Macroinvertebrate and zooplankton responses to ... - IngentaConnect

Arch. Hydrobiol.

162

2

187–210

Stuttgart, February 2005

Macroinvertebrate and zooplankton responses to emergent plant production in upper Mississippi River floodplain wetlands Michael B. Flinn1, Matt R. Whiles1, S. Reid Adams1, 2 and James E. Garvey1, 2 With 6 figures and 4 tables

Abstract: The upper Mississippi River is managed as a multiple use system, balancing the needs of industry, recreation, and navigation. Through water-level management (WLM), the U. S. Army Corps of Engineers manipulates water levels in some navigation pools to promote emergent vegetation for fish and wildlife in off-channel habitats. A manipulative experiment was conducted to assess responses of invertebrates to vegetation and associated organic matter associated with WLM in floodplain wetlands of Mississippi River navigation pool 25. Macroinvertebrate and zooplankton communities and benthic organic matter were compared in replicated paired plots consisting of experimentally de-vegetated and adjacent control plots. Hydrology was variable during the study, resulting in a strong vegetation response year (1999), a no response year (2000), and a moderate response year (2001). Differences in benthic organic matter between plot types were most pronounced in fall 2000, following the strong vegetation response year, where residual vegetation resulted in significantly higher total and coarse benthic organic matter in vegetated plots. Of the 2 years (2000, 2001) they were sampled, differences in macroinvertebrate communities were only evident in 2001 when there was a moderate vegetation response. Total macroinvertebrate densities during 2001 were similar between devegetated and vegetated plots, but responses of individual taxa varied. Total macroinvertebrate biomass was significantly higher in the vegetated plots during 2001. Of dominant groups, Oligochaeta biomass was significantly higher in vegetated plots, whereas total Chironomidae abundance and biomass were both significantly higher in devegetated plots. Community metrics reflected higher macroinvertebrate diversity in vegetated plots. Zooplankton were not abundant or diverse in plots, but total densities were significantly higher in vegetated plots during the 1999 strongest vegetation response year. 1

Authors’ addresses: Department of Zoology, Mailcode 6501, Southern Illinois University-Carbondale, Carbondale, IL 62901-6501, U. S. A; E-mail: [email protected] 2 Fisheries and Illinois Aquaculture Center, Southern Illinois University-Carbondale, Carbondale, IL, 62901-6501, U. S. A. DOI: 10.1127/0003-9136/2005/0162-0187

0003-9136/05/0162-0187 $ 6.00

 2005 E. Schweizerbart’sche Verlagsbuchhandlung, D-70176 Stuttgart

188

M. B. Flinn et al.

Results demonstrate that WLM to enhance emergent vegetation in off-channel habitats of this large river influences organic matter dynamics and invertebrates. Although many groups responded positively to vegetation, responses varied with the degree of vegetation response and the taxonomic resolution of analyses, suggesting that hydrologic variability, and associated spatial and temporal variability in emergent vegetation, can enhance invertebrate diversity in floodplain wetland habitats. Key words: large river, floodplain, wetland, hydrology, organic matter.

Introduction Management of the great rivers of the world is often challenging because of competing demands of navigation, recreation, industry, and water quality protection. The Upper Mississippi River (UMR) is important for both recreation and industry and also serves as a navigational link between the Gulf of Mexico and the midwestern U. S., with commercial towboats utilizing a system of 29 locks and dams. Managing for navigation has greatly influenced the river and its biological communities. Fremling & Claflin (1984) documented extensive impacts on the 1078-km reach from Minneapolis, Minnesota to St. Louis, Missouri. The future value of the Mississippi River for public, private, and wildlife use is linked directly to the health of the river itself, which in turn depends on development of a management plan that considers relationships between hydrology, physicochemical environments, and biological communities. Levee, dike, and dam construction and channel improvements have caused major changes in off-channel habitats in the UMR (Chen & Simons 1986). Construction of dams significantly altered the river, particularly portions directly upstream of locks and dams. Whereas the pre-dam floodplain of the UMR was heavily wooded and had numerous lakes and sloughs that underwent seasonal flooding and drying (Chen & Simons 1986), raised water-levels of the modern river shorten the dry season and reduce flood-intolerant vegetation (Sparks 1995). Increased erosion from anthropogenic activities in the drainage basin has increased sedimentation and further degraded off-channel habitats (Chen & Simons 1986). Degradation and loss of off-channel habitats including floodplain wetlands in the UMR (Sheehan & Konikoff 1998) have led to reduced species diversity, degraded water quality, declining fisheries resources, and hydrologic changes (Fremling & Claflin 1984, Grubaugh & Anderson 1988, Sparks 1995, Poff et al. 1997). Large rivers depend on annual connections between the main channel and floodplain for proper ecosystem function (e. g., Junk et al. 1989). Locks and dams on the UMR do not alter major floods because the gates are lifted once discharge is greater than the specified control range, which often occurs during spring high water events (Wlosinski & Hill 1995). However, with mod-

Macroinvertebrate response to plant production

189

ifications for navigation and flood control, connectivity with the floodplain is greatly reduced and hydrographs are now flashier (i. e., more rapid and extreme changes in discharge), fluctuations are less predictable, and minimum water levels are higher during the growing season compared to pre-modified conditions. Aquatic communities of the UMR evolved in a system with a predictable annual hydrologic cycle consisting of spring flooding and associated connectivity with the floodplain, summer drought, moderate fall flooding, and uniformly low discharge during winter (Lubinski et al. 1991, Sparks et al. 1998). The U. S. Army Corps of Engineers (USACE) has begun addressing these issues through water level management (WLM), which includes reducing summer water levels and better simulating historical hydrology. There is a need for river ecologists to work in conjunction with the USACE to improve hydrologic conditions for biota in the UMR within the constraints of a multi-use system (e. g., Woltemade 1997). The objectives of the USACE’s water-level management (WLM) plan are to maintain relatively low, stable water levels following maximum drawdown to promote establishment of emergent vegetation in floodplain wetlands without impeding commercial navigation. Emergent vegetation is then slowly re-flooded in late summer/fall, providing aquatic habitat for riverine biota. Water-level management was implemented by the USACE on 3 navigation pools in 1994 with the goals of consolidating substrates, re-establishing wetland biogeochemical processes, and enhancing food and habitat for fish and wildlife. Water-level manipulations could affect backwater communities such as aquatic invertebrates, both directly through hydrologic fluctuations and indirectly through vegetation production. WLM has increased moist-soil, wetland vegetation, mainly millets (Echinochloa spp.), chufa (Cyperus esculentus), and smartweeds (Polygonum spp.) (Garvey et al. 2003), which provide food for waterfowl directly from seeds, tubers, and leaves, and possibly indirectly by increasing invertebrate abundance (e. g., Fredrickson & Taylor 1982). Garvey et al. (2003) also documented young of the year fish use of vegetation associated with WLM. Although a variety of studies have examined responses of aquatic invertebrates to hydrologic fluctuations (e. g., Murkin & Kadlec 1986, Neckles et al. 1990, Boulton et al. 1992, Whiles & Goldowitz 2001) and different types of vegetation (e. g., Voigts 1976, Smock & Stoneburner 1980, Schramm & Jirka 1989, Hargeby 1990, Diehl 1992, Campeau et al. 1994, Olson et al. 1995, Shriver et al. 1995, Parsons & Matthews 1995, De Szalay & Resh 2000, Brodersen et al. 2001), few have quantified differences between vegetated and non-vegetated areas (e. g., Wrubleski 1989, Beckett et al. 1992). Further, few studies have examined relationships between hydrology, vegetation, and animal communities in large rivers (e. g., Theiling et al. 1996, Corti et al. 1997), and none that we know of have as-

190

M. B. Flinn et al.

sessed invertebrate responses to vegetation produced through WLM or similar management practices. To assess the influence of WLM on floodplain wetland communities, we examined macroinvertebrates and zooplankton responses to WLM-induced vegetation and associated organic matter by creating replicate devegetated plots and comparing them to adjacent, unmanipulated control plots. We predicted that emergent vegetation produced through WLM would increase invertebrate abundance, biomass, and diversity by enhancing organic matter resources and structural habitat.

Methods Study sites Mississippi River Pool 25 stretches from river km 390.3 to 440.2 and is confined by lock and dam 25 at Cap au Gris, Missouri. Lock and dam 25 is one of three dams operated by the USACE, St. Louis District. This stretch of river is denied access to its historic floodplain by levees on both the Missouri and Illinois sides, and historic floodplain habitat is very limited on the Illinois side of Pool 25 due to the presence of limestone bluffs. The surrounding land is used primarily for row crop agriculture and the majority of available wooded terrestrial habitat is located on islands within the pool (Wlosinski & Rogala 1996). Pool 25 drains an area of ~368 000 km2. Main channel water temperature averages 14 ˚C (31 ˚C max/0 ˚C min), mean discharge averages ~2500 m3/s, and mean water-level elevation is 131.9 m a. s. l. (Wlosinski & Rogala 1996). Substrates consist of sand and gravel in or near the main channel and fine silt and clay in off-channel habitats (Patrick 1998). Water levels in pool 25 are managed by the USACE at a mid-pool control point. Over a specific range of discharges, the target water-level at the control point is 133.2 m a. s. l., the water-level elevation required to maintain a 2.7 m navigation channel at zero discharge (Wlozinski & Hill 1995). At Lock and Dam 25 (lower end of the pool), dam gates are left open to allow spring water levels to fall from full pool elevation of 132.3 m a. s. l. to an elevation of 131 m a. s. l. (maximum drawdown) when discharge exceeds 1700 m3/s. Current USACE operating procedures allow no control over the magnitude or timing of maximum drawdown. If spring discharge remains within moderate ranges (500–1700 m3/s), a maximum drawdown is not required and water levels can be manipulated with the lock and dam. When water levels in the lower end of the pool are below full pool, mudflats are exposed allowing moist soil vegetation to germinate; the timing and duration of the drawdown determine the community composition and quantity of the vegetation in floodplain habitats of the lower pool (Garvey et al. 2003). Four (1999), five (2000), and six (2001) sites were selected in lower Pool 25 for quantifying invertebrate responses to vegetation produced through WLM. Jim Crow, Turner, Batchtown West, and Batchtown East were sampled in 1999, 2000, and 2001, and Dixon Pond and Hausgen were each added in 2000 and 2001, respectively to increase replication (Fig. 1). Sites were selected based on similar hydrology, substrates,


191

Fig. 1. Map of Mississippi River Pool 25 showing research sites. Batchtown East, Batchtown West, Jim Crow, and Turner were sampled in 1999, Dixon Pond was added in 2000, and Hausgen was added in 2001.

192

M. B. Flinn et al.

depth, and gradient at full pool water levels, greatly limiting the number of available sites. Paired plots consisting of a 400-m2 vegetated plot and adjacent 400-m2 devegetated plot, were established at each site. Vegetated plots were undisturbed, whereas the herbicide Rodeo (glyphosate 54 %) was used to remove vegetation from devegetated plots during the summer drawdown. Plots were not inundated during application and plant height was only 5 – 20 cm. The herbicide was applied with a backpack sprayer as per label instructions for a concentration of 2.3 L/ha. Glyphosate has relatively low persistence in the environment (Gardner & Grue 1996, Giesy et al. 2000, Williams et al. 2000). Vegetation removal was performed early in the drawdown cycle and plots were checked before inundation to confirm that vegetation did not persist or redevelop. Inspections revealed that in all cases during 1999 and 2001, when drawdown was sufficient to produce vegetation, devegetated plots were devoid of vegetation when water returned. During 2000, experimental de-vegetation was unnecessary because of extended inundation through summer and the absence of vegetation.

Organic matter and macroinvertebrates Macroinvertebrate communities and benthic organic matter were examined during the 2nd and 3rd years of the study (2000 and 2001). Three 314-cm2 stovepipe core samples were collected in each vegetated and devegetated plot in September following reflooding. Exact sample locations within plots were chosen in a stratified random manner to account for the depth gradient across the plots and the corer was plunged through the water and into the substrate. All water in the core was then bailed out and substrates were removed down to a depth of about 10 cm below the substrate surface. The entire contents [water, substrates, and vegetation (living and dead)] were placed into a 20 L bucket and elutriated through a 250-µm sieve in the field. Material retained on the sieve was placed in plastic bags and preserved in 8 % formalin solution with phloxine-b dye added to aid sorting. Basic water chemistry parameters (dissolved oxygen, conductivity, pH, turbidity) were also measured in each plot using handheld meters. In the laboratory, samples were further elutriated to remove inorganics and fractioned into coarse particulate organic matter (CPOM; > 1 mm) and fine particulate organic matter (FPOM; < 1 mm > 250 µm) using nested sieves. Following macroinvertebrate removal (see below), fractions were dried for a minimum of 48 hours at 50 ˚C, weighed, ashed at 500 ˚C for 1– 2 hours, and then reweighed to estimate ash-free dry mass (AFDM). For large samples, dried materials were homogenized and a subsample was used to determine percent ash; this percentage was then applied to total dry mass to estimate AFDM. Macroinvertebrates in coarse fractions ( > 1 mm) were sorted and removed by eye or under low magnification. Fine fractions ( < 1mm > 250 µm) were processed under a dissecting microscope. Fine fractions were occasionally subsampled (up to 1/16 of total) using a Folsom wheel when they were large.


193

Macroinvertebrates were identified to the lowest practical taxonomic level (usually genus) according to Merritt & Cummins (1996) & Smith (2001) and measured (total body length). Taxon-specific length-mass regressions (Rogers et al. 1976, Bottrell et al. 1976, Schoener 1980, Stagliano et al. 1998, Benke et al. 1999) were used to estimate biomass [dry mass (DM)] based on body length.

Zooplankton Zooplankton communities were sampled in plots during all three years using a 9-cm diameter littoral sampling tube (Pennak 1962). Three samples were collected in each plot after late summer/fall re-flood (Sept. 12 and Oct. 14, 1999; Sept. 30 and Oct. 21, 2000; Sept. 9 and Oct. 6, 2001). Samples collected in the same year were combined for analyses. Thus, there were six total samples collected in each treatment per year. Samples were rinsed through 80 µm Nitex mesh, placed in 100-ml plastic containers, and preserved in 5 % buffered formalin. In the laboratory, samples were rinsed through an 80 µm Nitex mesh sieve, placed under a dissecting microscope, and zooplankton were counted and identified to the lowest practical level (usually family) using Smith (2001). Abundance data were converted to individuals/L based on the original volume of water sampled.

Statistical analyses Organic matter, macroinvertebrate, and zooplankton data from each year were analyzed separately using one-way analysis of variance blocked over sites. The JMP statistical software package (SAS Institute 2001) was used for all analyses. Values were transformed [log (x + 1), or arcsine where appropriate] to normalize data and reduce heteroscedasticity prior to analysis. We tested for significance at α = 0.05. Because of replication constraints, low statistical power, and the inherent variability of these types of field data, we considered P values of 0.05 – 0.1 marginally significant. Analyses were limited based on a priori decisions to include major taxa and taxa of interest. We analyzed study years separately because of distinct differences in hydrology and vegetation during each year. We emphasize results from 2001 for interpreting

Table 1. Water-level management (WLM) on Mississippi River Pool 25 during 1999, 2000, and 2001. Numbers represent the number of days 0.15, 0.3, 0.6, 0.9, and 1.2 m equal to or below full pool. Re-flood rate is the difference in water levels between drawdown and 0.15 m below full pool divided by the number of days between those water levels. WLM Initiated Ended 1999 12-Jun 2000 1-Jul 2001 19-Jun

24-Aug 31-Jul 15-Aug

Number of days below full pool ( < 132.3 m a. s. l.) ≥ 0.15 m ≥ 0.3 m ≥ 0.6 m ≥ 0.9 m ≥ 1.2 m re-flood rate 69 31 58

60 28 53

54 22 27

36 12 7

21 7 2

4.1 cm/day 13.2 cm/day 3.6 cm/day

194

M. B. Flinn et al.

Table 2. Vegetation response in experimental plots of Mississippi River Pool 25 following late summer re-flood during 1999 – 2001. Data summarized from Garvey et al. (2003). High plant cover = > 70 %, moderate = 30 – 60 %, and low = < 20 %. Cover was estimated within three random locations (0.1 m2 sample frame) within each plot. Dashes indicate no plant cover. 1999 Dominant Plant Taxa Batchtown East

2000

Plant Cover Dominant Plant Taxa

2001

Plant Cover Dominant Plant Taxa

Plant Cover

Polygonum High Echinochloa

None

–

None

–

Batchtown West Polygonum High Echinochloa

None

–

Cyperus Amaranthus

Moderate

–

Echinochloa Moderate Amaranthus

Dixon Pond

not sampled

not sampled None

Hausgen

not sampled

not sampled not sampled not sampled Cyperus Amaranthus

Jim Crow

Echinochloa Moderate Polygonum

None

–

Cyperus Moderate Echinochloa

Turner

Leptochloa Cyperus

None

–

Eragostis Cyperus

Moderate

Low

Low

macroinvertebrate responses to vegetation treatments (sampled only during 2000 and 2001) because very little emergent vegetation was produced during 2000 (Tables 1, 2).

Results Water levels and emergent vegetation

Hydrology during 1999, 2000, and 2001 was variable, resulting in very different plant responses each year (Tables 1, 2, Fig. 2). In 1999, water levels remained low, well beyond the targeted mid-summer (July) drawdown period (Fig. 2 a). This extended low water period resulted in a strong vegetation response and plots were dominated by smartweeds (Polygonum spp.), millets (Echinochloa sp.), and flatsedges (Cyperus spp.) (Table 2). Conversely, in 2000 water levels remained high into July because of local operating procedures dictated by system-wide low discharge (Fig. 2 b). Thus, mud flats were exposed for only a short time (~20 d) during 2000 and little to no vegetation developed or was present during sampling. During 2001, intermediate conditions produced a moderate plant response dominated by pigweed (Amaranthus sp.) and flatsedges (Cyperus spp.) (Fig. 2 c, Tables 1, 2).


195

Fig. 2. Actual water levels above sea level (a. s. l.) during 1999 (A), 2000 (B), and 2001 (C) and depictions of water level management (WLM) targets in Mississippi River Pool 25.

Responses to emergent vegetation Organic matter

During 2000, following the large vegetation response of 1999, there was significantly higher total organic matter in the vegetated plots (P = 0.008; Fig. 3), even though there was no vegetation response during 2000. Significantly higher CPOM in the vegetated plots (P = 0.004) was the primary source of this difference, and there was no difference in FPOM between plots in 2000. In 2001, following no vegetation response in 2000, total organic matter was similar between plot types and CPOM was marginally higher in vegetated plots (P = 0.087; Fig. 3). There was no difference in FPOM between plot types in 2001.

196

M. B. Flinn et al.

Fig. 3. Benthic organic matter in vegetated and devegetated sites in Pool 25 of the Mississippi River (n = 5 of each in 2000, n = 6 of each in 2001). Bars represent mean organic matter [g ash-free dry mass (AFDM)/m2] ± 1 S. E. for total, coarse (CPOM), and fine (FPOM) particulate fractions. Different letters over 2000 values indicate significant differences between total organic matter (P = 0.009) and CPOM (P = 0.004). Different letters over 2001 CPOM values indicate a marginally significant difference (P < 0.087).

Macroinvertebrates

During 2000 and 2001, there were no significant differences in total macroinvertebrate abundances between treatments (Table 3), but there were numerous differences among individual taxa. Oligochaetes, mostly Tubificidae, Naididae, and Lumbriculidae, were the most abundant macroinvertebrates in plots during both 2000 and 2001 and there were no significant differences in oligochaete abundance between treatments during the two years (Table 3). In contrast, Chironomidae, which were the second most abundant taxon, were 2.5 × more abundant in devegetated plots compared to vegetated plots in 2001 when there was a vegetation response (P = 0.008; Table 3). Several chironomid genera also showed significant differences between plot types during 2001; Chironomus (P = 0.043), Cryptochironomus (P = 0.0159), Polypedilum (P = 0.0086), and Cladotanytarsus (P = 0.0204) all had higher abundance in the devegetated plots (Table 3). Several non-insect groups other than oligochaetes were also abundant and showed variable responses to vegetation manipulations. For example, physid snails (Physa/Physella) and ostracods were most abundant in vegetated plots during 2001 (P = 0.028 and P = 0.002, respectively), whereas the leech Helobdella was more abundant in devegetated plots (P = 0.0396) (Table 3). Many less common taxa showed non-significant trends


197

Table 3. Total macroinvertebrate abundance [no./m2 (1 S. E.)] at devegetated and vegetated sites in Mississippi River pool 25 during fall of 2000 and 2001. Numbers represent mean taxa abundance with standard error in parentheses (n = five sites in 2000 and six in 2001) (* denotes significance between plot types within year at P < 0.05). Dashes indicate absence of invertebrates. Note that column totals may not sum because rare taxa (those found in two or less sites and with low abundance) were not included. 2000 Taxa

Devegetated

BRYOZOA Plumatella sp. – NEMATODA 130 (78) OLIGOCHAETA 23466 (3917) HIRUDINEA Helobdella sp. 38 (31) GASTROPODA Physidae Physa/Physella sp. 2 (2) BIVALIVIA Sphaeridae Sphaerium sp. 19 (15) HYDRACHNIDA 30 (24) INSECTA Hemiptera Corixidae Trichocorixa sp. 192 (97) Coleoptera Hydrophilidae Berosus sp. 2 (2) Diptera Ceratopogonidae Bezzia/Palpomyia sp. 21 (17) Chironomidae 5088 (1083) Ablebesmyia sp. 4 (3) Axarus sp. 147 (94) Chironomus sp. 3586 (978) Cladotanytarsus sp. 183 (106) Coelotanypus sp. 190 (106) Cryptochironomus sp. 90 (35) Dicrotendipes sp. 115 (69) Polypedilum sp. 228 (65)* Procladius sp. 414 (152) Tanypus sp. 66 (36) Tanytarsus sp. 32 (17) OSTRACODA 3187 (1356) Total Taxa abundance

2001 Vegetated

Devegetated

Vegetated

– 46 (46) 124 (119) – 89 (46) 7139 (381) 26530 (4497) 33861 (6342) 29735 (5261) 49

(32)

155

(70)*

2

(2)

30

(28)*

34 2

(20) (2)

– –

108

(53)

19

(17)

28 4708 21 4 3392 102 228 81 384 60 339 60 26 5314

(20) (975) (17) (4) (761) (34) (143) (28) (275) (25)* (173) (49) (26) (1844)

473 (247)

43

67 4514 – – 1168 194 7 101 144 1130 101 19 1607 796

(18)

46

(22)*

618 (418)*

5 16

(3) (16)

555 (168)

160

(38) 44 (1302)* 1678 – – (318)* 697 (121)* 30 (4) 4 (39)* 62 (99) 405 (434)* 210 (46) 181 (14) 20 (983)* 57 (415)* 10364

(62)

(27.9) (371)*

(210)* (28)* (2) (42)* (213) (84)* (89) (13) (57)* (4242)*

32393 (4580) 37600 (6266) 42158 (7930) 48330 (9933)

198

M. B. Flinn et al.

Fig. 4. Total biomass of total macroinvertebrates (A), Oligochaeta (B), and Chironomidae (C) in vegetated and devegetated sites in Pool 25 of the Mississippi River (n = 5 of each in 2000, n = 6 of each in 2001). Bars represent mean invertebrate dry mass (DM) ± 1 S. E. Different letters above 2001 values indicate a significant difference between treatments (A, P = 0.001; B, P = 0.009; C, P = 0.005).

of higher abundance in vegetated plots, including Odonata (~9 × higher), Berosus (Coleoptera: Hydrophilidae) larvae (~4 × higher), and Callibaetis (Ephemeroptera: Baetidae) (~39 × higher). Total macroinvertebrate biomass was similar between vegetated and devegetated treatments during 2000 (Fig. 4 a). However, during 2001 when vegetation was present, macroinvertebrate biomass was ~2 × higher in vegetated plots (P = 0.001; Fig. 4 a), primarily because of higher Oligochaeta biomass (P = 0.009; Fig. 4 b). In contrast to oligochaetes, Chironomidae biomass was ~2 × higher in the devegated plots during 2001 (P = 0.005; Fig. 4 c), mostly because of higher biomass of Chironomus, Polypedilum, and Tanytarsus (Fig. 5). Consistent with abundance patterns in 2001, there was significantly higher ostracod (P = 0.001) and Physa/Physella (P = 0.03) biomass in vegetated plots.


199

Fig. 5. Chironomidae genera biomass in vegetated and devegetated sites in Pool 25 of the Mississippi River (n = 5 of each in 2000, n = 6 of each in 2001). Bars represent mean biomass (mg DM/m2) ± 1 S. E. Different letters over 2001 values indicate a significant difference between treatments (A, P = 0.07; B, P = 0.003; C, P = 0.003).

Macroinvertebrate community metrics showed differences between treatments. Although taxa richness was similar between plot types during both years, diversity was significantly higher in the vegetated plots (H′ = 1.02) compared to devegetated plots (H′ = 0.75) during 2001 (P = 0.006; Table 4). Likewise, dominance, calculated as combined fraction of community of Oligochaeta and Chironomidae, was significantly higher in devegetated plots (0.90) compared to vegetated (0.73) in 2001 (P < 0.0001; Table 4).

200

M. B. Flinn et al.

Table 4. Macroinvertebrate community metrics for devegetated and vegetated sites in Pool 25, Mississippi River in 2000 (n = 5) and 2001 (n = 6). Numbers represent mean ( ± S. E). Numbers with asterisks are significantly different between plots within a year at P < 0.05. 2000 Devegetated Richness 9.67 Shannon Index (H′) 0.88 Dominance 0.88

(0.80) (0.10) (0.04)

2001 Vegetated

Devegetated

Vegetated

8.73 (0.73) 0.76 (0.06) 0.84 (0.05)

9.72 (0.69) 0.75* (0.08) 0.90* (0.02)

10.22 (0.93) 1.02* (0.08) 0.73* (0.04)

Fig. 6. Zooplankton abundance in vegetated and devegetated sites in Pool 25 of the Mississippi River (n = 4 of each in 1999, n = 5 of each in 2000, n = 6 of each in 2001). Bars represent mean zooplankton abundance (number of individuals/L) ± 1 S. E. Different letters over 1999 indicate a significant difference between treatments (P = 0.043).

Zooplankton

Total zooplankton abundance in 1999 was 1.6 × higher in the vegetated plots than the devegetated plots (P = 0.04; Fig. 6), mostly due to higher abundance of the two most abundant cladocerans, Chydoridae (7 ×, P = 0.049) and Sididae (2 ×, P = 0.03), and cyclopoid copepods (1.6 ×, P = 0.056) in vegetated plots. However, during 2000 (no vegetation response) and 2001 (moderate vegetation response) there were no significant differences in total zooplankton abundance or abundances of individual taxa between plot types.


201

Discussion Our results indicate that emergent vegetation associated with WLM influenced invertebrate communities in floodplain habitats of the upper Mississippi River. During this multiyear investigation, we were able to quantify invertebrate responses to different amounts of vegetation because of annual hydrologic variability resulting from interactions between river discharge and WLM. We feel that the experimental manipulation (vegetation removal) did approximate conditions without an extended drawdown. The interaction between hydrology, vegetation, and invertebrates is complex. However, the relatively short window for successful vegetation germination and production, invertebrate life history, and the natural variability in hydrology may lessen potentially confounding factors such as sediments drying or differences in quantity and quality of residual organic matter. Analysis of basic water chemistry revealed no major differences within years. This study encompassed three years in which considerably different hydrologic regimes prevailed, providing insight on different river conditions. However, we also recognize the limitations of the manipulation. For one, the scale of the manipulations may not encompass all interactions, and extreme events like major floods or droughts likely fall outside the scope of our study. The USACE operating procedures are based on discharge and water levels, and WLM can only be implemented when these two factors allow. Additionally, the WLM schedule may be flexible in that given favorable conditions for water level manipulation, different water levels and drawdown periods may be used to achieve management goals. Benthic organic matter in vegetation manipulation plots

The distribution of benthic organic matter can be an important determinant of macroinvertebrate diversity and productivity in freshwaters (e. g., Egglishaw 1964, Whiles & Wallace 1995, Silver et al. 2000, Baer et al. 2001), and some important patterns were evident during this study. Despite a lack of vegetation response in 2000, differences in benthic organic matter were evident during this year. These differences were probably a result of residual organic matter from the strong vegetation response during 1999. This is an important finding because it suggests that vegetation responses in floodplain habitats during a given year may influence food and habitat resources for macroinvertebrates during subsequent years. Although residual organic matter may serve as an important resource for macroinvertebrates and other groups, we did not observe substantial differences in macroinvertebrate communities between vegetated and devegetated treatments during 2000, indicating residual organic matter in some plots was

202

M. B. Flinn et al.

not having a substantial influence on macroinvertebrate distributions, at least at the scale of our study. This may be because benthic organic matter did not offer the structure and complexity of standing emergent vegetation, and thus may not have influenced macroinvertebrate communities as much as standing vegetation that was present during other years. Further, collector-gatherers dominated macroinvertebrate communities in these habitats and differences would likely be most evident where there were differences in FPOM, their primary food source (e. g., Wallace et al. 1995, Baer et al. 2001). Throughout this study, FPOM showed the least variability in space and time. For example, average combined treatment FPOM values differed by < 1.0 g between 2000 and 2001. Hence, FPOM is much less spatially and temporally variable in this system compared to coarser particles and is less likely a limiting resource in this system. Despite the lack of an apparent link between macroinvertebrates and benthic CPOM in our plots, benthic CPOM produced through WLM is still potentially important in this system, particularly where interactions between the main channel and floodplain are now limited by anthropogenic modifications. Organic matter associated with WLM is available both locally and for export during high water events and should enhance overall energy and habitat resources. This is particularly important in this system and other large rivers that have lost so much of their original floodplain connections and associated sources of allochthonous organic materials (e. g., Junk et al. 1989). Macroinvertebrate responses to vegetation

Patterns of macroinvertebrate abundance and biomass in vegetated and devegetated plots appeared different between 2000 and 2001, and this was likely a product of large differences in river flows that resulted in very different vegetation responses between these years. In 2000, when hydrologic conditions did not allow for WLM and associated vegetation production, total macroinvertebrate abundance and biomass did not differ across treatments and there were no differences in individual taxa, including the Chironomidae and Oligochaeta that dominated communities. This complete lack of differences between treatments in 2000 was most likely related to the overall absence of a vegetation response, which resulted in both types of plots lacking vegetation. In effect, both types of plots were devegetated plots during 2000 because river hydrology precluded successful WLM. In contrast to 2000, differential responses to the presence of vegetation were evident among major macroinvertebrate groups in 2001, although total macroinvertebrate abundance and Oligochaeta abundance still did not differ across treatments. Similarities in oligochaete abundance in devegetated and control plots are likely related to similar FPOM resources between treatments


203

because oligochaete taxa in our sites (mostly Tubificidae and Lumbriculidae) are primarily benthic collector-gatherers (Smith 2001). Oligochaetes dominated total abundance in 2001, outnumbering other taxa by more than 10 ×, and thus obscured other differences in macroinvertebrate abundances between treatments. Despite similar abundances, Oligochaeta biomass was significantly higher in vegetated plots in 2001, indicating that generally larger oligochaetes were present in vegetated habitats. This may be a result of taxonomic differences in oligochaete communities in the two habitats, a factor we did not examine at a fine enough scale to accurately assess. Alternatively, oligochaetes could live longer, and thus grow larger, in vegetated sites due to fewer disturbances and/ or reduced predation associated with plant beds (e. g., Beckett et al. 1992, Chen & Barko 1988, Ságová-Marecková 2002). Chironomids, the second dominant macroinvertebrate group in this system, showed variable responses to vegetation. Three genera, Chironomus, Polypedilum, and Tanytarsus dominated midge communities. These three genera have been associated with a variety of habitats (Epler 2001), and possibly a species-level preference for devegetated areas could explain patterns we observed. Gilinsky (1984) observed increased abundance of several chironomid groups in sediments not covered by macrophytes and suggests that migration onto the plants may occur when available. However, this was not the case in our study because comparisons were made between samples that contained the associated vegetation in the same habitat. Higher predation by other macroinvertebrates in vegetated plots seems unlikely since these genera of chironomids burrow in sediments, and most macroinvertebrate predators observed in the vegetated plots (e. g., many odonates) were clingers or climbers associated with the vegetation. Further, predaceous macroinvertebrates were not abundant in our vegetated plots, constituting < 1 % of total abundance and < 1 % of total biomass during 2001. Fish predation could also reduce chironomid abundance in vegetated plots, but this explanation also seems unlikely because vertebrate predator effectiveness is often lower in dense vegetation due to physical interference (Hershey 1985). Abundance of fishes, including predatory species, was higher in our vegetated plots than in our devegated plots (Garvey et al. 2003). These fishes were likely feeding in the vegetated plots, but were obviously not suppressing macroinvertebrate populations, at least compared to adjacent devegetated areas. Differences in community metrics (e. g., Shannon diversity) between plot types during 2001 suggested that macroinvertebrates responded to structural habitat diversity associated with emergent vegetation in control plots. Kedzierski & Smock (2001) suggest that higher abundance and production of macroinvertebrates in an area of a low gradient stream with abundant macrophytes was due to increased surface area. There was a trend of higher numbers

204

M. B. Flinn et al.

of clinger-climber taxa such as Odonata, Berosus, and Callibaetis in vegetated plots compared to devegetated plots, and these taxa contributed to higher evenness and reduced dominance, and thus higher diversity, in the vegetated plots during 2001. Although these taxa were not dominant, their responses are important because they represent increased diversity in the type and size classes of macroinvertebrate prey available to predators such as fish, wading birds, and waterfowl. Further, these taxa represent prey items that are available in different microhabitats (e. g., on vegetation in the water column vs. buried in sediments). Although we treated sites as replicates, some differences between sites became evident during this study and may have influenced some results. For example, Jim Crow slough had consistently high taxonomic richness in the vegetated plot (mean = 15.6) compared to vegetated plots at the other sites (mean = 9.2). This site appeared unique and in some analyses it may have disproportionately influenced our results. One unique characteristic of Jim Crow compared to the other sites is that it becomes disconnected from the main channel before others, but then retains water for an extended period of time during summer WLM. Along with responses to vegetation associated with water level management, we also observed evidence that hydrology directly influenced floodplain macroinvertebrate communities during this study. For example, during 2000 there was no vegetation and high water inundated the study sites for most of the growing season. This longer hydroperiod likely influenced macroinvertebrate communities in a variety of ways, such as allowing larger, longer-lived taxa to successfully complete life cycles. For example, higher overall chironomid biomass in 2000 (~2 × higher) compared to 2001 may have been a product of the relatively long period of uninterrupted inundation that allowed midges to grow larger and/or favored larger, longer-lived midge taxa. In a prior study in Mississippi River floodplain ponds, Corti et al. (1997) found differences in macroinvertebrate communities associated with hydroperiod length, including high overall abundance in ponds with relatively low permanence (dried 4 – 5 times throughout the year), and larger, longer-lived taxa in ponds with higher permanence (permanent or dried once). Zooplankton responses to vegetation

Zooplankton responded to vegetation over the gradient of conditions produced during the 3 years they were sampled. In 2000 (no vegetation response) there were no differences in zooplankton abundance, in 2001 (moderate vegetation) there was a non-significant trend of higher abundance in devegetated plots, and in 1999 (strong vegetation response) zooplankton abundance was significantly higher in vegetated plots. These results indicate that relationships be-


205

tween zoolplankton abundance and emergent vegetation may be complex, but that very dense vegetation may be most effective for promoting zooplankton populations in backwaters. Patterns of zooplankton abundance in 1999 seem counterintuitive based on differences in dissolved oxygen between plot types. Spot checks during sampling events indicated that dissolved oxygen was often low (mean = 2.5 mg L –1) in the vegetated treatments during 1999, presumably because the exceptionally dense vegetation during this year limited wave action, water movements, and light penetration. However, generally lower turbidity in vegetated plots (mean 43 NTU) compared to devegetated plots (mean 63 NTU) during 1999 may have also influenced zooplankton abundances in plots as suspended fine clay and silt particles can impede zooplankton feeding if they overwhelm algal particles in the water (Threlkeld 1986, Kirk 1992). Additionally, Chydoridae and Cyclopoidea were the two most abundant taxa and were thus primarily responsible for differences between plot types. Chydorids are typically not planktonic per se, but move along surfaces where they scrape or filter detrital particles (Dodson & Frey 1991) and shading effects of the 1999 vegetation may not have affected chydorids. Likewise, many cyclopoid copepods are carnivorous (Williamson 1991), and thus would also be unaffected by low phytoplankton abundance from shading. Regardless of causal mechanisms, the positive response of zooplankton to dense vegetation during 1999 indicates that yet another important food source for fishes, particularly larval and juvenile fishes, which were more abundant in the vegetated plots during this study (Garvey et al. 2003), may be influenced by emergent vegetation associated with WLM. Management of the Upper Mississippi River

Water-level management in the UMR, as well as in any other regulated river, has ecosystem-wide implications, as hydrology of the river influences vegetation and animal communities. In the context of a highly managed and degraded system like the UMR where hydrologic variability was once the norm, management plans that include a summer drawdown, such as WLM, appear to be an effective management strategy that more closely approximates the natural hydrologic regime of this regulated river. Among the many potential benefits of restoring a more natural hydrologic regime, results of this study indicate that the vegetation associated with WLM can influence organic matter, macroinvertebrates and zooplankton, and thus food resources for target management species such as fishes and waterfowl. However, our results also demonstrate the importance of spatial heterogeneity in off-channel habitats because many taxa, including some dominant groups such as chironomids, responded positively to devegetation. Further, both vegetation and invertebrate communities

206

M. B. Flinn et al.

varied substantially with the different hydrologic conditions of each study year, indicating that inter-annual hydrologic variability might also enhance biological diversity in backwater habitats on a temporal scale. The USACE’s goal for WLM to enhance fish and wildlife habitat by promoting development of emergent vegetation appears to be met when river hydrology allows for it (e. g., 1999 and 2001 during this study) and could prove beneficial for the UMR and other large river floodplain ecosystems. Acknowledgements The St. Louis District of the United States Army Corps of Engineers funded this research. We thank those who provided valuable field and lab assistance including E. Keyser, S. Peterson, M. Venarsky, D. Butler, K. Bires, K. Smith, K. Emme, J. Hiebert, and G. Adams. We would also like to thank the suggestions and comments of fellow lab personnel and two anonymous reviewers.

References Baer, S. G., Siler, E. R., Eggert, S. L. & Wallace, J. B. (2001): Colonization and production of macroinvertebrates on artificial substrata: upstream-downstream responses to a leaf litter exclusion manipulation. – Freshwat. Biol. 46: 347– 365. Beckett, D. C., Aartila, T. P. & Miller, A. C. (1992): Contrasts in density of benthic invertebrates between macrophyte beds and open littoral patches in Eau Galle Lake, Wisconsin. – Amer. Midl. Nat. 127: 77– 90. Benke, A. C., Huryn, A. D., Smock, L. A. & Wallace, J. B. (1999): Length-mass relationships for freshwater macroinvertebrates in North America with particular reference to the southeastern United States. – J. N. Amer. Benth. Soc. 18: 308 – 343. Bottrell, H. H., Duncan, A., Gliwicz, Z. M., Grygierek, E., Herzig, A., Hillbricht-Ilkowska, A., Kurasawa, H., Larsson, P. & Weglenska, T. (1976): A review of some problems in zooplankton production studies. – Norwegian J. Zool. 24: 419 – 456. Boulton, A. J., Peterson, C. G., Grimm, N. B. & Fisher, S. G. (1992): Stability of an aquatic macroinvertebrate community in a multiyear hydrologic disturbance regime. – Ecology 73: 2192 – 2207. Brodersen, K. P., Odgaard, B. V., Vestergaard, O. & Anderson, N. J. (2001): Chironomid stratigraphy in the shallow and eutrophic Lake Søbygaard, Denmark: chironomid-macrophyte co-occurrence. – Freshwat. Biol. 46: 253 – 267. Campeau, S., Murkin, H. R. & Titman, R. D. (1994): Relative importance of algae and emergent plant litter to freshwater marsh invertebrates. – Can. J. Fish. Aquat. Sci. 51: 681– 692. Chen, R. L. & Barko, J. W. (1988): Effects of freshwater macrophytes on sediment chemistry. – J. Freshwat. Ecol. 4: 279 – 289. Chen, Y. H. & Simons, D. B. (1986): Hydrology, hydraulics, and geomorphology of the Upper Mississippi River System. – Hydrobiologia 136: 5 – 20.


207

Corti, D., Kohler, S. L. & Sparks, R. E. (1997): Effects of hydroperiod and predation on a Mississippi River floodplain invertebrate community. – Oecologia 109: 154–165. De Szalay, F. A. & Resh, V. H. (2000): Factors influencing macroinvertebrate colonization of seasonal wetlands: responses to emergent plant cover. – Freshwat. Biol. 45: 295 – 308. Diehl, S. (1992): Fish predation and benthic community structure: the role of omnivory and habitat complexity. – Ecology 73: 1646–1661. Dodson, S. I. & Frey, D. G. (1991): Cladocera and other Branchiopoda. – In: Thorp, J. H. & Covich, A. P. (eds): Ecology and Classification of North American Freshwater Macroinvertebrates. – Academic Press. Inc., pp. 723–786. Egglishaw, H. J. (1964): The distributional relationship between the bottom fauna and plant detritus in streams. – J. Anim. Ecol. 33: 463 – 476. Epler, J. H. (2001): Identification Manual for the Larval Chironomidae (Diptera) of North and South Carolina. A guide to the taxonomy of the midges of the southeastern United States, including Florida. – Special Publication SJ2001-SP13. North Carolina Department of Environment and Natural Resources, Raleigh, NC, and St. Johns River Water Management District, Palatka, FL. 526 pp. Fredrickson, L. H. & Taylor, T. S. (1982): Management of seasonally flooded impoundments for wildlife. – U. S. Department of the Interior Fish and Wildlife Service, Resource Publication, 148, 27 pp. Fremling, C. R. & Claflin, T. O. (1984): Ecological history of the Upper Mississippi River. – In: Wiener, J. G., Anderson, R. V. & McConville, D. R. (eds): Contaminants in the Upper Mississippi River. – Proceedings of the 15th Annual Meeting of the Mississippi River Research Consortium, Butterworth Publishers, Boston, pp. 5 – 24. Gardner, S. & Grue, C. (1996): Effects of Rodeo and Garlon 3 A on Nontarget Wetland Species In Central Washington. – Environ. Toxicol. Chem. 15: 441– 451. Garvey, J. E., Dugger, B. D., Whiles, M. R., Adams, S. R., Flinn, M. B., Burr, B. M. & Sheehan, R. J. (2003): Responses of fish, waterbirds, invertebrates, vegetation, and water quality to environmental pool management: Mississippi River Pool 25. – Final report submitted to the U. S. Army Corps of Engineers, St. Louis District, 180 pp. Geisy, J. P., Dobson, S. & Solomon, K. R. (2000): Ecotoxicological Risk Assessment for Roundup Herbicide. – Rev. Environ. Contam. Toxicol. 167: 35–120. Gilinsky, E. (1984): The role of fish predation and spatial heterogeneity in determining benthic community structure. – Ecology 65: 455 – 468. Grubaugh, J. W. & Anderson, R. V. (1988): Spatial and temporal availability of floodplain habitat: long-term changes in Pool 19, Mississippi River. – Amer. Midl. Nat. 199: 402 – 411. Hargeby, A. (1990): Macrophyte associated invertebrates and the effect of habitat permanence. – Oikos. 57: 338 – 346. Hershey, A. E. (1985): Experimental studies on predatory sculpin on the chironomid communities in an arctic lake. – Ecology 66: 1131–1138. Junk, W. J., Bayley, P. B. & Sparks, R. E. (1989): The flood-pulse concept in riverfloodplain systems. – In: Dodge, D. P. (ed.): Internat. Large River Symp. – Can. Spec. Publ. Fish. Aquat. Sci. 106: 110–127.

208

M. B. Flinn et al.

Kedzierski, W. M. & Smock, L. A. (2001): Effects of logging on macroinvertebrate production in sand-bottomed, low-gradient stream. – Freshwat. Biol. 46: 821– 833. Kirk, K. L. (1992): Effects of suspended clay on Daphnia body growth and fitness. – Freshwat. Biol. 28: 103–109. Lubinski, K. S., Carmody, G., Wilcox, D. & Drazkowski, B. (1991): Development of water level regulation strategies for fish and wildlife, Upper Mississippi River System. – Regul. Riv.: Resear. Managem. 6: 117–124. Merritt, R. W. & Cummins, K. W. (eds) (1996): An Introduction to the Aquatic Insects of North America, 3rd ed. – Kendall-Hunt Publishing, Dubuque, IA. Murkin, H. R. & Kadlec, J. (1986): Responses of benthic macroinvertebrates to prolonged flooding of marsh habitat. – Can. J. Zool. 64: 65–72. Neckles, H. A., Murkin, H. R. & Cooper, J. A. (1990): Influences of seasonal flooding on macroinvertebrate abundance in wetland habitats. – Freshwat. Biol. 23: 311– 322. Olson, E. J., Engstrom, E. S., Doeringsfeld, M. R. & Bellig, R. (1995): Abundance and distribution of macroinvertebrates in relation to macrophyte communities in a prairie marsh, Swan Lake, Minnesota. – J. Freshwat. Ecol. 10: 325 – 334. Parsons, J. K. & Matthews, R. A. (1995): Analysis of the associations between macroinvertebrates and macrophytes in a freshwater pond. – Northwest Sci. 69: 265 – 275. Patrick, R. (1998): Rivers of the United States. Vol IV, Part A, The Mississippi River and Tributaries North of St. Louis. – John Wiley and Sons, Inc. New York, NY. USA. Pennak, R. W. (1962): Quantitative zooplankton sampling in littoral vegetation areas. – Limnol. Oceanogr. 7: 487– 489. Poff, N. L., Allan, J. D., Bain, M. B., Karr, J. R., Prestegaard, K. L., Richter, B. D., Sparks, R. E. & Stromberg, J. C. (1997): The natural flow regime: a paradigm for conservation and restoration of river ecosystems. – Bioscience 47: 769 – 784. Rogers, L. E., Hinds, W. T. & Buschom, R. L. (1976): A general weight vs. length relationship for insects. – Ann. Entomol. Soc. Amer. 69: 387– 389. SAS Institute (2001): JMP Statistical Software for Windows 4.0.3. – SAS Institute, Cary, North Carolina, USA. Ságová-Marecková, M. (2002): Distribution of benthic macroinvertebrates in relationship to plant roots, sediment type and spatial scale in fishponds and slow streams. – Arch. Hydrobiol. 156: 63 – 81. Schoener, T. W. (1980): Length-weight regressions in tropical and temperate forestunderstory insects. – Ann. Entomol. Soc. Amer. 73: 106–109. Schramm Jr., H. L. & Jirka, K. J. (1989): Effects of aquatic macrophytes on benthic macroinvertebrates in two Florida lakes. – J. Freshwat. Ecol. 5: 1–12. Sheehan, R. J. & Konikoff, M. (1998): Wetlands and fisheries resources of the Mississippi River. – In: Majumdar, S. K., Miller, E. W. & Brenner, F. J. (eds): Ecology of wetlands and associated systems – Pennsylvania Academy of Science, pp. 628 – 647.


209

Shriver, P., Bøgestrand, J., Jeppesen, E. & Søndergaard, M. (1995): Impact of submerged macrophytes on fish-zooplankton-phytoplankton interactions: largescale enclosure experiments in a shallow eutrophic lake. – Freshwat. Biol. 33: 255 – 270. Silver, P., Cooper, J. K., Palmer, M. A. & Davis, E. J. (2000): The arrangement of resources in patchy landscapes: effects on distribution, survival, and resource acquisition of chironomids. – Oecologia 124: 216 – 224. Smith, D. G. (2001): Pennak’s Freshwater Invertebrates of the United States, 4th Ed. – Porifera to Crustacea. – John-Wiley & Sons, Inc. New York, NY. Smock, L. A. & Stoneburner, D. L. (1980): The response of macroinvertebrates to aquatic macrophyte decomposition. – Oikos 35: 397– 403. Sparks, R. E. (1995): Need for ecosystem management of large rivers and their floodplains. – Bioscience 45: 169–182. Sparks, R. E., Nelson, J. C. & Yin, Y. (1998): Naturalization of the flood regime in regulated rivers, the case of the upper Mississippi River. – Bioscience 48: 706 – 720. Stagliano, D. M., Benke, A. C. & Anderson, D. H. (1998): Emergence of aquatic insects from 2 habitats in a small wetland of the southeastern USA: temporal patterns of numbers and biomass. – J. N. Amer. Benth. Soc. 17: 37– 53. Theiling, C. H., Maher, R. J. & Sparks, R. E. (1996): Effects of variable annual hydrology on a river regulated for navigation: Pool 26, Upper Mississippi River System. – J. Freshwat. Ecol. 11: 101–114. Threlkeld, S. T. (1986): Life table responses and population dynamics of four cladoceran zooplankton during a reservoir flood. – J. Plankton Res. 8: 639 – 647. Voigts, D. K. (1976): Aquatic invertebrate abundance in relation to changing marsh vegetation. – Amer. Midl. Nat. 95: 313 – 322. Wallace, J. B., Webster, J. R. & Meyer, J. L. (1995): Influence of log additions on physical and biotic characteristics of a mountain stream. – Can. J. Fish. Aquat. Sci. 52: 2120 – 2137. Whiles, M. R. & Goldowitz, B. S. (2001): Hydrologic influences on insect emergence production from central Platte River wetlands. – Ecol. Appl. 11: 1829–1842. Whiles, M. R. & Wallace, J. B. (1995): Macroinvertebrate production in a headwater stream during recovery from anthropogenic disturbance and hydrologic extremes. – Can. J. Fish. Aquat. Sci. 52: 2402 – 2422. Williams, G. M., Kroes, R. & Munro, I. C. (2000): Safety Evaluation and Risk Assessment of the Herbicide Roundup and Its Active Ingredient, Glyphosate, for Humans. – Reg. Tox. Pharm. 31: 117–165. Williamson, C. E. (1991): Copepoda. – In: Thorp, J. H. & Covich, A. P. (eds): Ecology and Classification of North American Freshwater Macroinvertebrates. – Academic Press, Inc., pp. 787– 822. Wlosinski, J. H. & Hill, L. (1995): Analysis of water level management on the Upper Mississippi River. – National Biological Service, Environmental Management Technical Center, Onalaska, Wisconsin, May 1995. LTRMP 95-T001. 43 pp. Wlosinski, J. H. & Rogala, J. (1996): Pool 25: water-level management alternatives and their effects on habitat. – National Biological Service, Environmental Management Technical Center, Onalaska, Wisconsin, LTRMP 96-T004, 84 pp. Woltemade, C. J. (1997): Water level management opportunities for ecological benefit, Pool 5 Mississippi River. – J. Amer. Wat. Res. Assoc. 33: 443 – 454.

210

M. B. Flinn et al.

Wrubleski, D. A. (1989): The effect of waterfowl feeding on a chironomid (Diptera: Chironomidae) community. – In: Sharitz & Gibbons, J. W. (eds): Freshwater Wetlands and Wildlife. – CONF-8603101, DOE Symposium Series No. 61. Submitted: 4 May 2004; accepted: 17 September 2004.