WETLANDS, Vol. 22, No. 3, September 2002, pp. 451–466  2002, The Society of Wetland Scientists

HOW SEDGE MEADOW SOILS, MICROTOPOGRAPHY, AND VEGETATION RESPOND TO SEDIMENTATION Katherine J. Werner1 and Joy B. Zedler2 1 Illinois State Geological Survey 615 East Peabody Drive Champaign, Illinois, USA 61820 2

Botany Department and Arboretum, University of Wisconsin-Madison Madison, Wisconsin, USA 53706 E-mail: [email protected]

Abstract: The expansion of urban and agricultural activities in watersheds of the Midwestern USA facilitates the conversion of species-rich sedge meadows to stands of Phalaris arundinacea and Typha spp. We document the role of sediment accumulation in this process based on field surveys of three sedge meadows dominated by Carex stricta, their adjacent Phalaris or Typha stands, and transitions from Carex to these invasive species. The complex microtopography of Carex tussocks facilitates the occurrence of other native species. Tussock surface area and species richness were positively correlated in two marshes (r2 ⫽ 0.57 and 0.41); on average, a 33-cm-tall tussock supported 7.6 species. Phalaris also grew in tussock form in wetter areas but did not support native species. We found an average of 10.5 Carex tussocks per 10-m transect, but only 3.5 Phalaris tussocks. Microtopographic relief, determined with a high-precision GPS, measured 11% greater in Carex meadows than Phalaris stands. Inflowing sediments reduced microtopographic variation and surface area for native species. We calculated a loss of one species per 1000 cm2 of lost tussock surface area, and loss of 1.2 species for every 10-cm addition of sediment over the sedge meadow surface. Alluvium overlying the sedge meadow soil had a smaller proportion of organic matter content and higher dry bulk density than the buried histic materials. We conclude that sedimentation contributes to the loss of native species in remnant wetlands. Key Words: Carex stricta, invasive species, Phalaris arundinacea, runoff, stormwater, tussocks, Typha spp., floristic quality index, sedimentation

wetlands is a valuable ecosystem service, and even though some wetlands are used to remove sediments from stormwater (Kadlec and Knight 1996), the longterm consequences of sediment deposition are not well-understood (Boto and Patrick 1978). High sedimentation rates (greater than 0.3 cm/year, the rate for prairie wetlands) are known to bury organisms and affect germination rates of wetland plant species (van der Valk et al. 1983, Jurik et al. 1994). However, the changes in the physical properties of wetland soils receiving sediment from runoff and their effects on species distribution and abundance are not well-documented in the literature (EPA 1993). Sediment influx can potentially alter several physical soil properties that comprise the micro-environment of plants, including organic matter content and bulk density. Microtopographic variation is strongly correlated with the distribution and vigor of individual plant species and communities in wetlands (Watt 1947, Schlesinger 1978). In wetland experiments, species richness and the number of rare species were greater in all microtopographically heterogeneous treatments (VivianSmith 1997). Explanations for these patterns differ but

INTRODUCTION The urbanization of watersheds degrades wetlands by altering surface water flows and sedimentary loads (Ewing 1996). When construction activities cause losses in vegetation, increases in impermeable surfaces, and/or exposures of soil, wetlands downstream experience amplified peak stormwater flows, increased erosion, and augmented sediment loading (Kirk 1990). In wetlands of the upper Midwestern USA, the result is often a native species decline; species-rich wet prairies, sedge meadows, and fens tend to become monotypic stands of Phalaris arundinacea L. (reed canary grass) or Typha spp. (i.e., T. angustifolia L., T. latifolia L., or T. X glauca Godr., cattail) (Bedford et al. 1974, Galatowitsch et al. 2000). The underlying mechanisms by which runoff leads to a reduction in native species are not well-understood, although most managers and researchers attribute the changes to increasing nutrient levels, hydroperiod, and sedimentation (Volker and Smith 1965, Bedford et al. 1974, Owen 1999). This study focuses on the outcomes of increased sedimentation. Although the capture and retention of sediment by 451

452

WETLANDS, Volume 22, No. 3, 2002

Figure 1.

Locations of study areas in and near Madison, Wisconsin.

relate to variations in water levels, soil nutrient dynamics, and redox conditions. The ‘‘tussock’’ sedge (Carex stricta Lam.), with a cespitose habit, commonly dominates sedge meadows found in the Upper Midwest. Carex leaves surmount compact pedestals formed from dead stems and leaf debris held together by roots and rhizomes (Costello 1936); these pedestals, or tussocks, create microtopographic relief. We suspected that sedimentation fills in tussock interspaces or ‘‘hollows,’’ reducing the habitat’s spatial heterogeneity and subsequently influencing species richness. Silting associated with stormwater floods in Dane County, Wisconsin, USA reportedly smothered one sedge meadow and allowed the invasion of Phalaris, but effects were not quantified (Kirk 1990). Land surveys and long-term field observers claim that Carex may persist through sedimentation events (Glocker and Patzer 1978, Q. Carpenter, personal communication). Associations between sedimentation and species richness in sedge meadows have yet to be documented rigorously. Hence, we characterized the impacts of sediment accumulation on species richness in sedge meadows. We addressed the following research questions in three sedge meadows in Dane County, WI experiencing stormwater additions. (1) How does sediment accumulation alter micro-

topography and physical soil properties (organic matter content and bulk density)? (2) How strongly are sediment accumulation and species richness correlated? (3) Is there a positive correlation between microtopographic variation and species richness? METHODS Site Descriptions We chose three wetlands that have a history of urban stormwater inputs, sediment accumulation, and stands of Phalaris or Typha adjacent to Carex meadows. Each site was divided into three subsites according to qualitative observations of the plant community: Phalaris or Typha stand, Carex meadow, and transition between Phalaris or Typha stand and Carex meadow. The field study began in April 2000. The UW-Madison Arboretum’s Southeast Marsh (43⬚2⬘30⬙ N, 89⬚24⬘E, elevation ⫽ 260 m) receives sedimentation via stormwater leaving the retention basins in the southwestern and eastern corners of the marsh (Kline 1992). A monotypic Phalaris stand encompasses the stormwater basin, its outflow structure, and the eastern half of the site (Figure 1). The Carex meadow under study persists in the elevated north-

Werner & Zedler, SEDGE MEADOW RESPONSES TO SEDIMENTATION western corner of the site. The marsh was acquired by the Arboretum in 1969 to address the primary management objective of surface water retention. The marsh was disturbed by agricultural activities from the early 1900s to 1950s and hydrologically altered by a berm constructed in the 1980s to retain urban stormwater runoff in the wetland, thereby protecting nearby homes (Kline 1992). Lower Greene Prairie (43⬚1⬘40⬙ N, 89⬚26⬘15⬙ E, elevation ⫽ 285 m), also in the Arboretum, was agricultural land that Henry Greene restored by planting in 1952 (Allsup 1977, Blewett 1981). In the last 15 years, development to the south and west has brought urban runoff to the site. Subsequently, Phalaris invaded and formed a monotypic stand to the detriment of native species (Kline 1992, Zedler 2000). The Phalaris stand occupies the southern area of the study site and transitions into smaller Carex meadows to the north (Figure 1). Pheasant Branch (43⬚06⬘80⬙ N, 89⬚28⬘45⬙ E, elevation ⫽ 197 m) located in Middleton, WI differs from the other sites in that the sampling site contains stands of Typha and limited cover of Phalaris. The site may have been grazed for a brief period in the late 1930s, and the adjacent uplands to the east, southeast, and northeast have been converted from agricultural to urban development since the 1980s (M. A. Braunschweig and N. Hirabayashi, unpublished data). The Typha stand encompasses the outlet structure of the northeastern retention basin and transitions into the Carex meadow in the southwestern corner of the site (Figure 1). Sampling Methods and Calculations Sites were divided into three subsites (subsequently referred to as ‘‘communities’’) to sample the native vegetation, the invasive species, and the transition between them by randomly placing three transects in each subsite for a total of 9 transects per site (27 total transects). Two random coordinates located the start of the 10-m transects, except for transition transects in Greene Prairie and Southeast Marsh that were established by laying out a tape along the ‘‘invasion front’’ where Phalaris met Carex meadow (identified by aerial photographs), since transitions were not as distinct as at Pheasant Branch. The transect was placed perpendicular to the tape along the ‘‘invasion front’’ in locations determined by random numbers. We extended the transect length until we did not encounter Phalaris for 3 meters into the Carex meadow and vice versa for the Phalaris stand. These transition transects varied from 10 to 18 meters in length. Despite variations in length, data collected from these transects were comparable across sites because the unit areas of sampling for each of the variables were held constant.

453

Transects in Greene Prairie were extended to 16 m in Carex meadows and 28 m in the Phalaris stand to gather microtopographic and growth-form data across a longer elevation gradient. Measurements of vegetation, microtopography, and soil characteristics were made along each transect, as described below. Vegetation. At each transect, 5 quadrats (1-m2) were randomly placed at 1.5-meter intervals on alternate sides of the transect for a total of 135 quadrats. In each quadrat, we identified the species present and estimated cover using a Log 2 scale (Gauch 1982): 0.5, 1, 2, 4, 8, 16, 32, 64, ⬎64%. With the data, species-sequence curves for each community were created by graphing the mean percent cover of each species from highest to lowest cover. Species richness was also measured on individual Carex tussocks. We measured diameter and height of six randomly chosen tussocks to calculate the cylindrical surface area of tussocks along each Carex meadow transect (54 total tussocks). Coefficients of Conservatism. To assess floristic quality, we asked expert botanists to assign coefficients of conservatism or ‘‘C values’’ for plants in a region using the methods of Swink and Wilhelm (1994) based on the concept that remnant natural plant communities tend to be inhabited by concentrations of species conservative to them. C values represent ‘‘an estimated probability that a plant is likely to occur in a landscape relatively unaltered from what is believed to be a presettlement condition.’’ Two expert botanists, Q. Carpenter and T. Cochrane of UW-Madison, rated species on a scale of 0–10 based on their confidence that species came from a remnant natural plant community (Herman et al. 1996); a 0 meant no confidence, and a 10 meant certainty. Native species found in both degraded and non-degraded sites were rated as 5. In all, 183 species were rated. Where ratings differed by ⬍ 4, values were averaged and rounded. Ratings that differed ⬎ 4 were rated by a third expert, G. Smith of UW-Whitewater, and the three ratings were averaged and rounded. For each community, we present the mean C value and the Floristic Quality Index (FQI ⫽ (mean C) * 兹n), where n is the number of species in the site); FQI has a scale of 0 to infinity (Andreas and Lichvar 1995, Taft et al. 1997). Soil Profile Description. Soil characteristics were observed from organic and mineral soil layers using a bucket auger (10-cm diameter) to an impenetrable layer. We described the soils at a random location adjacent to each transect for a total of 27 samples. We partitioned the profile into horizons according to Schoeneberger et al. (1998). Soils were categorized as one of three types according to a threshold value of organic matter content (McKinzie 1974): alluvium

454 over histic material, alluvial deposits with high organic matter, or histic material. These three categories are subsequently referred to as ‘‘alluvium over histic,’’ ‘‘mixed,’’ and ‘‘histic.’’ Soil Organic Matter Content and Nutrients. Samples for organic matter measurement in the top 20 cm of soil were obtained using a hammer auger (5-cm diameter) in two random locations at each transect (totaling 54 samples). Samples were placed in separate plastic bags, transported to the laboratory, dried at 105⬚C (for 12 to 14 hours), and then ashed at 500⬚C in a muffle furnace for 3 hours (Council on Soil Testing and Plant Analysis 1992). Organic matter content was determined as the percentage loss in weight between 105 and 500⬚C. We collected another core (5 ⫻ 20 cm deep) from each transect to measure the macro-organic matter content (MOM; Craft et al. 1999), measured as the dry (70⬚ C) weight of the residue left after washing on a 2-mmmesh screen. Nutrient levels (Available P and Total N) in the top 20 cm of soil from each transect were measured by the UW Plant and Soil Science Lab. Soil Bulk Density. Two samples for bulk density were collected at every transect with a hammer auger (5-cm diameter), taking care not to disrupt the samples internally. If the 20-cm-deep core collected more than one layer, then the core was separated into an MOM layer composed of living and dead root and rhizome mat and the histic or alluvial layer below. We dried and weighed each sample, reporting bulk density as the dry weight divided by the volume of the auger. Microtopographic Variation. Changes in microtopography were measured along each transect using a Leica SR530 Global Positioning System (GPS) unit (⫹ or ⫺ 1.4 cm horizontal and vertical accuracy) and realtime differential correction with a radio link to a base station located at a benchmark in or near the wetland. Elevation measurements were based on GEOID99. Using the GPS, the elevation and description of the wetland surface (tussock, hollow, or neither) were recorded at each significant topographic break, a method similar to those used in cedar swamps (Chimner and Hart 1996). We defined a significant topographic break as the top of a tussock or the bottom of a hollow. When no topographic breaks were encountered, we recorded the elevations at meter-interval points along the transect. After the data were downloaded, the spatial distance between consecutive tussock top and hollow elevations was computed, and the sum of those distances along a transect provided a measure hereafter referred to as ‘‘microtopographic relief’’ in meter/meter units of transect. Other measures of microtopographic variation were tussock height and number of tussocks per 10-m transect.

WETLANDS, Volume 22, No. 3, 2002 Redox Potential. We measured redox potential using installed platinum electrodes, a calomel reference electrode, and a portable pH/ mV meter (Patrick et al. 1996). We adjusted meter values by adding ⫹244 mV on the basis of the standard hydrogen electrode (SHE). We conducted a survey of redox potential at Greene Prairie to compare Phalaris tussocks to Carex tussocks in the last week of August 2000. Readings were obtained at 5- and 10-cm distances from the outer surface toward the tussock interior at tussock top, middle, and hollow positions for a total of 82 measurements. Statistical Analyses Parameters were analyzed individually with General Linear Models using the methods of least squares including ANOVAs and Ls means for comparisons. Microtopographic relief, organic matter content, bulk density, and soil nutrients were analyzed individually using two-way ANOVAs to test the main effects of plant community type and site in addition to any interactions. Soil profile type, species richness, and percent cover of species were examined using contingency tables with Chi-Square and Fisher’s Exact, in addition to the two-way ANOVAs. All tests of significance were conducted at the p ⬍ 0.05 level. Environmental parameters were simultaneously examined for correlations using Pearson’s method for all continuous variables (e.g., soil properties, species richness, nutrients) and Spearman’s method for discrete variables (e.g., soil profile type, community type). Height and redox potential of tussocks were analyzed using General Linear Models. The number of tussocks per 10 m was compared with a t test. Tussock surface area and species richness per tussock in the Carex community type were tested according to three regression models (Cochran-Mantel-Haenszel, Mantel-Haenszel Chi-Square, and standard linear analysis). Because all models found significant correlations between surface area and species richness, we report only standard analysis r2- and p-values. Minitab (1996) and SAS Institute (1999) were used for all tests and summary statistics. RESULTS Species Richness For all three sites combined, there was a decrease in the species richness from the Carex meadows to the stands of Phalaris and Typha (Figure 2, Table 1). Within each site, we observed that Carex meadows were more species-rich than Phalaris or Typha stands; however, transitions were not consistently different from other community types. There was also a decrease in the total percent cover of species in the understory (termed ‘‘subordinate species’’) from the Carex mead-

Werner & Zedler, SEDGE MEADOW RESPONSES TO SEDIMENTATION

455

Despite the lack of statistically significant differences in soil profile type between community types, we observed a correlation (r2 ⫽ 0.57, p ⬍ 0.001) between community type and profile type for Southeast Marsh and Greene Prairie whereby Carex stricta and transition cores were histic and mixed and Phalaris cores were alluvium over histic. All Pheasant Branch profiles were alluvium over histic, and we observed an increase in alluvium depth (p ⬍ 0.0001) from the Carex meadow to the transition to the Typha stand (Figure 4). Figure 2. Species richness in the Carex to Phalaris to Typha zones across sites. SE ⫽ Southeast Marsh, GP ⫽ Greene Prairie, PB ⫽ Pheasant Branch. Means with the same letter were not significantly different (see text for results of statistical analysis).

ows (20.7 ⫾ 2.86 %) to the Typha stands (7.6 ⫾ 1.74 %) and Phalaris stands (1.4 ⫾ 0.47 %) (Figure 3). Average Coefficients of Conservatism We observed a significant decrease in the C values and FQIs from the Carex meadow to the Phalaris and Typha stands (Table 2). The vegetation in the Carex meadows is more indicative of a remnant plant community, whereas species in the Phalaris and Typha stands may be found in many disturbed habitats. Soil Profile All sites are underlain by interbedded glaciofluvial and glaciolacustrine deposits from the Late Wisconsin (12,000 YBP). Overlying these deposits are histic and alluvial deposits with medium, very fine to fine granular structure and very friable to firm friability. The type of core found in each site varied significantly (p ⬍ 0.0005, df ⫽ 4, Chi-square value ⫽ 19.8). When analyzed according to soil profile type, overall communities were not significantly different (p ⬍ 0.09; df ⫽ 4, Chi-square value ⫽ 8.5), although the majority of the Phalaris and Typha cores are alluvium over histic. In the Carex meadows, there were equivalent numbers of cores in each of the three soil profile types (histic, mixed, alluvium over histic) (Figure 4). In contrast, only one of nine cores in transition areas was histic. Eight out of nine cores from areas dominated by Phalaris and Typha had ⬎ 40 cm of alluvium over histic. Sedimentation depths were particularly variable in transition areas of Southeast Marsh and Greene Prairie: exploratory coring 10-cm apart along transects revealed profile changes from 0 to 8 cm of alluvium overlying histic material. Personal observations confirmed that sediment depth decreased with distance from the stormwater inflow in all sites.

Bulk Density, Organic Matter Content, MacroOrganic Matter Mass, and Nutrients Plant community types had different soil properties. Bulk density increased significantly from the Carex meadows to the Phalaris and Typha stands as percent organic matter decreased (Figure 5). Carex meadows were growing on soils of consistently lower bulk densities and greater organic matter than Phalaris and Typha stands according to within-site analyses, except for Pheasant Branch, which did not display a significant trend in organic matter content, as all community types had experienced ⬎20 cm of sedimentation and histosol burial. Most MOM in all sites was in the 0– 10 cm depth. The dry weight of MOM was not significantly different among community types at overall or site-specific levels (p ⬍ 0.4471, df ⫽ 24, F ⫽ 1.74). Available P and total N in soils were not directly correlated with community type. There was no difference in total N or available P from the community type. Sites were not significantly different in total N (p ⬍ 0.057, df ⫽ 26, F ⫽ 3.25). Pheasant Branch had significantly more available P than Greene Prairie and Southeast Marsh (p ⬍ 0.029, df ⫽ 26, F ⫽ 7.50). Microtopography Tussock Morphology and Characteristics. We observed tussocks in both Phalaris stands and Carex meadows. Within Phalaris stands, we also observed a ‘‘sward’’ or lawn-like form. Personal observations of rooting structures and soil underneath Phalaris tussocks along drainage ditches in Dane County, WI indicated that Phalaris was capable of creating tussocks and did not appear to overgrow pre-existing tussocks from other species. Based on this preliminary observation that requires further study, we refer to tussocks within Phalaris stands as Phalaris tussocks. Phalaris tussock bases occurred at an average elevation 10 ⫾ 3 cm lower than sward forms in Greene Prairie (Table 3). The number of tussocks per 10 m decreased from the Carex meadows to the Phalaris stands (Carex ⫽ 9.5 ⫾ 1.2 tussocks per 10 m and Phalaris ⫽ 3.4 ⫾ 0.7; p ⫽ 0.0004, df ⫽ 16, t ⫽ 4.4).

6.0 ⫾ 1.07 na

⬎ ⫽ ⬎ ⫽ ⫽ ⫽ ⫽

10.8 ⫾ 1.16 1.06 ⫾ 0.07 25.1 ⫾ 2.83 854 ⫾ 48 0.64 ⫾ 0.07 11.6 ⫾ 2.96 9273 ⫾ 1048

Species richness (species) Microtopographic relief (m/m) Sediment accumulation (cm) Organic matter content (g/100 g) Macro-organic matter content (g/m3) Bulk density (g/cm3) Available P (g/m3) Total N (g/m3) 0.69 ⫾ 0.07 21.0 ⫾ 1.00 5476 ⫾ 2807

1182 ⫾ 561

18.3 ⫾ 2.30

transition

Carex

Greene Prairie

Parameter

Table 1. Extended.

⬍ ⫽ ⫽

⫽

⬎

⬎ ⫽

1.03 ⫾ 0.04 18.7 ⫾ 2.96 3689 ⫾ 677

1002 ⫾ 384

7.84 ⫾ 0.42

1.6 ⫾ 0.21 1.02 ⫾ 0.01

Phalaris

0.27 ⫾ 0.05 10.0 ⫾ 5.13 26298 ⫾ 4716

406 ⫾ 7

61.6 ⫾ 3.50

5.1 ⫾ 0.39 1.11 ⫾ 0.02

Carex

Comparisons within sites

⫽ ⫽

⫽ ⫽ ⫽

⫽

⬎

0.4 ⫾ 0.06 19.6 ⫾ 5.24 52801 ⫾ 41866

1089 ⫾ 245

28.0 ⫾ 2.68

4.6 ⫾ 0.55 1.14 ⫾ 0.04

transition

Southeast Marsh

⫽⬍ ⫽ ⬎

⫽

⬎

⬎ ⬎

0.51 ⫾ 0.04 21.7 ⫾ 6.23 9477 ⫾ 239

970 ⫾ 339

19.2 ⫾ 3.05

1.3 ⫾ 0.16 1.02 ⫾ 0.01

Phalaris

0.95 ⫾ 0.04 25.0 ⫾ 4.51 2204 ⫾ 186 ⬍ ⫽ ⫽ 0.85 ⫾ 0.04 28.0 ⫾ 4.36 3492 ⫾ 229 ⬍ ⫽ ⫽

0.66 ⫾ 0.05 35.3 ⫾ 1.45 4741 ⫾ 412

0.83 ⫾ 0.06 21.8 ⫾ 2.55 5123 ⫾ 1323

⬍ ⫽ ⫽

0.65 ⫾ 0.06 22.9 ⫾ 2.37 20590 ⫾ 14548

⬍ ⫽ ⫽

0.51 ⫾ 0.05 19.0 ⫾ 4.45 3437 ⫾ 3567

353 ⫾ 138 ⫽

0.86 0.01 3.71 0.29 299 ⫾ 113

⫽

460 ⫾ 250

775 ⫾ 186

⫽

816 ⫾ 205

⫽

538 ⫾ 108

10.8 ⫾ 1.75

⬎

17.7 ⫾ 2.39

⬎

31.9 ⫾ 5.52

⫾ ⫾ ⫾ ⫾

5.8 ⫾ 0.78 1.00 ⫾ 0.001 121 ⫾ 3.33 5.35 ⫾ 0.31

Typha ⬎ ⫽ ⬍ ⫽

8.2 1.02 92.7 6.60

⫽ ⬎ ⬍ ⫽

0.77 0.03 6.01 0.53

⫾ ⫾ ⫾ ⫾

8.7 1.16 36.7 8.82

2.9 ⫾ 0.41 1.02 ⫾ 0.01

⬎ ⬎

6.3 ⫾ 0.53 na

⬎ ⬎

8.2 ⫾ 0.58 1.11 ⫾ 0.04

Species richness (species) Microtopographic relief (m/m) Sediment accumulation (cm) Organic matter content (g/100 g) Macro-organic matter content (g/m3) Bulk density (g/cm3) Available P (g/m3) Total N (g/m3)

transition

Carex

Pheasant Branch Phalaris, Typha

transition

Carex

Parameter

Comparisons across sites

Table 1. Means, standard errors, and significant relationships of parameters across sites and within each site. Alpha ⫽ 0.05 (unless otherwise stated). Refer to Werner (2001) for p-values. ⬍ significantly less than, ⬎ significantly greater than, ⫽ no significant difference. * Phalaris is significantly greater than Carex but not significantly different from the transition. Data on macro-organic matter are from 20-cm-deep soil cores, as are P and N concentrations (g/m3 ⫽ parts per billion).

456 WETLANDS, Volume 22, No. 3, 2002

Werner & Zedler, SEDGE MEADOW RESPONSES TO SEDIMENTATION

457

Werner 2001). Available P, total N, and MOM were not associated with other parameters in pooled or siteby-site analyses. Soil profile type, organic matter content, and bulk density were all closely linked (r2 ⬎ 0.60, p ⬍ 0.01), indicating that sediment accumulation significantly altered physical attributes of wetland soil.

Figure 3. Species-sequence curves for communities dominated by Carex, Phalaris, and Typha. Species are arranged in descending order of percent cover. Communities dominated by Carex (solid squares) and Typha (open diamonds), both at Pheasant Branch, had many species at low cover, while the stand dominated by Phalaris at Southeast Marsh (solid triangles) was nearly monotypic.

Microtopographic Relief. Microtopographic relief decreased from at least 1.11 to 1.02 m/m from the Carex meadow to the transition to the Phalaris and Typha stands of Southeast Marsh and Pheasant Branch (p ⬍ 0.0001, df ⫽ 17, F ⫽ 28.34). In Greene Prairie, there was not a detectable decrease from Carex meadows to Phalaris stands (p ⬍ 0.17, df ⫽ 8, F ⫽ 2.34) (no transition transects were measured). Within the sites, the comparison of microtopographic relief in subsites varied, most likely due to Phalaris’ apparent ability to create tussock or sward forms. In Southeast Marsh and Pheasant Branch, the Carex meadows contained significantly greater relief (p always ⬍ 0.0018, df ⫽ 26, F ⫽ 13.59) than the Phalaris stands. Redox Potential in Carex and Phalaris Tussocks in Greene Prairie Redox potentials in tussocks were recorded after a wet period of 4.6 cm precipitation over 11 days. Soils were more oxygenated in Carex tussocks (mean ⫽ 352 ⫾ 12.8 mV) than Phalaris tussocks (137 ⫾ 18.2 mV) (p ⬍ 0.0001, df ⫽ 53, F ⫽ 22.43). In Phalaris and Carex tussocks, Eh varied significantly from the top to the middle to the hollow (p ⬍ 0.0405, df ⫽ 2, F ⫽ 3.44). The range of Eh levels decreased from Carex tussocks to Phalaris tussocks (210 mV vs. 108 mV). Correlations Between Parameters Pooled Correlations. After pooling the data from all sites, soil and species richness varied significantly and consistently according to community type (data in

Site-by-Site Correlations. In Southeast Marsh, species richness was correlated with microtopographic relief (r2 ⫽ 0.67, p ⬍ 0.046) and soil profile category (r2 ⫽ ⫺0.77, p ⬍ 0.014). Soil profile type was correlated with bulk density (r2 ⫽ 0.82, p ⬍ 0.007) and organic matter content (r2 ⫽ ⫺0.73, p ⬍ 0.026) such that histic cores had lower bulk density and higher organic matter content than alluvium cores. Mixed cores were in between. In Greene Prairie, species richness was not correlated with microtopographic relief but was correlated with soil profile category (r2 ⫽ ⫺0.89, p ⬍ 0.04), as more species were found in mixed cores than in alluvium cores. Soil profile type was correlated with organic matter content (r2 ⫽ ⫺0.82, p ⬍ 0.007) and bulk density (r2 ⫽ 0.82, p ⬍ 0.007). In Pheasant Branch, species richness was correlated with depth of alluvium (r2 ⫽ ⫺0.60, p ⬍ 0.09) but not microtopographic relief. Depth of alluvium was correlated with bulk density (r2 ⫽ 0.61, p ⬍ 0.081), organic matter content (r2 ⫽ ⫺0.95, p ⬍ 0.0001), and microtopographic relief (r2 ⫽ ⫺0.88, p ⬍ 0.0018), indicating that sediment accumulation from stormwater runoff is associated with the gradient of environmental parameters witnessed between community types. Surface Area of Tussocks. We found positive correlations between tussock surface area and species richness in Carex meadows. Standard linear analyses found significant linear correlations (r2 ⫽ 0.27, p ⬍0.0001 for all sites pooled, Figure 6) between the number of plant species growing on a tussock and its surface area such that a decrease of 1000 cm2 could eliminate approximately one species (site was not included in the pooled model). In site-by-site analyses, significant correlations existed in Southeast Marsh (r2 ⫽ 0.57, p ⫽ 0.0003) and Pheasant Branch (r2 ⫽ 0.41, p ⫽ 0.004), but not in Greene Prairie (r2 ⫽ 0.03, p ⫽ .47; Figure 6). DISCUSSION Stormwater retention and sediment capture are notable ecosystem services provided by wetlands, but there are limits to the amount of sediments that can accrete without causing a loss in physical, chemical, or biological integrity. In three sedge meadows, we documented the influence of sediment accumulation on soil properties, microtopography, and vegetation. In every site, we found that sediment accumulation was

Agropyron trachycaulum (Link) Malte Bromus ciliatus L. Campanula aparinoides Pursh Carex sartwellii Dewey Carex sterilis Willd. Cirsium muticum Michaux Galium tintorium L. Galium trifidum L. Lathyrus palustris L. Lycopus uniflorus Michaux Lysimachia quadriflora Sims. Lysimachia thyrsiflora L. Muhlenbergia glomerata (Willd.) Trin. Poa palustris L. Thalictrum dioicum L. Thelypteris palustris Schott Viola cucullata Aiton Viola nephrophylla Greene

Angelica atropurpurea L. Ascelpias incarnata L. Aster puniceus L. Aster puniceus var. firmus (Nees) T. & G. Boehmeria cylindrica (L.) Sw. Calamagrostis canadensis (Michaux) Beauv. Caltha palustris L. Carex hystericina Willd. Carex lanuginosa (pellita) Michaux Carex scoparia Willd. Carex stricta Lam. Eleocharis erythropoda Steudel Eupatorium maculatum L. Eupatorium perfoliatum L. Glyceria striata (Lam.) Hitchc. Helenium autumnale L. Impatiens capensis Meerb. Iris virginica L. Juncus dudleyi L. Leersia oryzoides (L.) Sw.

C ⫽ 6-4

Species

C ⫽ 10-7

C group

6 5 6 6 4 6 6 5 5 4 5 4 6 6 6 6 4 5 6 5

7 8 7 8 10 8 7 8 7 7 9 7 8 7 7 7 7 9

C

0.70 2.10

0.10 8.07 0.53

9.00

6.13 19.33 2.40 0.07

31.93

0.13 4.67

13.53

0.07

0.13

0.13 21.67 4.27

0.07

0.80 0.07

1.07 0.03

0.10

transition

0.03 0.30

0.07

0.03

Carex

PB

0.53 5.77

0.03 0.17

0.27

0.13

0.07 0.03

9.20

0.13

0.70

Typha

0.67 0.03 0.03 0.10 0.27

0.57 49.33 5.33 0.10

0.40

0.40 1.07

0.57 1.20 0.67 1.13 0.20 0.03

0.97 0.37

0.23 0.30

1.07

0.07

Carex

0.07 2.73

0.07 0.90

15.77 2.53

0.07

0.27 0.27

1.40 0.07

0.03

0.93

0.03

0.87

0.30 0.17

transition

GP

Sites

Phalaris

8.70

0.67

42.13

26.20 0.03

0.10

0.07

0.20

Carex

7.43

26.33

4.27

4.67

0.07

0.07

0.03

transition

SE

0.07

0.53

Phalaris

Table 2. Coefficients of Conservativism (C) for each species and their mean percent cover in the community type and site in which they were found. Species are grouped according to conservativeness: species with C ⫽ 10-7 are indicative of relatively unaltered communities. C ⫽ 6-4 are found in both altered and unaltered communities. C ⫽ 3-0 are found in degraded sites. Note that the Pa stands had few species with C ⬎ 3.

458 WETLANDS, Volume 22, No. 3, 2002

C ⫽ 3-0

C group

Mean C Floristic Quality Index

1 0 0 3 2 0 0 3 3 3 3 0 0 1 0 0 1 3 3 1 1

6 5 4 5 4 6 6 6 5 4 6 5 6 4 6

Lycopus americanus W. P. C. Barton Mentha arvensis L. Monarda fistulosa L. Pilea pumila (L.) A. Gray Polyganum punctatum Ell. Polygonum sagittatum L. Pycnanthemum virginianum (L.) B. L. Rob & Fernald Rumex orbiculatus A. Gray Salix bebbiana Sarg. Scirpus atrovirens Willd. Scutellaria galericulata L. Stachys palustris L. Stellaria longifolia Willd. Verbena hastata L. Veronicastrum virginicum (L.) Farw.

Acer negundo L. Agrostis gigantea (alba) Roth Ambrosia artemisifolia L. Carex stipata Willd. Carex vulpinoidea Michaux Cirsium arvense (L.) Scop. Convolvulus arvensis L. Cuscuta gronovii Willd. Epilobium coloratum Biehler Equisetum arvense L. Helianthus grosseserratus M. Martens Lythrum salicaria L. Phalaris arundinacea L. Phragmites australis (Cav.) Steudel Poa compressa L. Poa pratensis L. Solidago canadensis L. Solidago gigantea Aiton Typha latifolia L. Typha x glauca Godron Urtica dioica L.

C

Species

Table 2. Continued.

5.2 27.1

0.07 1.07 0.53

0.10 0.50

0.53

0.77

0.03 0.10

4.2 21.6

0.47

0.23

0.03 1.60

0.03

2.13 3.20

19.20

0.53

0.53 0.03

0.03

3.5 17.9

18.13

0.73 0.80 0.53 0.27

0.60

0.23 4.57

0.67 0.03

6.07 0.20

2.13

0.13

0.47

0.10

Typha

4.03

1.53

0.37 0.60

transition

1.07

0.10

0.33

Carex

PB

5.2 32.7

0.13 1.90

0.13 0.03 0.07 0.07 36.93

0.07 2.77

0.07

0.87 0.33

4.27

1.73

0.07

0.27

Carex

4.8 25.6

0.17

83.47

0.27

0.27

0.07

0.67 0.13

0.27

0.03 0.07

transition

GP

Sites

2.4 4.8

0.07

0.27

0.03

Phalaris

4.9 18.8

22.90

0.30

0.03

0.07

0.03

0.03

0.03

Carex

4.5 17.3

0.03

0.03

66.40

0.07

0.10

0.20

0.47

0.03

transition

SE

2.0 3.5

0.07

Phalaris

Werner & Zedler, SEDGE MEADOW RESPONSES TO SEDIMENTATION 459

460

WETLANDS, Volume 22, No. 3, 2002

Figure 4. Diagrams of soil profiles in the communities of each site. Alluvium over histic profiles were observed most often, and at greater depths, in Phalaris or Typha communities.

associated with lower species richness, greater dry soil bulk density, and lower soil organic matter. Sediment accumulation was negatively correlated with microtopographic relief in two out of three sites. Surface area of tussocks was positively linked to species richness for all Carex meadows. Characteristics of Sediment Accumulation

Figure 5. Soil parameters by community and site. Groups with similar letters did not differ significantly. Mean dry bulk density in the top 20 cm of soil increased from Carex meadows to the Phalaris and Typha stands across sites, but sites were not significantly different from one another in bulk density. Mean organic matter content in the top 20 cm decreased from Carex meadows to Phalaris and Typha stands across sites, and sites were significantly different from one another.

The histic deposits overlying glacial lakebed sediments are like other undisturbed sedge meadow soil profiles throughout the state of Wisconsin and began forming after the mid-Holocene warm-up (6,000 YBP) (Costello 1936, Curtis 1959, F. Madison, personal communication). Mixtures of histic and mineral material also occur in sedge meadows that incorporate detritus with modest amounts of mineral soils transported from the uplands (Stout 1914, Curtis 1959). We documented that sediment has accreted on the order of 0.4 to 1.3 meters above histic horizons. The color and thickness of the silt loam overlying histic material suggest that it is top soil washed in from the surrounding uplands following agricultural and urban development that began approximately 150 years ago (Glocker and Patzer 1978, Johnston et al. 1984, Werner 2001). In Greene Prairie, it was difficult to separate sediment resulting from agricultural versus urban impacts. In Southeast Marsh, the coarse sand and granules that buried patches of vegetation in the Phalaris stand are indicative of street run-off, as they are consistent with the grain size reported for adjacent watershed moni-

Werner & Zedler, SEDGE MEADOW RESPONSES TO SEDIMENTATION

461

Table 3. Tussock heights and elevations. Base elevation (m above means sea level). For substrate, elevation was measured where tussocks were ⬎1 m apart or where plants formed a sward. X means that all tussocs were ⬍1 m apart. Pooled mean

Site PB

Height of tussock (cm) Elevation of tussocks Elevation of substrate

GP

SE

Carex

Typha

Carex

Phalaris

Carex

Phalaris

17.2 ⫾ 0.01

0.04 ⫾ 0.01

11.4 ⫾ 0.01

12.8 ⫾ 0.01

16.0 ⫾ 0.01

10.7 ⫾ 0.01

196.5 ⫾ 0.02

197.6 ⫾ 0.01

285.1 ⫾ 0.02

285.0 ⫾ 0.03

259.8 ⫾ 0.01

260.4 ⫾ 0.11

X

X

285.2 ⫾ 0.03

285.0 ⫾ 0.03

X

X

toring (Waschbusch et al. 1999). In Pheasant Branch, we observed a change in sediment texture, color, and character that distinguished the agricultural from urban sediment (Werner 2001). Where Phalaris has invaded alluvial silt loam, it has created a thick rhizome mat (mean thickness ⫽ 14 cm) intermixed with silt. Silt that coated Phalaris and Typha vegetation and created a fresh soil surface indicates that sediment accreted during the study period in all sites. Therefore, sedimentation in wetland remnants is an on-going process. Coring revealed a high degree of spatial variation, particularly in the transition subsites of Southeast Marsh and Greene Prairie. We attribute variations in accumulation to subtle changes in topography over short distances that generate different flow velocities

Figure 6. Species richness correlated with Carex tussock surface area in (A) pooled sites (r2 ⫽ 0.27, p ⬍0.0001), (B) Southeast Marsh (r2 ⫽ 0.57, p ⫽ 0.0003), (C) Pheasant Branch (r2 ⫽ 0.41, p ⫽ 0.004), but not in (D) Greene Prairie (r2 ⫽ 0.03, p ⫽ 0.47). Statistical relationships were determined by standard linear regression model (total SS explained by regression/total SS adjusted for mean).

PhalCarex aris 14.9

11.8

(Hupp and Bazemore 1993, K. McSweeney, personal communication). At the larger scale, the spatial distribution of sediment was directly related to the inflow, whereby the thickest areas of accumulation were observed to be closest to the stormwater entry, correlated with the locations of Phalaris and Typha stands. Sediment Accumulation is Correlated with a Loss in Species We documented a decrease in species richness with increased sediment accumulation in each site. We also documented a decrease in conservatism in Phalaris stands, where few species have a C greater than 4. According to Herman et al. (1996), the mean C and FQI values convey that the Phalaris and Typha stands are composed of opportunistic species that are not unique to the southern Wisconsin region or sedge meadow habitat. The Carex meadows possess more conservatism and richness than the Phalaris and Typha stands, but overall, the moderate values indicate that all areas have been degraded. Galatowitsch et al. (2000) contended that the decrease in natives is a predictable response to wetlands exposed to urban and agricultural impacts. Below, we discuss the possible mechanisms by which stormwater sedimentation alters sedge meadow habitat and influences the decrease in natives (Table 4). In each site, the degree of sediment accumulation increased from Carex meadows to the Phalaris and Typha stands. This increase was correlated with an increase in bulk density and decrease in organic matter. As in other studies (e.g., Craft and Casey 2000), we attribute the observed decrease in proportion of organic matter and increase in bulk density to accretion. Stormwater sedimentation alters soils in sedge meadows through mineral additions to the organic-rich system. The accumulation of organic matter is the net effect of vegetation production and decomposition (SanchezCarrillo et al. 2001). Alluvium cores represent younger soils (Entisols) than the histic and mixed cores (Histo-

462

WETLANDS, Volume 22, No. 3, 2002

Table 4. Direct effects of sediment accumulation and their indirect effects. Major Component of Wetland Habitat Soil properties

Vegetation

Direct Effects decrease in organic matter content increase in bulk density disrupted/buried soil surface nutrient additions plant burial inhibited seedling growth inhibited germination increase in invasives

Reference

Indirect Effects

this study this study this study Boto and Patrick 1978 Johnston et al. 1984 Martin and Hartman 1986 van der Valk 1983 Jurik et al. 1994 Dittmar and Neely 1999 Keddy and Constabel 1986 this study

Hunt et al. 1999

increase in invasive species increase in invasive species

Hobbs 1989 Woo 2000

decline in species richness increase in invasive species

this study this study

increase in sediment accumulation

Bernard and Lauve 1995 this study

decline in microtopographic variation Topography

decreased plant decomposition reduction in microtopographic variation

Vargo et al. 1998 this study

decline in species richness

decline in redox potential variability

sols and Mollisols) (Werner 2001), and their low organic matter content and high bulk density indicate they have not accumulated organic C reserves from overlying vegetation compared to the pre-settlement histosols. MOM was fairly constant across community types in all sites, consistent with the findings of Craft et al. (1999), who observed MOM quantities in constructed marshes equal to or greater than in natural marshes after 5 to 10 years. In our sites, MOM may break down over time and contribute to soil organic C reserves. However, as sedimentation slows the process of plant decomposition (Vargo et al. 1998), the soil surface may accumulate thick fibric MOM accumulations with little incorporation of more-decomposed organic material to the profile. We speculate that changes in organic matter and bulk density caused by stormwater sedimentation may influence the decrease in species richness by altering soil structure and the moisture regime. Due to their well-developed structure (Schwartzendruber et al. 1954, Richardson and Hole 1978), organic soils have greater water-holding capacities than mineral soils (Mitsch and Gosselink 1993, Skaggs et al. 1994), so the capillary fringe can extend meters above the water table, keeping the histic soils saturated close to the surface (Hunt et al. 1999). As sediment accumulation creates a fresh soil surface,‘‘resetting the clock’’ for soil development, sedimentary deposits will have minimal structure compared to the pre-settlement histic horizons influenced by abiotic and biotic factors for approximately 6,000 years. This loss in structure may lead to a decrease in soil moisture. Experimental research demonstrates that a reduction

Reference

decrease in soil water-holding capacity

this study Bratton 1976 Schlesinger 1978 Vivian-Smith 1997 this study

in soil moisture has a negative impact on the ability of Carex to compete with Typha and Phalaris (Wetzel and van der Valk 1998). Low soil moisture (especially when ⬍ 50 g H2O/100 g soil) is also a major impediment to the establishment of Carex spp. (van der Valk et al. 1999). Ashworth’s (1997) surveys demonstrate that native sedge species were most abundant in soils of thicker histic material. Therefore, by altering soil structure and moisture, sediment accumulation may slow the growth of Carex. Sediment accumulation also impacts species richness through smothering plants. Van der Valk et al. (1983) documented a decrease in shoot density in areas of experimental sediment addition. In a controlled greenhouse experiment with seedlings grown in flats receiving small amounts of sediment (0.25–0.5 cm), Jurik et al. (1994) demonstrated that the number of wetland species per flat and the total number of seedlings decreased significantly with increasing depth of sediment. By limiting oxygen availability (Leck 1996), burial and changes in soil particle size can also have a negative effect on recruitment (Keddy and Constabel 1986), thereby disrupting the composition of a species-rich wetland by inhibiting regeneration of sensitive species. While burying the pre-existing vegetation in sedge meadows, sediment accumulation creates gaps in the native canopy, disrupts the soil surface, and adds phosphorus and toxins to the soil (e.g., Boto and Patrick 1978, van der Valk 1983). Such disruptions are often linked to invasions (Hobbs 1989). For example, Phalaris establishes best in disturbed canopies with high light availability (Lindig-Cisneros and Zedler, in press). Phalaris and Typha occurred more often in ar-

Werner & Zedler, SEDGE MEADOW RESPONSES TO SEDIMENTATION eas of alluvium and areas of deepest alluvium, suggesting that they are more successful at establishing and persisting in areas of sediment accumulation. When compared to the agricultural setting, available P and total N were high in all sites (Kelling et al. 1991). High phosphorus levels indicate mineral additions and stormwater impact; high total nitrogen levels indicate abundant organic matter (Mitsch and Gosselink 1993). Soil disturbance enhances invasion, and nutrient additions enhance expansion (Hobbs 1989, Woo 2000). Although nutrient measurements were limited in scope and require further pursuit, the presence of both sedimentation and nutrient additions sets the stage for invasive species to establish. Our surveys demonstrated that species richness is lower where Phalaris and Typha dominate (Figure 2), supporting the findings of other studies in the upper Midwest (Bedford et al. 1974, Apfelbaum and Sams 1987). Therefore, by facilitating the establishment of invasive species, sedimentation facilitates a loss in species richness. Sedimentation creates canopy gaps, fresh substrate, and nutrient additions that appear to enhance a wetland’s invasibility. However, once established, Phalaris and Typha may not require these conditions to spread from a parent clone. Our transition data indicate that the invasives are expanding into relatively ‘‘speciesrich’’ canopies overlying heterogeneous distributions of histic, mixed, and alluvium over histic soils. Because Phalaris seeds are sensitive to light quality for germination, Phalaris germination is lower under complex species-rich canopies (⬎6 species) that reduce light at the soil level (Lindig-Cisneros and Zedler in press). However, shading (simulated complex canopies) did not prevent the spreading of Phalaris tillers attached to a parent clone (Maurer and Zedler 2002). Since Phalaris parent clones have already established in these sites, it is unlikely that species-richness and soil type inhibit the rate of movement at the ‘‘invasion front.’’ This evidence indicates that once Phalaris invades new areas, a homogenization of the habitat results. Our vegetation surveys and those of R. L. Veltman (unpublished data) demonstrate that 5–10 year old Typha canopies are not as devastating (a mean of 5.8 and 6.8 species, respectively) as Phalaris (mean ⫽ 1.5 species) to species richness. However, Reinartz and Warne (1993) suggested that species richness in Typha communities will decrease to one over time. Therefore, sedimentation likely plays a role in regional species loss by providing disturbance that facilitates invasive species establishment. Once established, invasives have the capability to expand and incur a loss in species richness. Sediment accumulation also alters the sedge meadows by diminishing microtopography. The reduction of microtopographic relief and increase in sediment accumulation from Carex meadows to Phalaris and

463

Typha stands in two of the three sites provides evidence that sediments diminish microtopography by burying tussocks. Other researchers suggest that sediments eliminate wetland habitat through filling, but they provide no data (Reed et al. 1977, Martin and Hartman 1987, Sanchez-Carrillo 2001). Further research (e.g., monitoring of concurrent microtopographic change and sediment accretion) is necessary. Microtopographic relief is not only influenced by levels of accretion, but also by vegetational cover (Watt 1947). We provide evidence of different tussock sizes and occurrences produced by Phalaris and Carex. Phalaris tussocks were significantly shorter than Carex tussocks in Southeast Marsh (10.7 vs. 15.9, p ⬍ 0.006, df ⫽ 156, F ⫽ 4.35). These heights are distinguished from hummocks in other wetland systems: sedge meadows, 5–90 cm; floodplains, ⬎50 cm; cedar swamps, 70–100 cm; and bogs 50–70 cm (Costello 1936, Titus 1990, Ehrenfeld 1995b). It is likely that hummock heights are determined by morphological attributes of the species, age, rate of accretion, and water levels, although these relationships have not been rigorously documented (Costello 1936). Although we cannot separate the influences of taxon morphologies and sedimentation on microtopographic variation from our surveys, we do have evidence indicating that tussock dimensions can influence the heterogeneity of the physical environment. For example, redox was significantly greater in Carex tussocks, suggesting that there was more available oxygen compared to Phalaris tussocks. We explain this difference in terms of tussock height. We found that Eh levels increased with elevation and varied significantly from tussock top, middle, and hollow positions, a trend also documented in hummocks of Chamaecyparis-Sphagnum wetlands (Ehrenfeld 1995a). Since Carex tussocks are taller than Phalaris tussocks, it follows that Eh would reach more oxygenated levels at the tops of the taller Carex tussocks and that the taller Carex tussocks would have a broader Eh range from top to hollow. In accordance, Bledsoe and Shear (2000) found that an elevation difference of 10 cm resulted in a 20% change in flooding frequency, emphasizing the potential difference in wetness and oxygen availability between the top and bottom of a tussock. Thus, we demonstrate how increased tussock height, a measure of microtopographic variation, increased redox variability–theoretically increasing the number of available microsites for species. A decrease in microtopography may reduce species richness in areas of sediment accumulation. According to our regression models, an additional 1000 cm2 of tussock surface area could support an additional species. With this relationship, we project the number of species lost as surface area is buried by sedimentary additions. For example, on the tallest tussock in Pheas-

464

WETLANDS, Volume 22, No. 3, 2002

Figure 7. Illustration of a sedge meadow experiencing (A) minimal accumulation where robust, tall Carex tussocks occur with high microtopographic variation, organic-rich soil, high species richness, and a high floristic quality index (FQI ⫽ 26.2), (B) partial smothering by sediment that decreases microtopographic variation, reduces organic content in soil, and reduces species richness and floristic quality, and (C) complete burial by sediment. Where Phalaris replaces Carex, microtopographic variation is minimal, soils are mineral-rich, and only 1–2 opportunistic species remain (FQI ⫽ 4.1).

ant Branch (height ⫽ 33 cm, basal circumference ⫽ 120 cm), we calculated a loss of 1.2 species for every 10 cm of sediment accumulation. We would lose an additional 4 species after the tussock top is buried because the top has a large surface area. Therefore, a cumulative loss of 7.6 species under 34 cm of sediment accumulation would result—a level of sedimentation exceeded in the Phalaris and Typha stands in every site (Werner 2001). Previous researchers report that microtopographic heterogeneity is a controlling factor on the diversity of vegetation (Titus 1990, Vivian-Smith 1997). Although Levine (2000) implied that the size of a Carex tussock is not important to diversity and invasion, we present data showing that the number of species is highly correlated with the tussock surface area, consistent with the ‘‘species-area’’ curve of MacArthur and Wilson (1967). Our data suggest that loss of microtopography at the tussock scale will cause a decrease in species richness. With fewer species, the sedge meadow might become more invasible (Lindig-Cisneros and Zedler, in press). The coarse-scale relationship is more complex. Microtopographic relief decreased as species richness decreased from the Carex meadows to the Phalaris and Typha stands in Southeast and Pheasant Branch Marshes. However, only in Southeast Marsh was the loss in microtopographic relief directly correlated with the loss in species richness. Microtopography may be highly important at the tussock scale but masked at the scale of entire wetlands by other site-specific biotic and abiotic conditions (Andrus et al. 1983, Vivian-Smith 1997). Microtopographic relief was not significantly different in community types at Greene Prairie, nor was the sur-

face-area vs. species correlation as strong. In this site, the Carex meadows had the lowest relief values due to their young age as restoration plantings. Also, the smaller Carex tussocks are located at a higher elevation than the Phalaris stand, indicating that they may be less exposed to stormwater flows, in contrast to the Carex meadows at the other two sites. We observed that stormwater sedimentation is accompanied by an increase in invasive species and a decrease in species richness. Some researchers contend that changes involving species at low frequencies and cover will not impact the functional properties of the wetland (Ehrenfeld and Schneider 1991). However, Phalaris prevents the recruitment of native species and alters the community composition with implications for ‘‘gross structural, and hence functional’’ aspects of the ecosystem (MacDonald et al. 1989). In addition, Phalaris is known for its capacity to produce numerous shoots, stabilize soils, promote the settling of solids, and grow in fine mineral substrates (Bernard and Lauve 1995), demonstrating its ability to influence the magnitude of sedimentation in the wetland. Vitousek (1990) claims that such biological invasions have a greater likelihood of changing and even controlling functions where they alter disturbance frequency or intensity. Therefore, stormwater sedimentation disrupts soils, microtopography, and species richness and promotes the spread of invasives that can perpetuate the degradation of sedge meadows and the functions they provide (Figure 7). Such evidence demonstrates the urgent need for watershed management and wetland restoration. Wetlands are effective sinks for nonpoint source pollution, including sediments, associated nutrients, and pollutants from urban and rural developments (Mitsch

Werner & Zedler, SEDGE MEADOW RESPONSES TO SEDIMENTATION 1994). However, as sediment loads increase, there are costs for these services. When the load becomes too great, the wetland is no longer able to protect downstream waters (Bedford et al. 1974). Our evidence regarding the impacts of sediment accumulation indicates that natural remnants are critically overburdened by the nonpoint source pollution of current land-use practices. Unless activities in the watershed change, we predict that the regional sedge meadows will fill in with sediment, and native species assemblages will be replaced by invasives. We recommend that managers reverse these trends by seizing opportunities to restore ecological integrity in remnants and to reduce soil loss and stormwater runoff throughout the watershed. ACKNOWLEDGMENTS Reviews of the manuscript by B. Svarstad, K. McSweeney, D. Wilcox and anonymous individuals are appreciated. We are grateful to R. Hozak for statistical advice and W. Wallace, T. McClintock, and the Center for Restoration Ecology for technical support. We appreciate the efforts of people at the Arboretum and UW Herbarium who made this effort possible. Thanks to D. Maurer, B. Miller, S. Kercher, K. Fye, A. Fuller, and many others who helped with data collection in the field. This research was supported by a research assistantship from the Aldo Leopold Chair endowment of the University of Wisconsin—Madison Arboretum and approved by the Friends of Pheasant Branch. LITERATURE CITED Allsup, M. 1977. Greene Prairie: a model for prairie restoration. M. S. Thesis. University of Wisconsin, Madison, WI, USA. Andreas, B. K. and R. W. Lichvar. 1995. Floristic index for establishing assessment standards: a case study for northern Ohio. U. S. Army Corps of Engineers, Waterways Experiment Station, Vicksburg, MS, USA. Technical Report WRP-DE-8. Andrus, R. E., D. J. Wagner, and J. E. Titus. 1983. Vertical zonation of Sphagnum mosses along hummock—hollow gradients. Canadian Journal of Botany 61:3128–3139. Apfelbaum, S. I. and C. E. Sams. 1987. Ecology and control of reed canary grass (Phalaris arundinacea L.). Natural Areas Journal 1: 69–72. Ashworth, S. M. 1997. Comparison between restored and reference sedge meadow wetlands in south-central Wisconsin. Wetlands 17: 518–527. Bedford, B. L., E. H. Zimmerman, and J. H. Zimmerman. 1974. Wetlands of Dane County, Wisconsin. Wisconsin Department of Natural Resources, Madison, WI, USA. Bernard, J. B. and T. E. Lauve. 1995. A comparison of growth and nutrient uptake in Phalaris arundinacea L. growing in a wetland and a constructed bed receiving landfill leachate. Wetlands 15: 176–182. Bledsoe, B. P. and T. H. Shear. 2000. Vegetation along hydrologic and edaphic gradients in a North Carolina coastal plain creek bottom and implications for restorations. Wetlands 20:126–147. Blewett, T. J. 1981. An ordination study of plant species ecology in the Arboretum prairies. M. S. Thesis. University of Wisconsin, Madison, WI, USA. Boto, K. G. and W. H. Patrick. 1978. Role of wetlands in the removal of suspended sediments. p. 479–489. In P. E. Greeson, J.

465

R. Clark and J. E. Clark (eds.) Wetland Functions and Values: the State of our Understanding. American Water Resources Association, Minneapolis, MN, USA. Braunschweig, M. A. and N. Hirabayashi. 2000. Soils and History of the North Branch of Pheasant Branch. Institute for Environmental Studies, University of Wisconsin, Madison, WI, USA. Chimner, R. A. and J. B. Hart. 1996. Hydrology and microtopography effects on northern white-cedar regeneration in Michigan’s Upper Peninsula. Canadian Journal of Forest Research 26:389–393. Costello, D. F. 1936. Tussock meadows. Botanical Gazette 97:610– 648. Council on Soil Testing and Plant Analysis. 1992. Handbook on reference methods for soil analysis. Soil and Plant Analysis Council, Inc., Athens, GA, USA. Craft, C. B., J. Reader, J. N. Sacco, and S. W. Broome. 1999. Twentyfive years of ecosystem development of constructed Spartina alterniflora (Loisel) marshes. Ecological Applications 9:1405–1419. Craft, C. B. and W. P. Casey. 2000. Sediment and nutrient accumulation in floodplain and depressional wetlands of Georgia. Wetlands 20:323–332. Curtis, J. T. 1959. Vegetation of Wisconsin. University of Wisconsin Press, Madison, WI, USA. Ehrenfeld, J. G. 1995a. Microsite differences and surface substrate characteristics in Chamaecyparis swamps of the New Jersey Pinelands. Wetlands 15:183–189. Ehrenfeld, J. G. 1995b. Microtopography and vegetation in Atlantic white cedar swamps: the effects of natural disturbances. Canadian Journal of Botany 73:474–484. Ehrenfeld, J. G. and J. P. Schneider. 1991. Chamaecyparis thyoides wetlands and suburbanization: effects on hydrology, water quantity, and plant community composition. Journal of Applied Ecology 28:467–490. U. S. Environmental Protection Agency (EPA). 1993. Natural wetlands and urban stormwater: potential impacts and management. U.S. EPA, Office of Wetlands, Oceans and Watersheds, Wetlands Division, Washington, DC, USA. Ewing, K. 1996. Tolerance of four wetland species to flooding and sediment deposition. Environmental and Experimental Botany 36: 131–146. Gauch, H. G., Jr. 1982. Multivariate Analysis in Community Ecology. Cambridge University Press, Cambridge, UK. Galatowitsch, S. M., D. C. Whited, R. Lehtinen, J. Husveth, and K. Schik. 2000. The vegetation of wet meadows in relation to their land-use. Environmental Monitoring and Assessment 60:121–144. Glocker, C. and R. Patzer. 1978. Soil survey of Dane County, Wisconsin. Department of Agriculture, Soil Conservation Service, Washington, DC, USA. Herman, K. D., L. A. Maters, M. R. Penskar, A. A. Reznicek, G. S. Wilhelm, and W. W. Brodowicz. 1996. Floristic quality assessment with wetland categories and computer application programs for the state of Michigan. Michigan Department of Natural Resources, Wildlife Division, Natural Heritage Program, Lansing, MI, USA. Hobbs, R. J. 1989. The nature and effects of disturbance relative to invasions. p. 389–406. In J. A. Drake and H. A. Mooney (eds.) Biological Invasions: a Global Perspective. SCOPE 37. John Wiley and Sons, Chichester, UK. Hunt, R. J., J. F. Walker, and D. P. Krabbenhoft. 1999. Characterizing hydrology and the importance of ground-water discharge in natural and constructed wetlands. Wetlands 19:458–472. Hupp, C. R. and D. E. Bazemore. 1993. Temporal and spatial patterns of wetland sedimentation, west Tennessee. Journal of Hydrology 141:179–196. Johnston, C. A., G. B. Lee, and F. W. Madison. 1984. The stratigraphy and composition of a lakeside wetland. Soil Science Society of America Journal 48:347–354. Jurik, T., S. Wang, and A. van der Valk. 1994. Effects of sediment load on seedling emergence from wetland seek banks. Wetlands 14:159–165. Kadlec, R. H. and R. L. Knight. 1996. Treatment Wetlands. Lewis Publishers, Boca Raton, FL, USA. Keddy, P. A. and P. Constabel 1986. Germination of ten shoreline

466 plants in relation to seed size, soil particle size, and water level: an experimental study. Journal of Ecology 74:133–141. Kelling, K. A., E. E. Schulte, L. G. Bundy, S. M. Combs, and J. B. Peters. 1991. Soil Test Recommendations for Field, Vegetable and Fruit Crops. Agricultural Bulletin, University of Wisconsin Extension, Madison, WI, USA. Kirk, E. (ed.). 1990. Urban Wetlands in the Yahara-Monona Watershed: Functional Classification and Management Alternatives. Institute for Environmental Studies, University of Wisconsin, Madison, WI, USA. Kline, V. 1992. University of Wisconsin Arboretum Long Range Management Plan. University of Wisconsin, Madison, WI, USA. Leck, M. 1996. Relationship between seeds, soil seed bank dynamics and establishment requires understanding of factors that control germination. Bulletin of Torrey Botanical Club 123:48–67. Levine, J. M. 2000. Species diversity and biological invasions: relating local process to community pattern. Science 288:852–854. Lindig-Cisneros, R. and J. B. Zedler. 2001. Effects of light on seed germination in Phalaris arundinacea L. (reed canary grass). Plant Ecology 155:75–78. Lindig-Cisneros, R. and J. B. Zedler. In press. Species-rich canopies limit the germination microsites for Phalaris arundinacea L. Oecologia. MacArthur, R. H. and E. O. Wilson. 1967. The Theory of Island Biogeography. Princeton University Press, Princeton, NJ, USA. MacDonald, I. A., L. L. Loope, M. B. Usher, and O. Hamann. 1989. Wildlife conservation and the invasion of nature reserves by introduced species: a Global Perspective. p. 215–256. In J. A. Drake and H. A. Mooney (eds.) Biological Invasions: a Global Perspective. SCOPE 37. John Wiley and Sons, Chichester, UK. Martin, D. B. and W. A. Hartman. 1987. The effect of cultivation on sediment composition and deposition in prairie pothole wetlands. Water, Air, and Soil Pollution 34:45–53. Maurer, D. A. and J. B. Zedler. 2002. Differential invasion of a wetland grass explained by tests of nutrients and light availability on establishment and clonal growth. Oecologia 131:279–288. McKinzie, W. E. 1974. Criteria used in soil taxonomy to classify organic soils. p. 1–10. In A. R. Aandahl, S. W. Buol, D. E. Hill, H. H. Bailey, M. Stelly, and R. C. Dinauer (eds.) Histosols: Their Characteristics, Classification, and Use. SSSA Special Publication Series no. 6. Soil Science Society of America, Inc., Madison, WI, USA. Minitab 11 for Windows. 1996. http://www.minitab.com Mitsch, W. J. and J. G. Gosselink. 1993. Wetlands, second edition. Van Nostrand Reinhold, New York, NY, USA. Mitsch, W. J. 1994. The nonpoint source pollution control funciton of natural and constructed riparian wetlands. p. 531–362. In W. J. Mitsch (ed.) Global Wetlands: Old World and New. Elsevier Science B. V., Amsterdam, The Netherlands. Owen, C. R. 1999. Hydrology and history: land use changes and ecological responses in an urban wetland. Wetlands Ecology and Management 6:209–219. Patrick, W. H., R. P. Gambrell, S. P. Faulkner. 1996. Redox measurements of soils p. 1255–1273. In D. L. Sparks, A. L. Page, P. A. Helmke, R. H. Loeppert, P. N. Foltanpour, M. A. Tabatabai, C. T. Johnston, M. E. Sumner, J. M. Bartels, and J. M Bingham (eds.) Methods of Soil Analysis. Part 3. Chemical Methods. Series 5. Soil Science Society of America and American Society of Agronomy, Madison, WI, USA. Reed, D. M., J. H. Reimer, and J. A. Schwarzmeier. 1977. Some observations on the relationship of floodplain siltation to reed canary grass abundance. p. 99–107. In C. DeWitt and E. Soloway (eds.) Wetlands Ecology, Values, and Impacts: Proceedings of the Waubesa Conference on Wetlands. Institute for Environmental Studies, University of Wisconsin, Madison. WI, USA. Reinartz, J. A. and E. L. Warne. 1993. Development of vegetation in small created wetlands in southeastern Wisconsin. Wetlands 13: 153–164. Richardson, J. L. and F. D. Hole. 1978. Influence of vegetation on water repellency in selected western Wisconsin soils. Soil Science Society of America Journal 42:465–467. SAS version eight. 1999. SAS Institute Inc., Cary, NC, USA. Sanchez-Carrillo, S., M. Alvarez-Cobelas, and D. G. Angeler. 2001.

WETLANDS, Volume 22, No. 3, 2002 Sedimentation in the semi-arid freshwater wetland Las Tablas de Daimiel (Spain). Wetlands 21:112–124. Schlesinger, W. H. 1978. On the relative dominance of shrubs in Okefenokee Swamp. American Naturalist 112:949–954. Schoeneberger, P. J., D. A. Wysocki, E. C. Benham, and W. D. Broderson. 1998. Field Book for Describing and Sampling Soils, version 1.1. http://www.statlab.iastate.edu/soils/nssc/fieldgd/ fieldgd.htm. National Soil Survey Center, Natural Resources Conservation Service, US Department of Agriculture, Lincoln, NE, USA. Schwartzendruber, D., M. J. De Boodt, and D. Kirkham. 1954. Capillary intake rate of water and soil structure. Soil Science Society of America Proceedings 18:1–7. Skaggs, R. L., D. Amatya, R. O. Evans, and J. E. Parsons. 1994. Characterization and evaluation of proposed hydrologic criteria for wetlands. Journal of Soil and Water Conservation 49:501–510. Stout, A. B. 1914. A biological and statistical analysis of the vegetation of a typical wild hay meadow. Wisconsin Academy of Sciences, Arts, and Letters: 405–469. Swink, F. and G. Wilhelm. 1994. Plants of the Chicago Region, fourth edition. Indiana Academy of Science, Indianapolis, IN, USA. Taft, J. B., G. S. Wilhelm, D. M. Ladd, and L. A. Masters. 1997. Floristic quality assessment for vegetation in Illinois, a method for assessing vegetation integrity. Erigenia 15:3–23. Titus, J. H. 1990. Microtopography and woody plant regeneration in a hardwood floodplain swamp in Florida. Bulletin of the Torrey Botanical Club 117:429–437. Vargo, S. M., R. K. Neely, and S. M. Kirkwood. 1998. Emergent plant decomposition and sedimentation: response to sediments varying in texture, phosphorous content and frequency of deposition. Environmental and Experimental Botany 40:43–58. van der Valk, A. G., S. D. Swanson, and R. F. Nuss. 1983. The responses of plant species to burial in three types of Alaskan wetlands. Canadian Journal of Botany 61:1150–1164. van der Valk, A. G., T. L. Bremholm, and E. Gordon. 1999. The restoration of sedge meadows: seed viability, seed germination requirements, and seedling growth of Carex species. Wetlands 19: 756–764. Veltman, R. L. 2001. The effects of hydroperiod variability on floristic diversity in created wetlands adjacent to the DesPlaines River, IL. Institute for Environmental Studies, University of Wisconsin, Madison, WI, USA. Vitousek, P. M. 1990. Biological invasions and ecosystem processes: towards an integration of population biology and ecosystem studies. Oikos 57:7–13. Vivian-Smith, G. 1997. Microtopographic heterogeneity and floristic diversity in experimental wetland communities. Journal of Ecology 85:71–82. Volker, R. and S. G. Smith. 1965. Changes in the aquatic vascular flora of Lake East Okoboji in historic times. Iowa Academy of Science 72:65–72. Waschbusch, R. J., W. R. Selbig, and R. T. Bannerman. 1999. Sources of phosphorus in street dirt from two urban residential basins in Madison, Wisconsin, 1994–5. U.S. Geological Survey. Water-Resources Investigations Report 99–4021. Watt, A. S. 1947. Pattern and process in the plant community. Journal of Ecology 35:1–22. Werner, K. J. 2001. Stormwater sedimentation alters soils and microtopography and facilitates invasive plants in sedge meadows of southern Wisconsin. Master’s Thesis. University of Wisconsin, Madison, WI, USA. Wetzel, P. R. and A. G. van der Valk. 1998. Effects of nutrient and soil moisture on competition between Carex stricta, Phalaris arundinacea, and Typha latifolia. Plant Ecology 138:179–190. Woo, I. 2000. Can nutrients alone shift a sedge meadow towards the invasive Typha X glauca. M. S. Thesis. University of Wisconsin, Madison, WI, USA. Zedler, J. B. 2000. (ed.). Adaptive Restoration of Lower Greene Prairie: a Baseline Study and Recommendations. Arboretum, University of Wisconsin, Madison, WI, USA. Manuscript received 5 July 2001; revisions received 10 January 2002 and 8 March 2002; accepted 28 May 2002.