science Institute
11400 American Holly Dr.
Laurel, MD 20708-4014
Office (301) 497-5933
E-mail: Norman Melvin (Norman_Melvin@usgs.gov)
T. Peter Lowe
Coordinator for Macroinvertebrates
USGS, Patuxent Wildlife Research Center
11510 American Holly Dr.
Laurel, MD 20708-4017
Office: (301) 497-5705
E-mail: T. Peter Lowe (Peter_Lowe@usgs.gov)
Case History for Assessment of Restored Wetlands in the Mid-Atlantic States
Purpose(s) of Project
- Develop a yardstick of biological metrics to assess the progress and condition of reconstructed wetlands in Maryland, Delaware, and Virginia
- Compare the suitability of different assemblages (plants, macroinvertebrates, amphibians) for assessing wetland condition.
- Evaluate the sources and magnitude of variance in data collected for biological metrics.
The project is a joint effort between the USGS, Patuxent Wildlife Research Center, the USDA Natural Resources Conservation Service Wetland Science Institute, and EPA. Project development started in 1995 in response to a lack of information about the success of wetland mitigation projects, especially those associated with wetland restoration on farmlands under set-aside programs. From 1996 to 1998, field work focused on a set of restored and existing wetlands. In 1999, a second set of wetlands was studied to evaluate the robustness of the metrics developed from the first set. All field work has now been completed and metric development is underway.
Because one of the objectives was to assess the sources and magnitude of variance in the different measurements, we intentionally over-sampled our sites compared to what would be done under typical regional or statewide assessments. For three years, work was done on a single set of 30 wetlands, including 22 restored and 8 natural wetlands or Delmarva Bays. All wetlands in the database were depressional, semi-permanent, or seasonal wetlands. Macroinvertebrates were sampled one to four times each year, depending on hydrology, macrophytes during spring and late summer, and amphibians almost continuously during the breeding season.
Sampling Methods and Analysis: Macrophytes
Data on plant species composition, abundance, and dominance were collected through the use of line transects. Pairs of 50 m transects were placed on opposite sides of the wetland at four different hydrological levels as determined at "full pool" levels. The hydrological levels were buffer, less than 25 cm, 25 to 45 cm, and greater than 45 cm. In wetlands larger than five acres, the number of transect pairs was doubled. Each transect was divided into five runs of 5 m each, alternating with 5 m of unsampled transect. Along each run, sampling occurred at specific points (as in point intercept methods) at 1 m intervals. Thus, each run of 5 m contained five sample points. Each transect consisted of five runs of five sampling points each. Each hydrological depth was sampled by two transects, bringing the total number of sample points per hydrological zone to 50. There were:
|(4 hydrological zones) x (2 transects each zone)
x (5 runs per transect) x (5 points per run)
|= (200 points per wetland)|
This was done in spring and fall for each wetland. At each point, the species name and number of individuals were recorded.
In addition to point data, incidental species along the line were recorded to include the species not intercepted at any points along the line. These species were included in species richness and attributes derived from presence or absence data, but not in dominance calculations.
Sampling Methods and Analysis: Macroinvertebrates
Aquatic invertebrates were sampled at approximately six week intervals beginning in late May and continuing until October and were conducted in association with the sampling of water quality, aquatic plants, hydrological and wetland dimensions. Invertebrate samples were collected along transects following compass coordinates originating from markers placed in the deepest part of every wetland before the sampling seasons began. Transect coordinates were randomly selected for each wetland and sampling time. The method of compass points was adapted to each wetland's morphology.
Samples were collected from three depth ranges along the transects to determine invertebrate relative abundance, diversity, and relative biomass in each wetland. As long as water depths were adequate, samples were collected from along the transects at the following water depths: less than 15.0 cm, 15.1 to 45.0 cm, and greater than 45.1 cm.
Samples were collected using a modified Gerking box sampler, which is a sheet aluminum box with a sliding screen door (1 mm mesh) at the bottom. The sampler has the advantage of allowing simultaneous collection of benthic, pelagic, neutonic, and plant-associated invertebrates. The sampler was lowered to the floor of the sampling area with the screen door open. The vegetation in the sampler was then cut at the mud-water interface and put into prelabeled plastic bags. Then the screen door was slowly closed as the sediments just in front of the advancing screen were stirred into the water column. After the screen was closed, soil materials were sieved through the screen by shaking the box. All invertebrates caught on the screen were then placed in prelabeled plastic bags. The bags of vegetation and invertebrates were placed on wet ice as soon as possible and stored in a refrigerator until the samples could be picked. Each individual was keyed to the lowest taxonomic level feasible, typically species or genus. Data on species composition and abundance were used to generate metrics in a manner similar to that for plants. An approved quality assurance/quality control process was followed, which included independent validation of 20 percent of our samples.
Sampling Methods and Analysis: Amphibians
Each site was sampled for amphibian larvae once every four weeks. Sampled areas consisted of the perimeter of the open water portion of the wetland and lightly vegetated areas that allowed a seine to pass. We used a 6 m x 8 m nylon mesh (1/16") seine to sample each wetland by wading out 3 to 5 m away from the shoreline and then moving in towards the shoreline in one continuous sweep. Sampling was time-constrained to two hours. If new species were caught during the last two sweeps, the sampling period was extended until no additional new species were found. Amphibian larvae were identified using published keys and by temporarily housing tadpoles until they metamorphosed and could be identified.
Drift fences were also used to supplement seining data and obtain information on adults and metamorphs. We initially considered surrounding each wetland with a drift fence but realized the impracticality of that idea. Therefore, each wetland was provided with 50-centimeter-tall and 15-meter-long drift fences. If possible, the fences were placed along drains, travel corridors, and other likely points of amphibian use in order to maximize capture of individuals entering and exiting the wetlands. Five-gallon plastic buckets were buried in pairs at the ends and midpoint on the inside and outside of the fence to capture adults entering the ponds and juveniles leaving the ponds. Wet sponges, rocks, and vegetation were placed in the buckets to prevent desiccation and provide some cover and refugia to captured individuals. All amphibians were identified, sexed, and returned to the inside of the fence at the wetland from which they were captured. Juveniles and metamorphs departing the wetlands were identified, counted, examined for malformations, and released on the outside of the drift fence.
General Analytical Methods
A fundamental component of this study is to devise a gradient of physical factors (e.g. land use in drainage area, management techniques, landscape features, method of restoration) that affect wetland health. From there, data on the frequency of occurrence and relative abundance of species, guilds, or trophic classes will be used to develop attributes for an Index of Biotic Integrity for each of the assemblages. An attribute will be considered a valid metric if it relates either positively or negatively to the physical gradient. We will then compare the IBIs developed and determine if they are consistent in ranking wetlands. Once acceptable IBIs have been developed on the initial wetland base, we will apply them to the second set of wetlands to validate the model. In addition, the variance within our sampling methodology will be assessed.
- Many of the wetlands in our bases were only a few years old when we started and may not have had ample time for ecological and anthropogenic factors to separate them along a physical gradient.
- As a result, development of a reliable and ecologically meaningful gradient has been one of the most difficult parts of this project.
- It would be very instructive to revisit these wetlands after 10 or 15 years and see how they have changed.
- There can be considerable differences between mid- and late summer in the ability to easily record and identify plants. This is especially true for graminoids, which are primarily identified by fruiting body characteristics. In addition, many legumes, composites, warm season grasses are present in late summer but not apparent in spring.
- The inclusion of incidental species added appreciably to the number of species identified in a particular wetland. We are evaluating whether this inclusion has an effect on the resulting metrics.
- Deep water areas (greater than 45 cm) have a much lower species richness than shallower zones and do not need to be sampled at the same level of intensity at the same site.
- Permanent transects are preferred if data collection can continue over several years. This will allow for the annual and seasonal changes that occur over time due to shifts in hydrology.
- Picking, sorting, and identifying aquatic insects was one of the most laborious aspects of the study. Investigators wishing to use invertebrates in their bioassessments should allocate sufficient resources to accomplish the task.
- Invertebrate species presence, and especially abundance, are seasonally quite variable. June to early July before the drying of mid- to late summer begins seems to be the best months for finding the greatest diversity and abundance of macroinvertebrates.
- The reduced species richness of amphibians, compared to macrophytes and macroinvertebrates, may limit the number and types of metrics that can be developed from this assemblage.
- Adequate sampling for amphibians requires more trips and techniques than other assemblages. This is due to their mobility, multiple life history strategies, and variable breeding periods among species. Sampling for one life stage only is probably not as effective as sampling for adults and tadpoles in determining amphibian usage of a wetland.