Abstract Introduction Study Methods Results References Tables PDF version

The U.S. Geological Survey is studying vegetative resistance to flow in the south Florida Everglades as part of a multidisciplinary effort to restore the South Florida Ecosystem. In order to test the flow resistance of sawgrass, one of the dominant species in the Everglades, uniform, dense stands of sawgrass were grown in a tilting flume at Stennis Space Center, Mississippi. Depth of water in the flume was controlled by adding or removing metal plates at the downstream end of the flume. A series of experiments were conducted at various flow depths, and the velocity, flow depth, and water-surface slope were measured. During each set of experiments, the sawgrass was sampled in layers from the sediment water interface for vegetative characteristics, biomass, and leaf area index. The results of the vegetation sampling are summarized in a series of tables.

As part of the South Florida Ecosystem Study, "Determination of Vegetative Resistance to Flow," uniform, dense stands of sawgrass were grown in pans that were fit tightly into a tilting flume at Stennis Space Center, Mississippi, to form a 61 m long, 1.8 m wide artificial sawgrass ecosystem (Lee and Carter, 1996). The depth of water in the flume is controlled by adding or removing metal plates at the downstream end. Several series of experiments were conducted at various flow depths between 0 and 90 cm, and vegetative resistance was calculated from velocity, flow depth, and surface-water slope.

During each experimental series, the vegetation in the flume was sampled to determine, as a function of distance from the bed or the sediment/water interface, the biomass per unit area, the number of live stems and leaves per unit area, leaf and stem width, and leaf area index. The general methods for measuring biomass and plant characteristics are outlined below. Measurements were made starting in September, 1995, when the plants were nine months old, and continued at intervals as each individual series of experiments were concluded.

This publication is the first of two planned Open-File reports summarizing the vegetation information by date and plant age. Following the series of experiments described here, part of the flume was converted to a wind tunnel, and several series of experiments were conducted at different flow velocities, depths, and wind speeds. The results of the later set of experiments will be described in the second report.

Measurement dates, type of measurement, and treatment of plants between measurements are summarized in Table 1. Biomass was measured in 37x55 cm quadrats; the number of quadrats varied by date. Leaves, culms, and dead material were cut and removed at 90, 60, 40, 20, and 0 cm from the sediment/water interface, starting at the top of the plants. The plant material from each layer was sorted (see plant description below) and dried at 105 °C for about 12 hours, weighed, and the weight expressed as grams dry weight per square meter (gdw/m²). This method, with variations in the number and positions of the quadrats, was used throughout the duration of the sawgrass experiments. For the first three sampling periods, all live leaves and culms were separated from dead standing leaves and culms and the remaining litter; thus, live biomass includes both leaves and culms, and dead biomass includes all dead material. In October, 1996, we began to separate live leaves from live culms and measure their biomass separately. The dead standing leaves and culms were still combined with the dead litter. By March, 1997, dead upright leaves and culms were present when we did the sampling, and the biomass measurements were further refined to include them. In March and June of 1997, we separated all components, live leaves, live culms, dead standing leaves, dead standing culms, and dead litter, and measured their biomass separately. Biomass data for individual quadrats were averaged to give biomass data for the flume for each date.

It was necessary to trim the tops of the sawgrass back to 1 meter total height frequently to permit the measuring cart to move across the top of the flume. For this reason, the >90 cm layer was not measured after September, 1995, until June 1997, when the tops of the plants were allowed to grow for wind simulation experiments. Visually, the plants were generally healthy and green with strongly stiff and upright leaves (the tips having been cut off). Some mortality occurred as time went on, and new plants also sprouted. During some periods between sampling, plants were thinned out or transplanted to fill gaps. The amount of litter in the bottom increased naturally, but was far less than we observed in the field. For this reason, we added to the bottom litter by throwing the cut-off tops of the plants into the flume in order to more closely simulate natural field conditions.

All leaves and culms in each layer were counted and dried. Leaves were separated into small, medium, and large classes, and six widths were measured for each size class, when possible. Likewise, culms were divided into small and large classes and six diameters were measured for each class. In October, 1996, March, 1997, and June, 1997, additional descriptive information was collected, including number of live culms, number of dead standing culms, number of live leaves, number of dead standing leaves, and, in June, 1997, composition of the vegetation above 90 cm. Descriptive data were summarized for each date. Leaf Area Index (LAI) in m²m^-2 (square meters of leaf opposing flow per square meter of sediment surface) was calculated for each layer using the equation:

LAI = (LL x AWLL + ML x AWML + SL x AWSL + LC x AWLC + SC x AWSC) x DL

where AW = average width in meters of leaves or culms, LL = number of large live leaves only or live plus dead leaves per m², ML = number of medium live leaves only or live plus dead leaves per m², SL = number of small live leaves only or live plus dead leaves per m², LC = number of large live culms or live culms plus dead culms per m², SC = number of small live culms or live plus dead culms per m², and DL = depth of the layer in meters. Use of live versus live plus dead depended on date of sampling. In September, 1995, January, 1996, May, 1996, and October, 1996, the number of leaves and culms used to calculate LAI included only live leaves and culms. In March, 1997, and June, 1997, the number of leaves and culms used to calculate LAI included both live and dead standing leaves and culms. In order to account for the resistance of the remaining dead material, we determined the ratio of dead material/standing biomass for each layer and then multiplied the LAI by the ratio to calculate a litter LAI. This litter LAI was added to the standing LAI to form a corrected LAI for each layer.

Selected results of the flume sampling are summarized Table 2. In general, the plants became larger and more robust as time went on; the culms extended into higher layers, the leaves became larger, and the number of leaves decreased. Table 3 summarizes the leaf area indices (LAI) for the six dates. Biomass and descriptive information for each date are summarized in Tables 4 through 9. Tables are numbered by date; thus Tables 4a through 4d are for September, 1995, and so forth. Figures 1 through 3 summarize the average live, dead, and total biomass in the flume respectively for the six sampling dates. Figure 4 shows the number of leaves, and Figure 5 shows the number of culms in the flume for the six sampling dates.

(left) Figure 1. Average live biomass in the flume, 1995 - 97. [larger image]	(right) Figure 2. Average dead biomass in the flume, 1995 - 97. [larger image]

(above) Figure 3. Average total biomass in the flume, 1995 - 97. [larger image]

---

(above) Figure 4. Number of live leaves in the flume, 1995 - 97. [larger image]

(above) Figure 5. Number of live culms in the flume, 1995 - 97. [larger image]

References Cited

Lee, J. K., and Carter, Virginia, 1996, Vegetation affects water movement in the Florida Everglades: U.S. Geological Survey Fact Sheet FS-147-96, 2 p.