2008 NORTHEASTERN NATURALIST 15(3):391–402 Possible Effect of Lock and Dam 19 on Phytoplankton Communities of the Upper Mississippi River Susan T. Meiers1,*, Sean E. Jenkins1, and Richard V. Anderson1 Abstract - Algal community composition at three sites above and five sites below Lock and Dam 19 on the upper Mississippi River was determined from samples taken during July, August, September, and November of 2003, and April and June of 2004, to compare the above-dam navigation-channel and vegetation-bed phytoplankton communities, and determine what effect the power plant and lock and dam may have on the mixing of these communities. We predicted there would be clear differences in community composition and abundance between the three above-dam locations and a more uniform community composition below the dam due to mixing. No sitespecific or predictable patterns to suggest an effect of the power plant or lock and dam were detected. Nonmetric multidimensional scaling ordination supports the idea that temporal factors may play a larger role in structuring localized phytoplankton communities in this section of the upper Mississippi River. Introduction Over the last century, much of the upper Mississippi River was leveed and channelized, creating a system of locks and dams to improve commercial navigation (USGS 1999). These actions, along with agricultural, municipal, and industrial runoff, and the dredging of barge channels, have greatly contributed to the continued alteration of associated aquatic and riparian communities within the river system. The locks and dams and other structures, such as power plants, have changed the river from a free-flowing river to a system comprised of a series of pools that in some sections may be viewed as more lacustrine in nature, with accompanying changes in all trophic levels within the ecosystem (Fremling 2005). Few studies have explicitly examined the effect of locks and navigation dams on the composition and dynamics of phytoplankton communities in rivers. Reinhard (1931) examined the plankton abundance and seasonal distribution in the upper Mississippi River from Minneapolis, MN, to Winona, MN, but did not specifically examine how the several locks and dams may have affected their abundance or composition. Wehr and Thorp (1997) observed that Ohio River navigation dams had less effect on phytoplankton communities than did water chemistry, and observed no clear general pattern for the effect of dams on phytoplankton communities. They did suggest that even small changes, such as a decrease in current velocity, may cause benthic algae to settle out from the water column above the dam. While these changes may have 1 Western Illinois University, Department of Biological Sciences, 1 University Circle, Macomb, IL 61455. *Corresponding author - st-meiers@wiu.edu. 392 Northeastern Naturalist Vol. 15, No. 3 negligible effects on the total phytoplankton community, they may affect the distribution of benthic species above dams more than at other sites. A number of studies have examined phytoplankton in various reaches and habitats along the upper Mississippi River. Reinhard (1931) studied a number of sites in Pools 7 and 9, and observed only negligible differences in species composition above and below Lake Pepin. Baker and Baker (1979, 1981) studied phytoplankton in the main channel near Prairie Island, MN, but observed few differences in abundance or community composition among sites. Huff (1986) examined two main channel and two backwater sites above Lock and Dam 7 and demonstrated that main channel sites had greater total biomass during September and that diatoms were the most common algal group at all sites. Diatoms and cyanbacteria were the most common algae in main channel sites and were more abundant in these habitats than in backwater sites. In a study of mid-reach Pool 8, Lange and Rada (1993) observed no differences in community composition or biomass among a side channel, an embayment, and shallow marshy habitats, and attributed this homogeneity to abnormally high river discharge during the 1986 sampling. They also observed that diatoms were the most abundant group at all locations, reaching a maximum during summer, with Chlorophyta and cyanobacteria the next most abundant groups. The present study was conducted above and below Lock and Dam 19 on the upper Mississippi River adjacent to Keokuk, IA. Pool 19, above the lock and dam, has a diversity of aquatic habitats including main/navigation channel, main channel border, tailwaters, side channels, rivers, and sloughs (Jahn and Anderson 1986). An examination of phytoplankton communities from the (navigation) channel, vegetation bed, and side channel sites of Pool 19 demonstrated an abundance of pennate diatoms and Euglena sp. in vegetated sites as opposed to non-vegetated sites (Engman 1984). This study also revealed that the navigation channel had lower numbers of algae than sites where vegetation beds and wing dams are located. Lock and Dam 19 at Keokuk, IA, was completed in 1913 and the pooled water created Pool 19. The approximately 12-m drop in elevation allows water to generate power in the Keokuk Power Plant, which also regulates water level in the pool using a series of gates. Rock wing dams, located just below the dam, direct most of the flow into the navigation channel located on the Iowa side of the river (Jahn and Anderson 1986). In order to better understand the phytoplankton communities of lower Pool 19 and the effects that pool management are having on these communities, we compared the abundance and species composition of phytoplankton communities in the swift-flowing navigation channel and slow-flowing vegetation bed above Lock and Dam 19, and at sites below the dam. Spatial and temporal comparisons of these sites were used to determine if phytoplankton communities varied: 1) between the navigation channel and the middle of the vegetation bed above the dam; and 2) above and below the power plant and dam, to determine the extent of mixing of phytoplankton communities 2008 S.T. Meiers, S.E. Jenkins, and R.V. Anderson 393 through the power plant and over the dam. We predicted there would be clear differences in community composition and abundance between the two above-dam locations and a more uniform community composition below the dam due to mixing. Materials and Methods Sampling sites Above-dam sampling sites were chosen to compare two distinct habitats above Lock and Dam 19 (Fig. 1): 1) the navigation channel (Site 1), distinguished by greater (9 m) depth and faster flowing water, and 2) the middle of a dense vegetation bed (Site 3) directly east of Site 1, distinguished by shallow and slower-flowing water. Vallisneria americana Michx. (Water Celery), Stuckenia pectinata (L.) Böerner (Sago Pondweed), Nelumbo lutea Willd. (American lotus), Elodea canadensis Michx. (Waterweed or American Elodea), Potamogeton crispus L. (Curly Pondweed), Potamogeton nodosus Poir. (American Pondweed), Ceratophyllum demersum L. (Coontail), and Myriophyllum L. spp. (Water Milfoil) were common macrophyte species. Site 2, midway between Sites 1 and 3 at the edge of the vegetation bed, was sampled to address potential above-dam interaction. Three sites Figure 1. Map of research site with collecting sites (Sites 1–8 as “S1”, “S2”, etc.), power plant intake pipe (“I”), and power plant outlet or discharge pipe (“O”) indicated in relation to Keokuk, IA and Hamilton, IL. 394 Northeastern Naturalist Vol. 15, No. 3 directly below the dam were chosen as the most likely to detect if mixing of navigation channel and vegetation bed communities occurred: Site 4 was located directly below where the power plant releases its discharge, Site 6 was located in the dam tailwaters, and Site 5 was midway between Sites 4 and 6. Two further sites, approximately 0.3 km downriver, were chosen as well-mixed since exiting the power plant: Site 7 was in the middle of the navigation channel and Site 8 in an unvegetated channel border dike field. Samples were collected above and below Lock and Dam 19 during the third week of the following months: July and August 2003 (summer), September and November 2003 (fall), and April and June 2004 (spring). Winter samples (December through February) were not collected due to the presence of ice on the river. Field sampling Surface light levels (Extech Instruments Model 40123 light meter), current velocity (Global Flow Probe, Model FP101), water temperature, and water transparency (Secchi depth) were measured when phytoplankton were collected at each site, except where unsafe water conditions prevented it (Table 1). Surface phytoplankton samples were collected to reflect the water taken into the power plant intake pipe and the water which flows through open dam gates. All samples were preserved in 1% Lugol’s solution and returned to the laboratory. Lab processing For identification and enumeration, sample bottles were mixed well, 1.0 mL was placed in a Sedgwick-Rafter counting chamber using a Henson- Stemple pipette, and the cells allowed to settle for 20 min. Identification and counts were performed using strip counts from four random strips the width of a Whipple grid at 200x (Clescerl et al. 1999). All individuals, colonies, and filaments that were more than halfway into the grid were counted. Colonies, coenobia, and individuals of species that normally occur as multiple cells were counted as single units; filaments and chains of cells were counted only when 10 or more cells were present (Clescerl et al. 1999). Statistical analyses Nonmetric multidimensional scaling (NMS; Kruskal 1964) was used to compare phytoplankton composition among sampling sites and dates using PC-ORD Multivariate Analysis of Ecological Data version 4.25 (McCune and Grace 2002, McCune and Metford 1999); this method is especially useful with non-normal data having many zeros (Clarke 1993). The slow-and-thorough mode with a random starting configuration of NMS was used with the best of 40 runs on real data with a random starting configuration. Monte Carlo simulations with 50 runs of randomized data were conducted to assess the probability of the actual configuration. The proportion of the variance explained by each axis was determined utilizing r2 values between the ordination space and the original 2008 S.T. Meiers, S.E. Jenkins, and R.V. Anderson 395 Table 1. Table of water temperature, light intensity, current velocity, and Secchi depth measurements of each site. N/A = water level was too high and current was too fast to take physical measurements on this day. Water Current Secchi Light Month/site temp (°C) velocity (m/s) depth (m) intensity (klux) July 2003 109.7 Site 1 27 7.0 0.46 Site 2 26 2.5 0.05 Site 3 27 0.25 0.38 Site 4 26 9.5 N/A Site 5 26 17.5 N/A Site 6 26 18.5 N/A Site 7 26 15.5 N/A Site 8 26 5.5 N/A August 2003 40.6 Site 1 27 2.0 0.53 Site 2 27 0.5 0.70 Site 3 26 0.0 +0.60A Site 4 27 10.5 N/A Site 5 27 11.5 N/A Site 6 N/A N/A N/A Site 7 27 8.5 N/A Site 8 27 1 0.8 September 2003 75.6 Site 1 N/A 3.5 0.62 Site 2 N/A 3.0 0.40 Site 3 N/A 1.0 0.52 Site 4 N/A N/A 0.47 Site 5 N/A 11.0 N/A Site 6 N/A 13.5 N/A Site 7 N/A 14.0 N/A Site 8 N/A 13.0 N/A November 2003 107 .0 Site 1 6 4.2 0.72 Site 2 6 3.2 0.50 Site 3 5 3.2 0.31 Site 4 5 9.0 N/A Site 5 5 5.0 N/A Site 6 6 1.5 0.50 Site 7 6 4.5 0.50 Site 8 6 N/A N/A April 2004 36.0 Site 1 15 5.0 0.64 Site 2 15 4.0 0.60 Site 3 15 4.5 0.40 Site 4 15 8.6 0.45 Site 5 16 5.0 0.50 Site 6 15 3.0 N/A Site 7 16 5.0 0.46 Site 8 15 9.5 0.60 June 2004 26.0 Site 1 21 N/A N/A Site 2 22 N/A 0.41 Site 3 21 N/A 0.20 Site 4 22 N/A N/A Site 5 N/A N/A N/A Site 6 N/A N/A N/A Site 7 N/A N/A N/A Site 8 N/A N/A N/A ASecchi disk was still visible resting on the bottom sediment. 396 Northeastern Naturalist Vol. 15, No. 3 space distance (McCune and Metford 1999). To test for significant differences in algal abundances among sites (data non-normal), Kruskal-Wallis tests were performed on: a) cells/mL per site, b) number of taxa, and c) the total number of cells/mL for the 15 most abundant taxa. A one-way Friedman test on ranks was used to compare Simpson’s diversity values and sampling sites (repeated measures basis) among dates. Spearman’s rank correlations were performed to examine possible relationships between Simpson’s diversity values and date, site, and physical variables. Due to an inability to record very high current velocities, flow rates were ranked as: 1 = no flow, 2 = low flow, 3 = medium flow, and 4 = too high to measure. Results The 254 algal taxa identified were composed of 228 species, 21 varieties, one centric diatom category, and four unidentified categories (eukaryotic unicells, pennate diatoms, eukaryotic filaments, and cyanobacterial filaments). Diatoms and Chlorophytes were the most abundant taxa at most sites. No significant differences were observed for phytoplankton densities (cells/mL) among sites or over time (Fig. 2), and no significant differences were observed for the number of taxa per site or among dates (Fig. 2). Similarly, Simpson’s diversity indices also were not significantly different among sites (Fig. 3). Algal densities within each division (Table 2) indicate that Chlorophytes were most abundant at most sites, except Site 8, where diatoms were more common. The Simpson’s diversity indices for each month are presented in Figure 3. The Monte Carlo randomization test indicated that the NMS ordination was best fit using a two-axis solution (P < 0.05; Fig. 4). The initial NMS ordination indicated that two samples were outliers: Site 3 in July 2003 and Site 1 in August 2003. These two samples were removed before the final NMS analysis. The first two ordination axes accounted for 71.2% of the variability in the data (Axis 1 = 63.7%, Axis 2 = 23.9%), with a final stress and instability of 4.01 and 0.00001, respectively. Axis 1 of NMS ordination indicated Table 2. Total number of cells/mL per site, sorted by division, for sites above (Sites 1-3) and below (Sites 4-8) Lock and Dam 19. Number of cells/mL/site Division Site 1 Site 2 Site 3 Site 4 Site 5 Site 6 Site 7 Site 8 Chlorophyta 64,407.7 68,720.4 84,971.6 64,765.8 87,135.4 58,005.7 89,567.9 12,540.0 Cryptophyta 14.9 14.9 0.0 44.8 0.0 0.0 0.0 14.9 Cyanobacteria 1984.8 1537.1 1357.9 10863.9 1283.4 820.8 1343.1 1328.2 Dinophyta 0.0 14.9 0.0 0.0 0.0 0.0 0.0 14.9 Euglenophyta 313.4 164.2 104.5 74.6 208.9 89.5 343.2 298.5 Ochrophyta Diatoms 49,887.6 22,429.3 18,862.7 27,667.2 23,384.3 18,698.5 27,010.6 28,935.7 Other 298.5 119.4 119.4 208.9 134.3 74.6 89.5 149.2 Unknown 223.9 44.8 56.7 44.8 14.9 0.0 0.0 14.9 Note: 0= less than 14.9 cells/mL 2008 S.T. Meiers, S.E. Jenkins, and R.V. Anderson 397 marked compositional differences between September, April, and the other dates (Fig. 4). In September there was an increase of two orders of magnitude in densities of Chlamydomonas sp., a flagellated green alga. A number of taxa were found only during that same sample period, including Cosmarium Figure 2. Mean +1 SE for number of cells/mL, mean number of taxa, and Simpson’s diversity index for Sites 1 through 8 (abbreviated “S1,” “S2,” etc.) at Lock and Dam 19 on the Mississippi River. Sites 1–3 were above-dam sites, and Sites 4–8 were below-dam sites. 398 Northeastern Naturalist Vol. 15, No. 3 Figure 3. Mean + 1 SE for Simpson’s diversity indices for July, August, September, and November 2003, and April and June 2004. Figure 4. Results of nonmetric multidimensional scaling (NMS) ordination analyses of phytoplankton communities grouped by species. 2008 S.T. Meiers, S.E. Jenkins, and R.V. Anderson 399 subcrenatum Hantzsch., Tetraedron trigonum (Naegeli) Hansgirg, Surirella sp., and Aphanizomenon flos-aquae (L.) Ralfs. Echinosphaerella limnetica G.M. Smith and Micractinium sp. were also common at several sites during September. The ordination further indicated that April sites were tightly grouped, and all contained several unique taxa, including Closterium gracile Brebisson, Cocconeis placentula Ehrenberg, Dictyosphaerium pulchellum Wood, Rhoicosphenia curvata (Kuetzing) Grunow, and Synedra ulna (Nitz.) Ehrenberg. The April and September samples had the highest average species richness with values of 49 and 45, respectively. Most sites in November 2003 clustered close together, except Site 7, which clustered more closely with the July 2003 and June 2004 sites. Characteristic taxa for most sites on this sample date were Epithemia ocellata (Ehrenberg) Kuetzing and Eunotia pectinalis (Kuetzing) Rabenhorst. November had the third-highest average species richness with a value of 28. Even though on this date the third-highest average diversity was observed, there were few species unique only to this sampling date. The composition of July 2003 and June 2004 sites were indistinguishable by NMS. July 2003 and June 2004 sample sites also had very similar species composition, with roughly 23 species identified from each. The August samples were closely grouped in the lower left corner of the ordination, having the fewest number of centric diatoms as compared with all other sample sites and dates. These three sampling dates had much lower total abundances and species richness than November, September, and April samples, though these differences were not significant. A one-way Friedman test indicated a significant difference in Simpson’s diversity among dates (P < 0.001; Fig. 2). Species diversity values for communities from September and April were significantly different from all other dates, and from each other (P < 0.001). The correlations with environmental variables indicated that algal diversity decreased with time (r = -0.599, P = 0.00001) and was positively related to temperature (r = 0.402, P = 0.006) and velocity (r = 0.357, P = 0.015). Discussion No clear or predictable pattern was observed among sites for the dates sampled. In particular, no statistical differences in mean densities of algae were identified among sites, although significant differences were observed among some dates. Results from this study differed from those of Engman’s (1984) study of Pool 19 algal communities, which identified clear differences between the above-dam navigation-channel and vegetation-bed communities. The current study differed also in lacking a cyanobacterial peak in spring or summer (2.56% vs. Engman’s 11% of species composition). It is possible that peaks in the abundances of some species may have been missed if they occurred between sampling periods (March and May) in the current study. This discrepancy could also have been due to changes in current velocity through the vegetation bed resulting in a mixing of vegetation- bed phytoplankton with those in the navigation channel, which may have inhibited bloom-forming cyanobacteria. 400 Northeastern Naturalist Vol. 15, No. 3 Our results differed from the findings of Chandler (1937) and Baker and Baker (1981), who observed a lower abundance of phytoplankton in regions of heavy macrophyte vegetation. In the present study, Sites 1 and 7, both in the center of the navigation channel, had the highest mean number of taxa per site, which perhaps was due to greater discharge and turbulence causing resuspension of benthic algae into the water column (Luttenton and Rada 1986, Luttenton et al. 1986, Wehr and Thorp 1997). Overall, Sites 3, 4, and 7 had the greatest diversity levels, which was perhaps caused by increased complexity of the vegetation bed. A close grouping of sites for September in the NMS analysis was associated with low algal diversity and a high percentage of green algae (92%). In our study, Site 1 differed from the other sites, especially in August (Fig. 4), likely due to the large number of Chlamydomonas cells. This higher count could be due to nutrient release by the fall senescence of upriver vegetation beds and locally dense Lemna (duckweed) populations (S.E. Jenkins, unpubl. data). Because the intake of the power plant is approximately 15 m east of the lock entrance and the output is directly upstream of Site 4, it was predicted that the community composition and abundance of Sites 1 and 4 would be similar, which in general was observed. The non-significant decline in numbers of Chlamydomonas cells indicates that they apparently were unaffected by passage through the turbines and the approximately 12-m drop from the intake pipe to the turning turbines, which may also damage cells. The level of Pool 19 is managed so that sufficient water is taken in through the intake pipe, and thus a significant flow of water is always moving through the power plant. This flow may be the only way for water to move from above to below the dam at low river water levels and may have had significant effects on abundances of algae immediately below the dam. In addition, local planktivorous fauna including Hypophthalmichthys molitrix (Valenciennes in Cuvier and Valenciennes) (Silver Carp), H. nobilis (Richardson) (Bighead Carp), Polyodon spathula (Waldbaum) (Paddlefish) (Coker 1923, Lazzaro et al 1992), bivalves (Gale and Lowe 1971, Raidow and Hamilton 2001, Welker and Walz 1998), and other river invertebrates (Lamberti and Moore 1984, Madenjian 1995) known to congregate along the power plant, lock wall, and river bottom in this area (R.V. Anderson, unpubl. data) may also have contributed to the observed reduction of green algal density. The correlation between diversity and date supports the findings of the NMS ordination and one-way Friedman tests. The clustering of some of the June and July sites together in the NMS analysis is due to their sharing similar abundances of certain species that did not occur on other dates. This pattern may be the result of warmer temperatures and moderate availability of nutrients at this time (Grubaugh et al. 1986). Similarities among Sites 4–8 may be explained by the fact that water from above the dam passes through the power plant discharge or outlet pipe, or the output pipe and the dam gates. Downstream wing dams mix the water and forces the flow toward the Iowa side of the river to maintain navigation channel depth, and may homogenize phytoplankton communities at sites below the dam. Consistent with other studies on the upper Mississippi River (Engman 1986), these analyses indicated significant temporal differences, with 2008 S.T. Meiers, S.E. Jenkins, and R.V. Anderson 401 diatoms being the most prevalent on all dates except September, and a spring and fall increase in the overall abundance of algae. The higher densities of algal cells in September and April were largely associated with greater abundances of green algal taxa. The September increase in Chlamydomonas sp., Echinosphaerella limnetica, and Scenedesmus quadricauda may be due to the senescence and collapse of the vegetation bed at Site 3 andother beds farther upriver, and the seasonal demise of local Lemna populations in side channels and sloughs with a subsequent release of nutrients into the river. In a study of macrophyte beds that encompassed the upriver locations and Site 3 of the current study, Grubaugh et al. (1986) observed no increase in sediment organic matter in their macrophyte sites after Sagittaria (arrow-head) and Nelumbo senesced and decayed, suggesting that nutrients had been released into the water column downstream following senescence. Another potential source of nutrients may be agricultural runoff from fertilizer application in April–May and September–October. These events result in a pulse of nutrient rich runoff that is readily available to microorganisms, phytoplankton, and other river organisms (Grubaugh et al. 1986, Scholze 1994). Despite the fact that there are many habitats that could allow the development of significantly different phytoplankton communities, no clear or predictable phytoplankton spatial patterns were observed in the present study. The river above and below Lock and Dam 19 and its incorporated power plant is a complicated system whose algal communities are apparently more influenced by temporal factors than by physical differences in habitat. More intensive sampling at the intake pipe, the power plant output, and slightly downstream may reveal details as to the impact of the power plant turbines and local planktivores on survival of local phytoplankton communities. Acknowledgments The authors wish to thank H. Courtois and Earthwatch SCAP (Student Challenge Award Program) volunteers for their assistance collecting some of the samples used in this study. Literature Cited Baker, A.L., and K.K. Baker. 1979. Effects of temperature and current discharge on the concentration and photosynthetic activity of phytoplankton in the upper Mississippi River. Freshwater Biology 9:191–198. Baker, K.K., and A.L. Baker. 1981. Seasonal succession of the phytoplankton in the upper Mississippi River. Hydrobiologia 83:295–301. Chandler, D.C. 1937. Fate of typical lake plankton in streams. Ecological Monographs 7(4):445–479. Clarke, K.R. 1993. Non-parametric multivariate analyses of changes in community structure. Australian Journal of Ecology 18:117–143. Clescerl, L.S., A.E. Greenberg, and A.D. Eaton. 1999. Standard Methods for the Examination of Water and Wastewater, 20th Edition. American Public Health Association, Washington, DC. 1325 pp. Coker, R.E. 1923. Methuselah of the Mississippi. Scientific Monthly 16:89–103. Engman, J.A. 1984. Phytoplankton distribution in pool 19, Mississippi River. M.Sc. Thesis. Western Illinois University, Macomb, IL. 184 pp. 402 Northeastern Naturalist Vol. 15, No. 3 Fremling, C.R. 2005. Immortal River: The Upper Mississippi River in Ancient and Modern Times. The University of Wisconsin Press, Madison, WI. 429 pp. Gale, W.F., and R.L. Lowe. 1971. Phytoplankton ingestion by the fingernail clam, Sphaerium transversum (Say) in pool 19, Mississippi River. Ecology 52:507–513. Grubaugh, J.W., R.V. Anderson, D.M. Day, K.S. Lubinski, and R.E. Sparks. 1986. Production and fate of organic material from Sagittaria latifolia and Nelumbo lutea on pool 19, Mississippi River. Journal of Freshwater Ecology 3:477–484. Huff, D.R. 1986. Phytoplankton communities in navigation pool no. 7 of the upper Mississippi River. Hydrobiologia 136:47–56. Jahn, L.A., and R.V. Anderson. 1986. The ecology of pools 19 and 20, upper Mississippi River: A community profile. US Fish and Wildlife Service, [PROVIDE LOCATION]. Biological Reports 85(7.6). Kruskal, J.B. 1964. Nonmetric multidimensional scaling: A numerical approach. Psychometrica 29:1–27. Lamberti, G.A., and J.W. Moore. 1984. Aquatic insects as primary consumers. Pp. 164–195, In V.H. Resh and D.M. Rosenberg (Eds.). The Ecology of Aquatic Insects. Praeger Publishers, New York, NY. 625 pp. Lange, T.R., and R.G. Rada. 1993. Community dynamics of phytoplankton in a typical navigation pool in the upper Mississippi River. Journal of the Iowa Academy of Sciences 100:21–27. Lazzaro, X., R.W. Drenner, R.A. Stein, and J.D. Smith. 1992. Planktivores and plankton dynamics: Effects of fish biomass and planktivore type. Canadian Journal of Fisheries and Aquatic Sciences 49:1466–1473. Luttenton, M.R., and R.G. Rada. 1986. Effects of disturbance on epiphytic community architecture. Journal of Phycology 22:320–326. Luttenton, M.L., J.B. Vansteenburg, and R.G. Rada. 1986. Phycoperiphyton in selected reaches of the upper Mississippi River: Community composition, architecture, and productivity. Hydrobiologia 136:31–46. Madenjian, C.P. 1995. Removal of algae by the Zebra Mussel (Dreissena polymorpha) population in western Lake Erie: A bioenergetics approach. Canadian Journal of Fisheries and Aquatic Sciences 52:381–390. McCune, B., and J.B. Grace. 2002. Analysis of Ecological Communities. MjM Software Design. Glendeden Beach, OR. 300 pp. McCune, B., and M.J. Metford. 1999. PC-ORD for Windows, Version 4.25. MjM Software Design. Gleneden Beach, OR. Raidow, D.F., and S.K. Hamilton. 2001. Bivalve diets in a Midwestern US stream: A stable isotope enrichment study. Limnology and Oceanography 46:414–522. Reinhard, E.G. 1931. The plankton ecology of the upper Mississippi, Minneapolis to Winona. Ecological Monographs 1:395–464. Scholze, R.J. 1994. A summary of best management practices for nonpoint source pollution. US Army Corps of Engineers, USACERL Technical Report EP-93/06. National Technical Information Service, Springfield, VA. US Geological Survey (USGS). 1999. Ecological status and trends of the upper Mississippi River System 1998: A report of the Long Term Resource Monitoring Program. LTRMP 99-T001. 236 pp. Wehr, J.D., and J.H. Thorp. 1997. Effects of navigation dams, tributaries, and littoral zones on phytoplankton communities in the Ohio River. Canadian Journal of Fisheries and Aquatic Sciences 54:378–395. Welker, M., and N. Walz. 1998. Can mussels control the plankton in rivers? A planktological approach applying a Langrangian sampling strategy. Limnology and Oceanography 43:753–762.