establishing relationships among in-stream nutrient ... · macroinvertebrate indices and...

Establishing Relationships Among In-stream Nutrient Concentrations, Phytoplankton and Periphyton Abundance and Composition, Fish and

Macroinvertebrate Indices, and Biochemical Oxygen Demand in Minnesota USA Rivers

July 2003

Establishing Relationships Among In-stream Nutrient Concentrations, Phytoplankton Abundance and Composition, Fish

IBI and Biochemical Oxygen Demand in Minnesota USA Rivers Final Report to USEPA Region V

Minnesota Pollution Control Agency Environmental Outcomes Division

Steven Heiskary and Howard Markus

Contributors Report: Fish IBI analysis – Scott Niemela, Mike Feist and Leah Class Macroinvertebrate data – Joel Chirhart Transparency tube analysis – Laurie Sovell Red River data analysis – Andrea Plevan, Student Intern Diurnal water quality measurement and periphyton collection – USGS, Moundsview office Manuscript Review –

Doug Hall, Environmental Outcomes, MPCA Dan Helwig, Environmental Outcomes, MPCA Kathy Lee, USGS, Mounds View

Bruce Monson, Environmental Outcomes, MPCA Scott Niemela, Environmental Outcomes, MPCA Bruce Wilson, Regional Environmental Management, MPCA

Data Management – Nancy Flandrick, Jean Garvin, and Linda Nelson Sampling – Numerous persons assisted with sampling of the rivers over the course of the three-year study including several MPCA staff and student interns and some local cooperators:

Student Interns: Andrea Plevan, Dan Barringer, Ann Wicklund, Phillip McDonald, Chris Klukas;

Staff: Beth Endersbe, Sandy Bissonette, Mike Vavricka, Laurie Sovell, Willis Munson, Louise Hotka;

Local Cooperators – Kelli Daberkow, Cottonwood County; diurnal monitoring and periphyton collection – USGS, Mounds View office

Typing: Jan Eckart

i

Table of Contents Page I. Introduction.............................................................................................................................1 A. Overview........................................................................................................................1 B. Study Area and Ecoregion Descriptions ........................................................................2 C. Material and Methods ....................................................................................................7 II. Results .................................................................................................................................12

A. USGS studies ...............................................................................................................13 Periphyton....................................................................................................................13 Diurnal Measurements .................................................................................................15 B. Water quality and aquatic community patterns and relationships within streams: 1999-2000...........................................................................................18 Crow Wing River......................................................................................................18 Mississippi River ......................................................................................................27 Rum River.................................................................................................................29 Crow River................................................................................................................33 Blue Earth River .......................................................................................................35 Red River of the North..............................................................................................38

III. Discussion.............................................................................................................................41 A. Water quality patterns and relationships among streams.............................................41 Relationships among nutrients, chlorophyll-a and BOD5............................................41 Algae community composition....................................................................................51 Relationships among total suspended solids, turbidity and transparency....................56 Relationships among nutrient enrichment, dissolved oxygen and fish IBI .................57 Fish IBI and macroinvertebrate EPT indices ...............................................................63 B. Approaches for nutrient criteria development .............................................................70 Low nutrient concentration - protection oriented approach: Crow Wing River..........70 Moderate nutrient concentration - slight reduction approach: Rum River ..................71 Nutrient-rich river - BOD reduction approach: Crow River........................................72 Nutrient-rich river, high algal response: Blue Earth River ..........................................74 Large river draining multiple ecoregions, mass balance approach: Mississippi River76 High nutrient and low algal response, focus on a downstream receiving water approach: Red River ................................................................................................80 IV. Summary and Recommendations .........................................................................................81 References...................................................................................................................................86 Appendices..................................................................................................................................91

ii

List of Figures

Page 1. River-nutrient study sample sites: statewide and basin-specific maps ................................5 2. Periphyton percent biovolume and biovolume by substrate type ......................................16 3. Periphyton chlorophyll-a by substrate type .......................................................................17 4. Dissolved oxygen flux at select sites .................................................................................17 5. Total phosphorus and flow for select sites.........................................................................22 6. Total chlorophyll-a (ChlT) and flow comparisons for Crow Wing (CWR-72), Rum (RUM-18), Crow (CR-23) and Blue Earth (BE-18) Rivers for 1999 and 2000. ......25 7. Phytoplankton composition for 1999 and 2000: (a) Crow Wing River,

(b) Rum River, (c) etc. .......................................................................................................30 8. Red River (a) TP and chlorophyll-a and (b) chlorophyll-a and flow comparisons

among sites ........................................................................................................................39 9. Summer-mean total phosphorus (TP) and total chlorophyll-a (ChlT) for (a) 1999 and 2000 sites and (b) including Red River sites...............................................................44 10. Summer-mean total Kjeldahl nitrogen (TKN) vs. total chlorophyll-a (ChlT) for Minnesota river-sites sampled in 1999 and 2000. .............................................................45 11. Summer-mean total chlorophyll-a (ChlT) versus watershed area for Minnesota river-sites sampled in 1999 and 2000. ...............................................................................46 12. Summer-mean total chlorophyll-a (ChlT) vs. BOD5 for Minnesota river-sites sampled in 1999 and 2000 .................................................................................................47 13. BOD5, chlorophyll-a, and pheophytin comparisons for Crow and Blue Earth River sites...........................................................................................................................47 14. Summer-mean total phosphorus vs. BOD5 for Minnesota river-sites sampled in 1999 and 2000....................................................................................................................49 15. Summer-mean total phosphorus (TP) vs. total chlorophyll-a (ChlT) for core (1999 & 2000) study sites compared to new 2001 sites. ...................................................50 16. Summer-mean total chlorophyll-a (ChlT) vs. BOD5 for core (1999 & 2000) sites as

compared to new 2001 sites...............................................................................................51 17. Mean algal composition based on abundance (units/mL) and biovolume for selected

Minnesota river-sites. Based on three summer samples per site from 1999 and 2000. Worldwide percentage drawn from Rojo et al. (1994). .....................................................52

18. Comparison of algal composition of periphyton rock and wood samples and phytoplankton samples for Summer 2000 .........................................................................55

19. Total suspended solids vs. turbidity...................................................................................56 20. Transparency tube measurements as compared to TSS, turbidity, TP and chlorophyll-a .....................................................................................................................58 21. Diurnal DO flux compared to TP and seston chlorophyll-a ..............................................60

iii

22. Community DO production and respiration as compared to TP and seston chlorophyll-a .....................................................................................................................61 23. Community DO production and respiration as compared to periphyton chlorophyll-a ...62 24. Fish IBI macroinvertebrate EPT compared to QHEI.........................................................65 25. Fish IBI and macroinvertebrate EPT compared to TP, TN chlorophyll-a,

BOD5 and diurnal DO flux ................................................................................................66 26. Mississippi River mean TP and flux for 1999 ...................................................................79

Tables Page 1. Watershed area and channel morphometry for study sites ................................................10

2. Laboratory methods and precision estimates for river-nutrient study ...............................11

3. Summer-means and maxima for 1999 and 2000 ...............................................................20

4. Regression equations derived based on river nutrient study..............................................41

5. Fish IBI and macroinvertebrate EPT values for 2000 .......................................................64

6. Crow Wing River data summary and comparison.............................................................71

7. Rum River data summary and comparison........................................................................72

8. Crow River data summary and comparison.......................................................................74

9. Blue Earth data summary and comparison .......................................................................75

10. Mississippi River ecoregion composition..........................................................................77

11. Mississippi River data summary and comparison .............................................................78

12. Mass balance comparison for Mississippi and Crow Rivers .............................................79

13. Red River data summary and comparison .........................................................................81

14. Summary of approaches for setting nutrient criteria..........................................................84

iv

Establishing Relationships Among In-stream Nutrient Concentrations, Phytoplankton and Periphyton Abundance and Composition, Fish and

Macroinvertebrate Indices and Biochemical Oxygen Demand in Minnesota USA Rivers.

Final Report to USEPA Region 5

Abstract Significant and predictable relationships were demonstrated among summer nutrient, chlorophyll a, and algal concentrations and biochemical oxygen demand (BOD) in five medium to large Minnesota, USA, rivers. Summer (June – September) flows in 1999 were significantly higher than 2000, and as a result, the “age” of the water (residence time) was greater in 2000. The lower and more stable flows generally resulted in higher algal concentrations in 2000 than in 1999. Algal composition varied not only in terms of origin: benthic vs. planktonic, but also along a gradient of nutrient enrichment. Benthic diatoms comprised a significant proportion of the algal community in clear low nutrient rivers but declined in significance in more nutrient rich rivers where planktonic green and blue-green algae became more prominent. In more turbid and high nutrient rivers highly tolerant blue-greens were dominant. Subsequent studies provide further insights into the linkages and relations among total phosphorus (TP) concentrations, chlorophyll-a and BOD. Water quality data from independent sites indicated that the previously-defined relationships were valid for most sites; however, these relationships did break down in the highly turbid Red River of the North because of extremely high inorganic turbidity that limited the growth of algae and subsequent nutrient assimilation. Diurnal dissolved oxygen (DO) flux (based on submersible data recorders over a period of three to six days) was found to be strongly positively correlated to summer mean TP and chlorophyll-a concentrations at 12 stream sites tested in 2000. Likewise, fish index of biotic integrity (IBI) scores were found to be inversely correlated with summer-mean TP. Linkages established here will contribute to nutrient criteria development and nutrient or DO-based TMDLs. Examples of approaches for setting river nutrient goals are offered.

1

I. INTRODUCTION A. Overview The purpose of this study was to document relationships among nutrient concentrations, phytoplankton concentration and composition, and biochemical oxygen demand in large Minnesota rivers. This understanding will support national efforts for ecoregion-based nutrient criteria development, enhance predictive modeling and load allocation studies for medium to large rivers, and provide an improved basis for setting nutrient goals for rivers. In this study we demonstrated significant and predictable relationships among nutrients, chlorophyll, and BOD in a range of Minnesota rivers. We also demonstrated the significant role that flow, suspended solids, watershed size, and residence time play in this relationship. We also found that algal community composition varied as a function of nutrient concentration, watershed size and flow. Linkages were also made among fish IBI, nutrients, chlorophyll-a, and diurnal dissolved oxygen flux. The role of excess nutrients (phosphorus and nitrogen) in the eutrophication of lakes and reservoirs has been long known and well documented in the literature (USEPA 2000a). This knowledge of cause (excess nutrients) and effect (e.g., nuisance algal blooms, reduced transparency, and low hypolimnetic dissolved oxygen) provides ample basis for the development of nutrient criteria for lakes and reservoirs (Heiskary and Wilson 1989) and for the development of nutrient-based wasteload allocations. The role of excess nutrients in the eutrophication of rivers is not, however, as well documented nor are there clear linkages between nutrients and response variables (e.g., algal abundance, turbidity, etc.). Work that has been done on North American rivers has focused primarily on periphyton growth in shallow (wadable) rivers (Dodds et al. 1997 and Dodds and Welch 2000). Much less work has been done on river phytoplankton in larger non-wadable rivers (USEPA 2000b). Rivers have a cycle of planktonic biomass fluctuation that is as regular and reproducible as any annual cycle from lakes (Reynolds, 2000). Various studies have sought to describe factors that might control the production of phytoplankton in rivers. Baker and Kromer-Baker (1979) note the role of temperature and stream discharge on the production of phytoplankton in the Mississippi River, south of the Minneapolis-St. Paul metropolitan area. They note further that algal concentrations have increased up to 40-fold the concentrations since the installation of locks and dams in the 1920’s (Kromer-Baker and Baker 1981). Soballe and Kimmel (1987) made comparisons among the responses of lakes, reservoirs, and rivers – noting a gradient in response to nutrients related to retention time and other factors. More recent studies from North America (Van Nieuwenhuyse and Jones 1996; Basu and Pick 1995, 1996) and Europe (e.g., Billen et al. 1994) document linkages between phosphorus and in-stream chlorophyll-a. Other studies describe phytoplankton dynamics in rivers and contributing environmental factors (Rojo et al. 1994, and DeRuijter van Steveninck et al. 1992). While these studies contribute significantly to our knowledge of phytoplankton in streams and environmental factors that might control abundance and growth they do not often provide further linkages with other variables such as biochemical oxygen demand (BOD) or other factors which could contribute to the identification of quantifiable thresholds of impairment in rivers.

2

Documenting these relationships will be useful for nutrient criteria development, load allocations, and establishing effluent limitations as a part of National Pollutant Discharge Elimination System (NPDES) permits. USEPA (1997), in guidance for developing total maximum daily loads (TMDLs), alludes to the role of algae in the consumption (respiration) and production (photosynthesis) of oxygen, however a link back to carbonaceous BOD appears to be missing. This linkage is later implied in the USEPA (1999a) protocol for developing nutrient based TMDLs but is not explicitly described. The Minnesota Pollution Control Agency (MPCA) has used the interrelationship of nutrients, chlorophyll-a and BOD5 in recent NPDES permits. The USEPA and MPCA established a TMDL for BOD at the Lower Minnesota River below River Mile 25 in 1988 (Anderson and Klang 1997). Subsequent work by MPCA staff, including Erwin Van Nieuwenhuyse (formerly at the MPCA), recognized linkages among phosphorus, algal production and BOD for this portion of the Minnesota River (MPCA unpublished data). This recognition led to point-nonpoint source trading for a new point source discharge in this reach. This trade called for the new discharger to fund upstream nonpoint source projects in order to reduce nutrient loading (algal production) in the Lower Minnesota River (Anderson and Klang 1997). This in turn would offset additional BOD they would contribute in this reach and provide for overall reductions in BOD as a part of the TMDL for this reach. This TMDL is also influencing other NPDES discharges in the Minnesota River Basin with respect to phosphorus control. The goals of this study include: a) establishing relationship between nutrients, and sestonic chlorophyll-a; b) establish relationships between chlorophyll-a and BOD5 and related factors; c) explore relationships among phytoplankton and periphyton composition and abundance and nutrients; d) begin to study interrelationship of above factors with fish and invertebrate communities; and provide some examples of how this might be used to develop criteria. Portions of this report were previously published in the Journal of Lake and Reservoir Management (Heiskary and Markus, 2001). This report considers those findings and additional work that was conducted in 2000 by the United States Geological Survey (USGS) (Lee, 2002) and MPCA’s bioassessment team on selected sites from the original study. In addition, several different rivers were sampled in 2001 (Appendix III) and that data is incorporated as well.

B. Study Area and Ecoregion Descriptions This study focused on medium to large rivers that are typical of several Minnesota ecoregions. Between-region differences in land use, soil characteristics and geomorphology influence water runoff, nutrient loading, and processing of nutrients in the rivers (USEPA 2000b). MPCA has previously described ecoregion-based differences in stream water quality based on minimally-impacted streams (McCollor and Heiskary, 1993). A summary of values from that study that will be used as one basis for comparison to data from the current study (Appendix I). Likewise, USEPA has recently compiled distributions of water quality variables by ecoregion as a part of their “Ambient Water Quality Criteria Recommendations” that were compiled for the various nutrient ecoregions (e.g., USEPA, 2000b). The rivers selected for our study (Fig. 1) drain one or more of the following Level IV ecoregions: Northern Lakes and Forests (NLF), North Central Hardwoods Forests (NCHF), Western Corn

3

Belt Plains (WCBP), Northern Glaciated Plains (NGP) or Red River Valley (RRV). These ecoregions correspond to three aggregated Level III (USEPA, 2000b) Nutrient ecoregions (VIII, VII, and VI, respectively) that characterize Minnesota and much of the Upper Midwest. Watershed areas are 2,590 km2 (1,000 mi2) or greater for 19 of 21 sites (Table 1). 1,000 mi2 has been used as a basis to differentiate large streams from small streams (Miltner and Rankin 1998). Selecting rivers-sites from each ecoregion and of varying watershed areas allowed us to capture a range of responses. At least two sites were sampled on each river to allow upstream and downstream comparisons of concentration and flux in the sampled parameters. Site designators are represented by an abbreviation for that river and miles from the river mouth, for example CWR-72 (Crow Wing River, 72 miles upstream from the mouth). Sites are located at, or near, USGS stream gage locations whenever possible to allow for accurate estimates of flow for each sample date, with a minimum of one USGS gage per river. Watershed Description As is evident in Fig. 1, a single river community drains more than one ecoregion. Therefore, the watershed for each river-site was characterized in terms of the percent composition by ecoregion in the following description. For some rivers, such as the Blue Earth, only one ecoregion is drained (WCBP ecoregion). For others, such as the Mississippi River, there is a transition from a watershed primarily characterized by forested, lake, and wetland dominated land use (NLF; Heiskary and Wilson, 1989) in the upper reaches to the increasingly agricultural and urbanized land uses that characterize the NCHF ecoregion. These differences in land use, soil characteristics and geomorphology influence water runoff and pollutant loading to the river. For example at UM-895 the contribution from the NLF portion is about equal to that from the NCHF portion of the ecoregion based on areal estimates. The upper site on the Crow Wing at CWR-72 drains about 2,668 km2 (1,030 square miles) on the western edge of the NLF ecoregion. The downstream site at CWR-35 has a watershed of about 5,517 km2 (2,130 square miles) of which about 58 percent is from the NLF and 42 percent is from the NCHF ecoregion (Fig. 1). The Crow Wing River watershed is relatively undeveloped and heavily forested. Fifty-one and thirty-nine percent of the drainage area of the upstream and downstream sites on the Crow Wing River were classified as forest land (Appendix II, Table 1). Disturbed land occupied less than half of the drainage area. The Mississippi River has its origin in the NLF ecoregion (Fig. 1). In 1999, the furthest upstream site monitored in this study was at UM-1004 in Brainerd, Minnesota. As the Mississippi River flows from the Brainerd area, the watershed becomes more characteristic of the NCHF ecoregion (Fig. 1). At the most downstream site, UM-872 (just north of the Twin Cities Metropolitan Area), the majority of the watershed (65 percent by area) lies in the NCHF ecoregion. In 2000, we replaced sites UM-1004 and UM-965 with two sites upstream of Brainerd (UM-1056 and UM-1029) that are more characteristic of the NLF ecoregion and exhibit less direct influence from upstream reservoirs. From upstream to downstream locations, the percent of agricultural land, range land, and urban areas increased steadily, while the percent of forest, wetland, and water progressively declined (Appendix II, Table 1), which is consistent with the transition from the NLF to NCHF portion of the watershed.

4

The Rum River, with the exception of Lake Mille Lacs at its headwater (accounting for about 15 percent of the watershed at RU-18), primarily drains the NCHF ecoregion. Two downstream sites at Isanti, Minnesota (RU-34) and St. Francis, Minnesota (RU-18) were monitored in 1999 and 2000. No major tributaries enter between the two sites and about 250 km2 additional watershed area is drained over this 26-km (16-mile) reach. There are no lakes or reservoirs on the main stem of this reach; however several small lakes are located in the watershed between these two sites. The Rum River empties into the Mississippi River in Anoka, Minnesota. The drainage area of these sites contains a moderate amount of disturbed land (42 percent). One-third of the watershed is forested (Appendix II-Table 1). The Crow River has two very distinct subwatersheds. The North Fork comprises about 53 percent of the watershed, drains from the NCHF ecoregion with numerous lakes and wetlands throughout the watershed (Fig. 1). The South Fork, comprising about 47 percent of the watershed, drains primarily from the agricultural WCBP ecoregion. The highly agricultural land use, combined with numerous small wastewater discharges dotted throughout both subwatersheds, contributes to the “nutrient-rich” conditions in the Crow. The two sites monitored in 1999 and 2000 (CR-23 and CR-0) are located on the mainstem downstream from the confluence of the North and South Forks at river mile 23 and near the mouth to the Mississippi river. No major tributaries enter between these two sites, however about 614 km2 of watershed are added over this 36-km (23-mile) reach. The Crow River watershed has a high percentage of agricultural land (68 and 69 percent). Eighty-four to eighty-five percent of the drainage area of sites on the Crow River is classified as disturbed (Appendix II-Table 1). The Blue Earth River, in south central Minnesota, has its headwaters in Iowa and drains an area of about 9,174 km2 (3,542 square miles) before entering the Minnesota River. Five sites arrayed from near the Iowa border (BE-94) to a site upstream of the Rapidan Dam (BE-18) were included in 1999. In 2000, the site at BE-77 was replaced with BE-100 to allow for an improved characterization of headwater conditions. Approximately ninety percent of the land use within the basin is engaged in agriculture (Appendix II-Table 1). Ninety-five to ninety-six percent of the drainage area for sites on the Blue Earth River is classified as disturbed. The Red River of the North was added to the study in 2000. The Red River flows north from Browns Valley in Traverse County to the Canadian border. Four sites were included in this study (Table 1). The Minnesota portion of the watershed covers just under 43,512 km2. The Red River receives most of its flow from eastern (Minnesota) tributaries as a result of regional patterns in precipitation, evapotranspiration, soils, and topography (Stoner, 1991). The total watershed at the most downstream monitoring site at Grand Forks ND (RE-298) is about 56,980 km2. The glacial lake plain that the watershed drains is relatively flat, with soils containing mostly clay particles.

5

Figure 1. River nutrient study sample sites: statewide and Mississippi, Minnesota and Red River basin maps.

Upper Mississippi Basin sites

(X

(X(X

(X(X

(X

(X(X(X(X(X

(X

(X

(X

(X

(X

(X(X

(X

(X(X

(X(X

CROW R

MISSISSIPPI R

CROW

WING

R

BLUE

EAR

TH R

RUM R

RED

R

N

October 20000 50 100 150 200 Miles

EcoregionRed River ValleyNorthern Minnesota WetlandsNorthern Lakes and ForestsNorth Central Hardwood ForestsDriftless AreaWestern Corn Belt PlainsNorthern Glaciated Plains

Major Watersheds/HUCBasinRiver

(X Sampling Site

Figure 1.MPCA River Nutrient Study Sites

Major Watershed and Ecoregion Boundaries Noted

#

#

#

#

#

#

#

#

#

#

#

#

CROW W

ING R

MISSISSIPPI R

RUM R

MISSISSIPPI R

CROW R

MISSISSIPPI R

Aitkin

Brainerd

Royalton

St. Francis

Rockford

Anoka

Nimrod

CWR-72.3

UM-1056UM-1029

UM-1004

CWR-35.5

UM-953

UM-895

CR-23

RUM-18

RUM-34

CR-0.2 UM-872

April 2003

N

0 20 40 60 Miles

Upper Miss Basin

Major WatershedMississippi R (Sartell)Mississippi RMississippi R (St. Cloud)North Fork Crow RRum RMississippi R (Brainerd)Crow Wing R

RiverCity

# Sampling Site

Basin

6

Minnesota River Basin – Blue Earth River sites

#

#

#

#

# BE-18.2

BE-73.2

BE-94.3

BE-100

BE-54

Winnebago

Blue Earth

Rapidan Twshp

Amboy

BLUE EARTH R

N

April 2003

0 20 40 60 Miles

Minnesota R Basin

Major WatershedWatonwan RLe Sueur RBlue Earth RMinnesota R (Mankato)

CityRiver

# Sampling Site

Basin

#

#

#

#

RED

R

RE-403

RE-298

RE-452

RE-536

East Grand Forks

Perley

Moorhead

N

April 2003

0 20 40 60 Miles

Red R Basin

Major WatershedMarsh RRed R of the NorthRed Lake RWild Rice R

CityRiver

# Sampling Site

Basin

Red River Basin sites

7

t = 0.08 * W^0.6 / Q^0.1

0

20000

40000

60000

Wat

ersh

ed S

ize

(km

2)

0 100 200 300Flow (m3/s)

1020

30

40

C. Materials and Methods

Morphometry

The morphometry of each river reach (study site) was determined based on available data and estimated values from published equations (Table 1). The length of each reach and watershed area above the reach was based on existing MPCA and the USGS data. The width of the river (wetted cross-section) was estimated over various river stages in 1999 and 2000, at each sample site, and an average value was included to provide a general description of the site. Mean depth at the sample site was estimated based on Soballe and Kimmel (1987) [after Leopold et al. 1964] where:

mean depth (m) = annual mean flow (m3/s) 0.42/2.95 (1)

To evaluate the effect of stream flow, we used USGS discharge data for the summers of 1999 and 2000. Residence time or “age of water in the system”, was estimated based upon equations provided by Soballe and Kimmel (1987) [after Leopold et al. 1964] where:

residence time (days) = 0.08* watershed area (km2) 0.6 / mean flow (m3/s) 0.1 (2)

The relationship between these variables is depicted below. For sites without actual flow data, residence time was estimated based on the watershed area of the site as a proportion of the watershed area for the nearest gauged site. Based on a comparison of upstream-downstream watershed area and mean flow (upstream/downstream) the percentage differences ranged from one percent to ten percent (e.g., the watershed at UM-1056 accounts for 34 percent of UM-872 watershed and 36 percent of mean flow).

Detailed channel morphometry for the Crow and Mississippi Rivers (MPCA unpublished data) allowed for calculation of travel time as a function of flow and provided a basis for evaluating “age of water” estimates. For the Crow, travel time over the reach from CR-23 to CR-0 varied from about 20 hours, at 42 m3/s (1,500 cfs) up to about 50 hours, at 5.7 m3/s (200 cfs). Considering average flows of about 25.5 m3/s (900 cfs) and 8.5 m3/s (300 cfs), respectively for 1999 and 2000 (Table 1), this would translate to travel (residence) time over this reach of about 25

to 35 hours. “Age of water” (residence time) estimates (based on extrapolated flow at CR-0) suggest that the “age of water” at CR-0 would be on the order of 12.3 days (24 hour increase compared to CR-23) at 25.5 m3/s (1999) and 13.7 days (26 hour increase compared to CR-23) at 8.5 m3/s (2000). Based on this comparison the “age of water” estimates provide a reasonable basis for estimating between-site changes in residence time for the Crow River. A similar time-of-travel comparison for the Mississippi River sites suggested that “age of water” estimates

days

8

might overestimate residence time for this much larger system. Time-of-travel measurements from UM-1004 to UM-872 averaged about 110 hours (4.6 days) during summer 1999. In comparison, “age of water” estimates indicated a time-of-travel on the order of 10 days. In summer 2000 time-of-travel from UM-1004 to UM-872 averaged 8.3 days, which was two-fold greater than summer 1999. For the Mississippi sites, time-of-travel estimates were more accurate and will be referred to later in the discussion. Stream order (after Strahler, 1957) was estimated based on 1:190,000 scale quadrangle maps in the Minnesota Atlas and Gazetteer (DeLorme 1994). BE-100, with the smallest watershed in this study (811 km2), was deemed a third-order stream. The remaining sites on the Blue Earth, Crow Wing at CWR-72, and Rum were deemed fourth-order. The Crow Wing at CWR-35, Crow River, and Mississippi River at UM-1004 to UM-1056 were deemed fifth-order. The remaining sites on the Mississippi (downstream of UM-1004) were deemed sixth order. Red River sites were presumed to be fifth order or higher based on watershed size (Table 1). To provide specific land use information on the watershed of each site, upstream land use was characterized using GIS land use coverages. The GIS land use theme was overlaid in Arcview onto the drainage area theme and clipped, producing a land use theme identical in shape and size to the drainage area of each site. Land uses were then summed across the entire drainage area and divided by the total area to produce percentages for each land use. The percent watershed disturbance was calculated by adding the percentages for the land use themes that were either agricultural, urban, grassland that was associated with pastured areas, mines, and open pits (Appendix II-Table 1).

Sample Collection Water samples were collected at mid-channel on five to six occasions from mid-July to mid –September in 1999 and seven to eight occasions from June through September in 2000. Samples were collected from bridges at each site by means of a bucket on a rope. The bucket was rinsed twice with ambient water prior to sample collection. For quality assurance purposes duplicate samples were collected on about ten percent of the visits. Though collection method was consistent among sites and over the study

period we cannot assume that the mid-channel collections are well-mixed and water quality (chemistry) may vary across the stream cross-section (Lee, personal communication). Nutrient samples were acidified upon collection with H2SO4. Chlorophyll-a samples were field filtered on the day of collection and the volume filtered was noted. The filter was then placed in a petri-dish and wrapped in foil. Samples were frozen prior to shipment to laboratory and analyzed for chlorophyll-a and pheophytin-a. Other samples such as total suspended solids and BOD5 were not preserved, but chilled to 4º C prior to shipment to the laboratory. Phytoplankton samples were subset from the grab samples and preserved with Lugol’s solution (APHA 1998). In 1999, three samples, typically one each from July, August and September, from nine sites (CWR-35, UM-1004, UM-953, UM-872, RU-34, RU-18, CR-23, BE-94 and BE-54) were forwarded to PhycoTech, Inc. for identification to genera and bio-volume calculation.

9

A similar approach was followed in 2000 as well. The methods may be found under “General Technical Approach” on the PhycoTech Web site (www.phycotech.com). Three slides per sample were prepared by the HPMA (2-hydroxypropyl methacrylate) method for analysis. An Olympus BHT compound microscope was used to identify the algae. Approximately 10 percent of the counts were recounted from the same slides for both abundance and biovolume to determine detailed counting efficiency. All water chemistry samples were analyzed at Minnesota Department of Health (MDH). Precision estimates were derived from the analysis of ten duplicate samples taken during our study (Table 2). The mean and percentage difference, for these duplicates samples, was equivalent to or better than routinely reported results of MDH laboratory duplicates. Dissolved oxygen, pH, temperature, and conductivity were measured in the field with a multi-parameter probe during collection of water samples. Transparency tube measurements were made with a 60 cm long, 3.8 cm diameter, clear plastic tube. A well-mixed sample is poured into the tube. While looking down into the tube, water is released from a valve at the bottom until the black and white (Secchi) symbol at the bottom of the tube is visible. The depth of the water when the symbol becomes visible is recorded. Typically, two separate readings are averaged to yield the recorded measurement. Field and laboratory data will be stored in STORET (STOrage and RETrieval), USEPA’s national water quality data bank according to MPCA protocol. A variety of techniques were used to examine the data. Data analysis was conducted primarily by EXCEL spreadsheet (V97, Microsoft, 1997). Linear regression was used to describe relationships between variables and F-tests were used to assess the significance of these relationships. T-tests and/or comparisons of means and standard error were used to determine significant between-site and between-year differences. USGS (Lee 2002) collected periphyton samples at twelve of the study sites in 2000: CWR-72.3 and 35.5, UM-1056 and 872, RU-34 and 18, CR-23 and 0.2, BE-73.2 and 54, and RE-536 and 452. Benthic algal samples were collected from rock and wood substrates at each site. Approximately 10 different rocks and 10 branches (wood) were sampled at each site according to the methodology described in Lee (2002). Rock and wood samples were processed separately for each site. Processed samples were forwarded to PhycoTech for identification and biovolume estimation. Separate samples were filtered through 0.47 mm glass fiber filters, placed in foil, and frozen prior to delivery to Minnesota Department of Health for chlorophyll-a analysis. Also for most of the sites, with the exception of the Red River, phytoplankton collections were made on three or more occasions that bracketed the periphyton collection (typically in June, August and September). These collections allowed for qualitative and somewhat quantitative comparisons between periphyton and phytoplankton composition at a given site, comparisons between sites on the same river, comparisons across a gradient of nutrient concentrations and for phytoplankton comparisons between-years for those sites sampled in 1999 and 2000.

10

Table 1. Watershed area, channel morphometry, and flow characteristics for study sites. Summer (June – September) flow and residence time unless otherwise noted. Annual and summer mean are for period of record. Site MPCA (USGS)a

Watershed Area

Mean Wetted Cross-section

Mean Depth

Flow Annualmean

Flow Summer

mean

Flow Summer

1999

Flow Summer

2000

Res. Timeb Summ. mean

Res.Time1999 mean

Res.Time2000 mean

units: km2 m m m3/s m3/s m3/s m3/s days days days Crow Wing CWR-72 2,668 40 1.0 13.7 12.7 22.4 14.2 7.1 6.7 7.0CWR-35 5,517 65 1.0 Mississippi UM-1056 (1055.9)

15,242 35 2.2 84.1 81.0 152.9 58.2 16.7 15.6 17.2

UM-1029 15,768 35

UM-1004 (1003.7)

18,959 75 2.3 104.1 98.0 183.1 77.1 18.6 17.5 19.1

UM-953 (956)

30,044 130 2.6 135.6 132.6 252.6 122.8 23.9 22.4 24.9

UM-895 34,076 205 25.0 24.0 26.0UM-872 (864.8)

44,289 300 3.3 233.1 228.3 338.6 152.9 28.5 27.4 29.7

Rum RU-34 3,294 55 7.6 7.6 8.2RU-18 (15.8)

3,546 60 1.1 17.8 16.6 16.9 7.8 8.1 8.1 8.7

Crow 12.4 12.3 13.7CR-23 (23)

6,527 70 1.3 23.1 23.8 25.1 8.5 11.3 11.3 12.6

CR-0 7,141 75 Blue Earth BE-100 811 20 0.6 3.7 3.5 3.7BE-94 2,082 30 0.7 5.9 5.6 5.9BE-73 3,541 40 0.6 7.7 7.2 7.7BE-54 3,603 46 7.7 7.3 7.8BE-18 3,955 48 1.5 30.3 33.5 68.0 30.4 8.1 7.6 8.2Redc RE-536 (549)

10,490 -- 1.25 19.2 19.0 30.0 22.3 14.7 15.2

RE-452 (453)

17,612 -- 1.49 26.5 19.5 42.5 33.9 19.4 19.8

RE-403 (375)

56,462 -- 2.76 60.1 56.0 110.4 148.0 35.5 34.4

RE-298 (298)

56,980 -- 3.10 114.6 78.7 238.9 195.0 33.0 33.7

a River mile for USGS gauge noted b Represents “age of water” estimates based on flow and watershed area. For site without flow data, residence

time is estimated, because water quality samples are collected upstream of Red Lake River. c Red River watershed areas estimated based on nearest USGS gage sites. RE-298 flow and watershed areas have

Red Lake River subtracted from USGS values, based on the watershed area of the site as a proportion of the watershed area for the nearest gauged site.

11

Table 2. Laboratory methods and precision estimates for Minnesota river-nutrient study.

Parameter Reporting Limit & Units

EPA method number

Precision: 1 mean difference

Difference as Percent

of observed Total Phosphorus 10.0 µg.L-1 365.2 4.8 µg.L-1 2.7 % Total Kjeldahl N 0.1 mg.L-1 351.2 0.05 mg.L-1 2.8 % NO2 + NO3 0.01 mg.L-1 353.1 Total Suspended Solids

0.5 mg.L-1 160.2 2.8 mg.L-1 9.6 %

Total Suspended Volatile Solids

0.5 mg.L-1 160.4 -- --

Turbidity 0.2 NTU 180.1 -- -- BOD5 0.5 mg.L-1 405.1 0.15 mg.L-1 6.6 % Chlorophyll-a 0.16 µg.L-1 446.0 1.7 µg.L-1 7.4 % Pheophytin 0.27 µg.L-1 446.0 -- --

1 Average of individual means of 10 duplicates and expressed as a % of measured concentrations. Diurnal water quality measurements were made at the same 12 sites by USGS during the summer of 2000. Measurements of specific conductance, pH, water temperature, and DO were recorded at 30 minute intervals over a period of 3-6 days in August 2000 using submersible data recorders (Lee, 2002). The probes were positioned in the euphotic zone in an area of streamflow of at least 1 cfs (0.028m3/s). The probes were calibrated according to manufacturer specifications before installation and after retrieval. Stream productivity was estimated by USGS by calculating the slope of the DO concentrations between 10 a.m. and 3 p.m. (Lee, 2002). These estimates define the net rate of oxygen accrual in milligrams of oxygen per liter per hour. Net community respiration was quantified by calculating the slope of the DO concentrations between midnight and 6 a.m. As a further complement to the ongoing water chemistry, plankton and diurnal monitoring; fish, macroinvertebrate, and habitat studies were conducted by MPCA staff at 10 of 12 sites that USGS monitored on the Crow Wing, Mississippi, Rum, Crow and Blue Earth Rivers. These investigations allowed for the calculation of fish Index of Biotic Integrity (IBI), macroinvertebrate indices, and habitat descriptions for each site. A habitat assessment was performed at each site to characterize the instream and riparian features of the stream. In wadable streams, a modified version of Wisconsin’s quantitative habitat assessment procedure (Simonson et al. 1994) was used. The habitat assessment included characterization of streambed substrate (e.g., boulders, cobble, silt), instream cover (woody debris, macrophytes), and riparian land use at 13 evenly-spaced transects. Channel morphology (riffles, runs, pools) throughout the reach was noted. In non-wadable streams we used the Ohio Qualitative Habitat Evaluation Index (QHEI) (Rankin 1989). The QHEI rates the habitat based on substrate quality, in-stream cover, riparian zone quality and bank erosion, and pool/glide and riffle/run quality. QHEI data were compiled for all wadable streams in this study based on results from the quantitative habitat assessment to allow for greater ease in comparing specific habitat and stream features among sites (Appendix II-Table 2). Habitat and fish collections were made during daylight hours between mid-June and September. Fish were collected using electro-fishing techniques following procedures described in Niemela

12

and Feist (2002). Depending on stream size and type, a towed stream electrofisher, mini-boom electrofisher, or boom electrofisher was used to sample the fish community. An IBI score was calculated for each sampling event using an IBI developed specifically for streams in each respective basin. The IBI uses multiple attributes of the fish community (termed metrics) to characterize the biological integrity of the stream reach (Karr 1981). For the Blue Earth River, the Minnesota River Basin IBI developed by Bailey et al. (1993) was used with one exception; metric classifications for fish species were updated following Niemela and Feist (2000). For sites in the Upper Mississippi River Basin (e.g., Crow, Rum, and Crow Wing), IBI scores were calculated using the IBI developed by Niemela and Feist (2002). Common metrics and IBI scores for all sites are included in Table 3 in Appendix II. The differences in IBI metrics between the two basins are noted in parenthesis. Because of the differing range in scores between the two IBIs [The total score range for the Minnesota River Basin IBI is from 12 to 60; for the Upper Mississippi River Basin IBI, overall scores range from 0 to 100], narrative descriptions were used to rate the biotic integrity of each site and to allow for the interpretation of overall IBI scores and comparison between river basins (Appendix II-Table 4). Macroinvertebrate samples were collected at ten sites in September 2000: CWR-72, UM-953, RU-18, CR-23, CR-0.2, BE-100, BE-94, BE-73, BE-54, and BE-18. All sites were considered wadeable at that time, with the exception of UM-953, which was deeper and wider (higher order) than the other nine sites (Table 1). Hence this site may not have been sampled as completely as the other sites. Multihabitat samples were taken by means of standard protocols according to the methods in Barbour et al. (1999) and MACS (1996). As with the MACS protocols, soft bottom substrates were not sampled. A 600 micron mesh, d-frame dipnet was used to collected samples. Complete samples were subsampled to a minimum of 300 organisms followed by a 20 minute large and rare pick to supplement taxa richness. Identifications were made to the genus level or higher depending on the maturity and condition of the specimens. Chironomids were identified to family. An invertebrate IBI had not been developed for the state at the time of this analysis. Rather, an index based on the Ephemeroptera (mayflies), Plecoptera (stoneflies), and Trichoptera (caddis flies), EPT index, provided a relative measure of the presence and diversity of pollution-sensitive macroinvertebrate groups. These taxa (and this index) are felt to be strong water quality indicators (Barbour et al. 1992). This will be primary index used later in the report to compare to fish IBI and water chemistry data. II. RESULTS The research documented in this report was conducted between 1999 and 2001. With the exception of the Red River sites, most monitoring sites (Table 1) were sampled in 1999 and 2000. In addition to the water quality monitoring, related studies by the USGS staff and MPCA biomonitoring team were conducted in 2000 at 10-12 of the core sites on the Crow Wing, Mississippi, Crow, Rum, Blue Earth and Red Rivers. In 2001 several additional river sites were monitored for the purpose of validating nutrient and chlorophyll-a regression models (developed from the 1999 and 2000 data, Table 3) and to help better understand the relationships between

13

these variables. Results are presented as follows: A) USGS diurnal monitoring and benthic algae collections; B) MPCA biomonitoring team 2000 fish and habitat study; and C) river-specific descriptions of flow, water quality, and phytoplankton and periphyton composition based on the 1999 and 2000 monitoring.

A. USGS Studies

Periphyton

Periphyton (benthic algae) samples were collected from rock and wood substrates at each of twelve separate sites (see Methods) in August 2000. Diatoms, greens and blue-greens were the predominant algal types found at each site, with diatoms being the most abundant (Figure 2a & b). Based on paired data for 12 sites, periphyton was greater on rock compared to wood at 7 of 12 sites (Figure 2c & d) and in terms of density at 9 of 12 sites. The total number of diatom

species found at any one site ranged from 25 (wood) at UM-872 to 8 (wood) at BE-54, with 13-15 species being typical for most sites. Blue-greens were frequently the next most abundant form and ranged from nine (rock) species at UM-872 to two species at several sites, with four to five species typically found at most sites. Greens varied from 12 (wood) species at CR-0.2 to one at UM-1056 and CWR 72.3, with three to four species typically found at most sites. Among the diatoms the most common genera found included Navicula, Nitzschia, Rhoicosphenia, Cocconeis, Cyclotella, and Melosira. These genera were often represented by three or more species. Some like Navicula were represented by 16 species and others like Rhoicosphenia curvata were represented by one species. Among the highest densities of diatoms were Cocconeis placentula (CWR 35.5) – 202.1E6 µm 3/cm2, Cyclotella meneghiniana (BE-54) - 494.1E6 µm 3/cm2, Gyrosigma spencerii (RE-452) – 157.1E6 µm 3/cm2 and Melosira varians (CWR-35.5)- 237E6 µm 3/cm2. The most common blue-green algal types were non-motile blue-greens (found at all sites), followed by the filamentous Lyngbya and Oscillatoria. Among the highest densities of Oscillatoria 377.6E6 µm 3/cm2 (wood) and 122.7E6 µm 3/cm2 were noted at RE-536 and RE-452 respectively. The highest density of Lyngbya was found at CWR-35.5: 146.8E6 µm 3/cm2. Typical densities of blue-greens at most sites were on the order of 10E6-20E6 µm 3/cm2. Green algae were represented by very few genera. Of the eight noted, only Chlorococcum was found at eight sites with densities ranging from 44.45E6 µm 3/cm2 (rock @ CWR-72.3) to 2.254E6 µm 3/cm2 (wood @ BE-73.3). Though filamentous greens such as Cladophora and Spirogyra were found infrequently they did exhibit extremely high densities when present. Cladophora accounted for 53 and 99 percent of the biovolume at RU-34 (rock) and CWR-72.3 (wood), respectively, with densities of 517.1E6 and 58.02E9 µm 3/cm2. Spirogyra , found only at RU-34, accounted for 87 percent of the biovolume at the site with a density of 2.494E9 µm3/cm2. Other forms such as Pyrrhophyta, Euglenophyta, Chrysophyta and Cryptophyta were seldom present in these samples. Only Euglena , which comprised 15 percent of biovolume at BE-73.2

14

and Batrachospermum vagum, a filamentous red alga, which comprised 31 percent of biovolume at RE-536 were found in any appreciable quantities. Hynes (1970) lists controlling factors for the presence or absence of benthic algal species, including current, substrate, light, and scour. Hynes (1971) stated that strength of current is the paramount factor for determining species composition. Once a scour event has occurred, often the initial algal “colonizers” can significantly affect which algae can secondarily attach. If a filamentous green alga can attach as a colonizer, there can be a large biovolume and biomass present (as was the case for Cladophora at RU-34 and CWR-72.3). However, if these algae are not present initially, they may have a difficult time gaining a holdfast, which would result in reduced biovolume, but may increase biodiversity. Three rock samples were collected at UM-872, near Anoka, for quality assurance purposes. This provided one basis for assessing precision in characterizing the density, biovolume and form of algae at a given site from a consistent substrate type. Some observations from this comparison follow:

Sample A Sample B Sample C

(rep. 1) (rep. 2) 1. # of species accounting for 80% of biovolume 9 12 12 2. Algal forms contributing to 80% of Diatom 8 (17, 59%) 7 (17, 52%) 11 (26, 60%)

biovolume (total species found, Green -- ( 5, 17%) 2 ( 7, 21%) -- ( 9, 21%) % of total) Blue-green 1 ( 7, 24%) 3 ( 9, 27%) 1 ( 8, 19%)

Sum 9 (29) 12 (33) 12 (43) 3. Total biovolume (µm3/cm2) 216,515,943 227,948,884 129,906,614 4. Total density (#cells/cm2) 1,188,846 1,567,368 736,078 Based on this comparison the three samples provide a similar indication of the number of species that comprise 80 percent or more of the biovolume. There is some variability, though, for the total number of species which ranged from 29-43. Of the species that comprised the upper 80 percent only four were common to all three samples and only the diatom Amphora pediculus was consistently among the dominant species in each sample at 15, 12, and 29 percent respectively. Variability in the estimates of total biovolume and total density was evident based on the three replicates. For example, total biovolume of sample C was 40 percent lower than sample A and total density was 38 percent lower, while sample B was five percent and 24 percent higher, respectively. This suggests care must be taken when comparing total biovolume and/or total density among sites. Comparisons of dominant algal forms were fairly comparable among the replicates. Benthic chlorophyll-a concentrations ranged from 2.1 to 150 mg/m2 among all samples (Lee, 2002). Individual measurements on wood substrate ranged from 4.0 mg/m2 at BE-73.2 up to 117 mg/m2 at UM-872 (Fig. 4). Individual measurements on rock substrate ranged from 2.1 mg/m2 at UM-1056 to 150 mg/m2 at UM-872. There was no consistent relationship between chlorophyll-a concentrations found on wood vs. rock substrate based on Figure 3 and linear regression of paired data (R2 = 0.001).

15

Diurnal measurements Diurnal measurements of DO, pH, temperature, and specific conductance were made at 12 sites on the Crow Wing, Mississippi, Crow, Rum, Blue Earth and Red Rivers during a period of 3-6 days in August 2000. River flow for this period in 2000 was below the long term summer-average on all rivers with the exception of the Red River where flows were slightly above the long-term average (Lee, 2002). In general, DO concentrations, DO percent saturation, pH, and temperature values increased, and specific conductance values decreased during daylight hours (Lee, 2002). These trends were reversed during the nighttime period. Water temperatures generally ranged between 17-28 degrees C (across all sites) over the 3-6 days of measurement. Water temperature variance over the period ranged from about 7 degrees C (17-24 degrees C) at CWR-72.3 to about 2 degrees C (23-25 degrees C) at UM-1056. pH values were generally between 8.0 – 8.8 at most sites with pH values up to 9.3 noted at RU-18. Specific conductance ranged from about 290-300 µmhos/cm at CWR-72 to 590-690 at CR-23. Specific conductance often declined in response to precipitation events. Diurnal DO flux, measured as the mean daily difference between the minimum and maximum DO, varied considerably among the different sites (Fig. 4). Minimal DO variation was noted at UM-1056 and RE-452 with a mean flux of 0.5 mg/L. At these sites DO typically ranged between 6.5-7.5 mg/L and DO saturation was about 80-90 percent. In contrast at the Crow and Blue Earth River sites DO flux ranged between 5-7 mg/L (Fig. 4). During the daytime hours DO was often supersaturated at these sites and DO concentrations up to 16 mg/L were measured at BE-73.2. Net community primary production (4-day mean) ranged from 0.05-0.07 gO2/m3/hr at RE-452 and UM-1056 respectively up to 0.93 gO2/m3/hr at BE-54. Conversely, net community respiration ranged from .02- 0.04 gO2/m3/hr at UM-1056 and the Red River sites to 0.43-0.44 gO2/m3/hr at BE-73.2 and the Crow River sites.

16

Figure 2. Periphyton percent biovolume by substrate type: (a) Rocks (b) Wood. Sites sorted based on summer-mean TP for 2000. (a)

(b)

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

% B

iovo

lum

e

CWR-35 CWR-72 UM-1056 UM-872 RU-34 RU-18 BE-54 BE-73 CR-0.2 CR-23 RE-536 RE-452

Biovolume on rocks. Sorted by TP

Greens BGA Diatom Family Crypto Misc

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

% B

iovo

lum

e

CWR-35 CWR-72 UM-1056 UM-872 RU-34 RU-18 BE-54 BE-73 CR-0.2 CR-23 RE-536 RE-452

Biovolume on wood. Sorted by TP

Greens BGA Diatom Family Crypto Misc

17

Figure 3. Periphyton chlorophyll-a by substrate type

Figure 4. Dissolved oxygen flux based on continuous measurement.

Periphyton chlorophyll-a by substrate type: 2000. Mean TP in parenthesis

0

20

40

60

80

100

120

140

CWR-72

(34)

CWR-35

(49)

UM-1056(59)

UM-872(84)

RUM-18

(133)

RUM-34

(143)

BE-73(205)

BE-54(207)

RE-536

(208)

CR-0(284)

RE-452

(312)

CR-23

(349)

Chl

-a m

g/m

2

Wood Rock

DO Flux (min, max, & median): Based on 3-6 days, August 2000.

0.02.0

4.06.08.0

10.012.014.0

16.018.0

CWR-72

CWR-35.5

UM-1056

UM-872

RUM-34

RUM-18CR-23

CR-0.2

BE-73.2

BE-54

RE-536

RE-452

DO

ppm

18

B. Water Quality and Aquatic Community Patterns and Relationships within Streams: 1999 & 2000

Data for each stream were analyzed individually to aid in characterizing relationships among TP, chlorophyll, BOD5 and algal composition. These analyses serve to describe and compare: (a) water quality between sites on the same stream; (b) changes in concentration as a function of flow; (c) changes as a combined function of time and flow over the course of both summers, and (d) fish IBI. Results proceed from the nutrient-poor NLF sites (Crow Wing and upper sites on Mississippi), to more nutrient-rich NCHF-NLF transition sites (lower sites on Mississippi and Rum), to the highly nutrient-rich NCHF-WCBP transition (Crow), WCBP (Blue Earth) and Red River sites.

Crow Wing River

The upper site on the Crow Wing at river mile (RM) 72.3 drains about 1,030 square miles (2,668 km2) on the western edge of the NLF ecoregion. USGS annual and summer flow, at this site, average 484 and 449 cfs respectively. The downstream monitoring site at CWR-35.5 has a watershed of about 2,130 square miles. Much of the additional watershed entering between the upstream and downstream site is from the NCHF ecoregion (Fig. 1 and Table 1). Flow in summer 2000

was near average and this contributed to a slightly longer residence time (7.0 days vs. 6.7 days at CWR-72.3) in comparison to 1999 when flows (Table 2) and precipitation (Appendix IV) were well above average. Average residence time at CWR-72.3 and CWR-35.5 was estimated at 6.7 days (based on measured flow and watershed area) and 9.7 days (based on estimated flow and watershed area) respectively for 1999 which suggests about 3 days (72 hours) of residence time between the two sites. The habitat features of the Crow Wing River are indicative of the forested and relatively undeveloped nature of its watershed (Appendix II-Table 1). Both sites were dominated by a variety of substrate materials, ranging in size from sand to boulders, and were free of silt (Appendix II-Table 2). A moderate amount of instream cover was present at both sites, including woody debris, boulders, and aquatic macrophytes. The downstream site also had undercut banks and overhanging vegetation. Riffles containing coarse substrate material (boulders, cobble, and gravel) were present at both locations. With the exception of the left bank at the downstream location, an undisturbed riparian width greater than 50 meters was present. Land use beyond the riparian area was primarily forested. QHEI scores for the upstream and downstream sites were 79 and 80, respectively, which is reflective of sites with good habitat and stable channels. TP and flow were weakly related at the upstream site (CWR-72.3) in 1999 (Fig. 5a, e) as is common for sites without significant upstream point source discharges. TP concentrations declined from 40 µg/L in July and to about 20 µg/L in September as flows declined. Elevated TP concentrations were noted at the downstream (CWR-35.5) site.

19

TP concentrations at the downstream site ranged from 50 µg/L in early July increased to 80 µg/L in mid-July as flow increased and declined as flows declined. A similar range of concentrations was noted in 2000. Chlorophyll-a concentrations were low over the sampling period, ranging from about 7 µg/L in July to about 3 µg/L in September (Fig. 6). Total chlorophyll increased with flow and declined as flow subsided. A comparison of chlorophyll-a between the two sites revealed similar patterns and no overall significant difference in concentration was noted (Table 2). Based on Basu and Pick (1995) a residence time of 72 hours may not provide adequate time for growth of phytoplankton over this reach. Since there was no significant between-year difference in TP concentrations for these two sites (Table 3) it is presumed that the lower and more stable flows in 2000 contributed to the slight increase in chlorophyll. BOD5 concentrations were typically low at both sites and were generally at or below 1.5 mg/L. No distinct difference in BOD5 was evident between the two sites based on 1999 and 2000 data (Table 3). Diatoms were the predominant form of phytoplankton on all dates in 1999, based on three algal samples from CWR-35.5 for July 20, August 17 and September 9 (Fig. 7). The most common genera were the benthic forms: Cocconeis, Nitzschia, Acananthus and Navicula. Green algae, represented by Scenedesmus and Ankistrodesmus, were the second most common form. Blue-green algae were present in low numbers later in the summer and were characterized as Oscillatoria and “non-motile” blue-greens. As flows declined in 1999 periphyton was visibly evident on rocks and substrate on the stream bottom downstream from the sampling point (Plate 2). Diatoms were again the most common form in 2000 with Cocconeis, Navicula, Surirella and Fragilaria the most abundant genera at CWR-72.3. The Cryptomonads, including Cryptomonas and Rhodomonas and some filamentous blue-greens, including Aphanizomenon and Oscillatoria were among the dominants as well. The downstream sites (CWR-35.5) shared many of the same dominant species as noted for CWR-72.3. Also several diatoms including Cocconeis, Nitzschia and Navicula were common to the periphyton as well.

20

Table 3. Summer-means and maxima from Minnesota river-nutrient study for selected parameters in 1999 and 2000.

1999 mean mean Max mean mean mean mean mean mean mean mean mean mean mean mean max Site Tube Temp. Temp. pH Cond. Turb. BOD5 TSS TSV TKN NO3 TP Chl-a Pheo ChlT ChlT

cm C C SU umhos NTU mg.L-1 mg.L-1 mg.L-1 mg.L-1 mg.L-1 ug.L-1 ug.L-1 ug.L-1 ug.L-1 ug.L-1CWR-72 60 16 26 8.1 283 3 1.0 3.5 1.4 0.58 0.21 32 3.1 1.4 4 7CWR-35 60 16 25 7.8 386 4 1.0 6.0 2.8 0.77 0.22 59 2.4 1.9 4 9

UM-1004 25 20 25 7.4 216 18 1.1 23.3 3.4 0.81 0.09 71 4.5 3.8 8 14UM-965 40 20 26 7.6 259 12 1.2 15.4 2.7 0.75 0.14 63 4.0 3.9 8 12UM-953 37 21 26 7.7 263 10 1.1 13.1 2.7 0.72 0.15 62 4.4 2.9 7 11UM-895 48 21 28 7.7 275 9 1.2 12.8 3.1 0.76 0.20 67 5.2 4.5 10 15UM-872 37 22 28 7.9 278 12 1.5 18.9 4.9 0.88 0.38 92 15.6 6.6 22 33

RU-34 46 20 26 7.5 272 10 1.6 16.8 4.2 1.14 0.27 137 13.3 6.6 20 36RU-18 49 21 27 7.7 279 8 1.8 15.5 4.9 1.11 0.26 131 18.8 7.9 27 55

CR-23 12 22 29 8.1 588 53 4.5 73.0 18.0 2.06 2.11 359 83.4 21.2 105 154CR-0 15 21 28 8.0 574 49 4.0 75.0 17.0 1.92 1.86 329 74.1 19.6 94 132BE-94 15 19 25 7.9 668 59 2.1 125.0 17.8 1.17 7.64 247 29.1 12.0 41 84BE-73 13 20 25 7.9 636 57 3.6 110.0 17.7 1.55 6.28 243 47.7 14.7 62 99BE-54 12 21 26 8.0 636 68 3.4 126.0 20.4 1.47 6.41 248 64.4 17.0 81 150BE-18 13 22 27 8.1 621 68 3.4 135.0 20.9 1.44 6.28 240 57.6 16.1 74 147

2000 mean mean max mean mean mean mean mean mean mean mean mean mean mean mean max Site Tube Temp. Temp. pH Cond. Turb. BOD5 TSS TSV TKN NO3 TP Chl-a Pheo ChlT ChlT

cm C C SU umhos NTU mg.L-1 mg.L-1 mg.L-1 mg.L-1 mg.L-1 ug.L-1 ug.L-1 ug.L-1 ug.L-1 ug.L-1CWR-72 >60 20 25 8.4 313 2.5 1.2 3.3 2.3 0.57 0.12 34 3.4 1.5 5 14CWR-35 >60 20 24 8.2 372 3.3 1.2 6.1 2.7 0.70 0.23 49 3.7 1.7 5 9

UM-1056 53 20 24 8.0 262 11.7 1.1 19.5 2.9 0.72 0.07 59 4.7 2.1 7 9UM-1029 42 20 24 7.9 256 14.0 1.0 22.4 3.1 0.74 0.08 60 5.1 2.7 8 10UM-953 53 21 25 8.3 295 7.6 1.3 9.1 2.4 0.72 0.11 54 7.6 4.8 12 22UM-895 53 23 27 8.4 269 6.9 1.6 10.9 3.4 0.82 0.19 77 11.0 6.2 17 30UM-872 47 23 27 8.4 342 8.8 2.1 15.7 5.0 0.93 0.23 84 22.7 6.4 29 41

RU-34 56 18 24 8.1 337 6.2 1.8 10.6 3.7 0.97 0.28 143 20.5 7.2 29 41RU-18 52 18 24 8.4 329 6.3 2.3 11.2 4.8 0.97 0.18 133 31.4 10.7 44 65CR-23 13 21 26 8.5 736 40.6 6.6 74.6 18.1 1.94 1.69 349 120.3 25.5 143 213CR-0 15 19 24 8.5 721 31.5 7.0 63.8 17.6 1.76 1.56 284 112.4 26.6 136 207BE-100 42 18 24 8.3 723 16.4 1.1 35.0 5.4 0.57 7.18 116 6.4 4.8 11 23BE-94 23 19 25 8.1 698 30.6 2.7 60.8 10.7 1.14 6.11 192 41.8 6.7 48 121BE-73 17 20 26 8.3 675 40.6 5.1 74.4 15.3 1.63 5.55 205 87.4 15.4 101 188BE-54 15 21 26 8.3 662 46.3 6.3 90.9 18.1 1.60 5.35 207 86.7 11.5 97 195BE-18 15 21 27 7.3 620 57.4 5.3 108.3 18.9 1.63 5.37 223 73.1 10.2 82 177RE-536 22 22 7.5 351 27.0 2.8 55.0 8.6 1.3 0.2 208 18.9 8.3 27 43RE-452 14 21 7.2 389 69.0 2.1 144.0 19.1 1.4 0.3 312 23.2 13.5 38 96RE-403 7 23 6.9 443 151.0 4.7 374.0 45.6 2.9 0.6 602 16.4 21.5 38 67RE-298 7 20 6.7 537 148.0 2.9 324.0 32.6 1.7 0.6 502 10.4 15.1 26 100

1 Notes: Cond.=specific conductivity in µmhos/cm; TSS=total suspended solids, TSV=total suspended volatile solids; ISS (inorganic suspended solids) = TSS-TSV; TN=TKN+NO3; Total chlorophyll-a (ChlT) = chlorophyll-a (chl-a) + pheophytin (pheo)

21

Low periphyton diversity was noted for the CWR-72.3 and CWR-35.5 sites in 2000. The dominants at the upstream site (rock) were comprised of Chlorococcus, Cocconeis, and non-motile blue-greens. The wood sample at CWR-72.3 represented a stark contrast as it was completely covered by the filamentous green alga Cladophora The downstream site (CWR-35.5) exhibited slightly more diversity than the upstream site, with the diatoms Melosira, Cocconeis, Nitzschia, Navicula and Rhoicosphenia among the dominants. Lyngbya, a filamentous blue-green, was fairly abundant as well. Several of these genera, diatoms in particular, were common to the phytoplankton community as well. Both sites on the Crow Wing River had a fish IBI rating of excellent (Appendix II-Table 3). There was a relatively high number of intolerant darter-sculpin-Notorus species present. Intolerant species are those known to be sensitive to environmental degradation. Species of darter, sculpin, and Notorus are generally found in higher quality streams in the Upper Mississippi River Basin (Niemela and Feist 2002). Twenty-seven species were collected from the downstream site on the Crow Wing River, which was the greatest species richness from any sampling event included in this study.

22

Figure 5. Total phosphorus and flow for June – Sept. 1999 and 2000 selected sites (a,b,c,d). TP and flow regressions (e). (a) (b)

Crow Wing (CWR-72.3)

0

200

400

600

800

1,000

1,200

1-Jun

16-Ju

n1-J

ul

16-Ju

l

31-Ju

l

15-A

ug

30-A

ug

14-S

ep

29-S

ep

13-Ju

n

28-Ju

n13

-Jul

28-Ju

l

12-A

ug

27-A

ug

11-S

ep

26-S

ep

flow

cfs

0

10

20

30

40

50

60

TP p

pb

Flow TP

Rum (RU-18)

0

200

400

600

800

1,000

1,200

1-Jun

16-Ju

n1-J

ul

16-Ju

l

31-Ju

l

15-A

ug

30-A

ug

14-S

ep

29-S

ep

13-Ju

n

28-Ju

n13

-Jul

28-Ju

l

12-A

ug

27-A

ug

11-S

ep

26-S

ep

flow

cfs

0

50

100

150

200

250

300

TP p

pb

Flow TP

23

(c) (d)

Crow (CR-23)

0

500

1,000

1,500

2,000

2,500

1-Jun

16-Ju

n1-J

ul

16-Ju

l

31-Ju

l

15-A

ug

30-A

ug

14-S

ep

29-S

ep

13-Ju

n

28-Ju

n13

-Jul

28-Ju

l

12-A

ug

27-A

ug

11-S

ep

26-S

ep

Flow

cfs

050100150200250300350400450

TP p

pb

Flow TP

Blue Earth (BE-18)

0

2,000

4,000

6,000

8,000

10,000

12,000

1-Jun

16-Ju

n1-J

ul

16-Ju

l

31-Ju

l

15-A

ug

30-A

ug

14-S

ep

29-S

ep

13-Ju

n

28-Ju

n13

-Jul

28-Ju

l

12-A

ug

27-A

ug

11-S

ep

26-S

ep

flow

cfs

050100150200250300350400450500

TP p

pb

Flow TP

24

Crow Wing TP vs. Flow 1999 & 2000

R2 = 0.22

0

20

40

60

80

100

0 200 400 600 800 1,000 1,200

Flow cfs

TP u

g/L

RU-18 TP vs. Flow: 1999 & 2000

R2 = 0.12

0

50

100

150

200

250

300

0 200 400 600 800 1,000 1,200

Flow cfs

TP p

pb

CR-23 TP vs. Flow: 1999 & 2000

R2 = 0.05

0

100

200

300

400

500

0 500 1,000 1,500 2,000 2,500

Flow cfs

TP p

pb

BE-18 TP vs. Flow: 1999 & 2000

R2 = 0.59

0

100

200

300

400

500

0 2,000 4,000 6,000

Flow cfs

TP p

pb

(e)

25

Figure 6. Total chlorophyll-a and flow comparisons for selected sites.

Crow Wing

0

200

400

600

800

1,000

1,200

1-Jun

16-Ju

n1-J

ul

16-Ju

l

31-Ju

l

15-A

ug

30-A

ug

14-S

ep

29-S

ep

13-Ju

n

28-Ju

n13

-Jul

28-Ju

l

12-A

ug

27-A

ug

11-S

ep

26-S

ep

Flow

cfs

0

10

20

30

40

50

60

70

Chl

T pp

b

Flow ChlT

1999 2000

Rum

0

200

400

600

800

1,000

1,200

1-Jun

16-Ju

n1-J

ul

16-Ju

l

31-Ju

l

15-A

ug

30-A

ug

14-S

ep

29-S

ep

13-Ju

n

28-Ju

n13

-Jul

28-Ju

l

12-A

ug

27-A

ug

11-S

ep

26-S

ep

Flow

cfs

0

10

20

30

40

50

60

70C

hlT

ppb

Flow ChlT

1999 2000

(a)

(b)

26

Crow

0

500

1,000

1,500

2,000

2,500

1-Jun

16-Ju

n1-J

ul

16-Ju

l

31-Ju

l

15-A

ug

30-A

ug

14-S

ep

29-S

ep

13-Ju

n

28-Ju

n13

-Jul

28-Ju

l

12-A

ug

27-A

ug

11-S

ep

26-S

ep

Flow

cfs

0

50

100

150

200

250

Chl

T pp

b

Flow ChlT

1999 2000

Blue Earth

0

2,000

4,000

6,000

8,000

10,000

12,000

1-Jun

16-Ju

n1-J

ul

16-Ju

l

31-Ju

l

15-A

ug

30-A

ug

14-S

ep

29-S

ep

13-Ju

n

28-Ju

n13

-Jul

28-Ju

l

12-A

ug

27-A

ug

11-S

ep

26-S

ep

Flow

cfs

0

50

100

150

200

250

Chl

T pp

b

Flow ChlT

1999 2000

(c)

(d)

27

Mississippi River

The farthest upstream site monitored in 1999 was at UM-1004 in Brainerd (Fig. 1). The river at this site drains about 7,320 square miles in the NLF ecoregion. As the river flows from the Brainerd area more drainage is derived from the NCHF ecoregion (Fig. 1). At UM-953.7 (Royalton) about 20 percent of the 11,600 square mile watershed is derived from the NCHF ecoregion. This is also supported by flow data whereby the upstream Mississippi (UM-1004, Aitkin) flow combined with the NLF portion of the Crow Wing (CWR-72.3) flow would account for about 80 percent of the flow measured at UM-953.7. At UM-

895 (Monticello) the watershed, on an areal basis, is characterized almost equally by both ecoregions. At the most downstream Mississippi site monitored, UM-872 (Anoka, photo), the majority of the watershed (65 percent) lies in the NCHF ecoregion but about 50 percent of the flow at UM-872 is derived from the NLF portion of the watershed. Flow was high at all sites in 1999 but tended to be highest, relative to the long-term mean, at the upstream (NLF) sites (Table 2) where precipitation was well above normal (Appendix IV). TP concentrations at the four upstream sites ranged from about 60 to 80 µg/L during 1999 and in general fluctuated positively with flow (Fig. 5). No significant difference in TP concentrations was noted based on mean ± standard error for these four sites (Table 3). The most downstream site (UM-872) exhibited a wider range in concentrations (60 to 120 µg/L) and consistently higher TP concentrations as compared to the upstream sites (Table 3). TP loads from more nutrient-rich watersheds, such as the Crow River, that enter between UM-895 and UM-872 contribute to the higher concentration noted at UM-872. TP concentrations fluctuated positively with flow at UM-872 as well during 1999 (Fig. 5). These patterns were similar in 2000 as well. Summer-mean chlorophyll concentrations were generally between 5-10 µg/L and maxima did not exceed 15 µg/L at the four upstream sites in 1999 (Table 3). In contrast, chlorophyll concentrations at UM-872 were consistently higher, ranging between 10 - 33 µg/L, and averaged 22 µg/L (Table 3). Concentrations peaked in July and generally declined thereafter during the high flows of August and September. For sites monitored in both years chlorophyll concentrations were much higher in 2000 as compared to 1999. Water residence time at UM-872 averaged 27.4 and 29.7 days, respectively, for 1999 and 2000. Relative to UM-953.7, this represented increases of about five days – more than adequate to allow for increases in chlorophyll-a over this reach. BOD5 concentrations were quite low at all sites and generally ranged between 0.8 to 1.3 mg/L (Table 3). The highest concentration (2.1 mg/L) noted was at UM-872 and coincided with the peak chlorophyll concentration on July 15. A similar pattern (upstream-downstream) was evident in 2000, however BOD5 concentrations were slightly higher as compared to 1999 (Table 3).

28

Phytoplankton composition shifts somewhat as we move from the NLF portion of the Mississippi at UM-1004 to the transitional UM-953.7 and UM-872 sites based on collections from 1999. While diatoms, including Cyclotella, Cocconeis, Fragilaria, and Melosira, and Cryptomonads, including Chryptomonas and Rhodomonas, were a significant portion of the population at UM-1004, they decline in significance at UM-953.7 and UM-872 (Fig. 7). Greens, including Ankistrodesmus and Scenedesmus, and blue-greens including Oscillatoria and Anabaena were more prominent at the latter two sites. Phytoplankton at UM-1056 in 2000 was comprised of diatoms, greens, blue-greens, Cryptomonads and Euglenoids. Diatoms were represented by several species with Nitzschia, Cocconeis, and Stephanodiscus contributing the most to overall biovolume. Of these, only Cocconeis was found in the periphyton samples. Green algae were prominent as well with Chlamydomonas and Oocystis being the most common. The phytoplankton samples at UM-872 in 2000 were equally diverse. Aulacoseira, Nitzschia, and Cyclotella were the most common diatoms. Blue-green algae represented by Oscillatoria and Aphanizomenon were common on two sample dates. Greens, including non-motile Chlorococcales, Scenedesmus, Oocystis, and Closterium were present but were often a small portion of the biovolume. Two Cryptomonads, Rhodomonas and Cryptomonas, were among the dominant forms on two sample dates. Likewise the Euglenoids, Euglena and Phacus, were present as well. Periphyton samples were collected at UM-1056 and UM-872. The periphyton community at UM-1056 exhibited minimal diversity and both the rock and wood samples were dominated by diatoms. The diatoms Epithemia and Rhoicosphenia accounted for 80 percent of the biovolume on the rock sample, while Melosira, Cocconeis and Rhoicosphenia accounted for 38 percent of the wood biovolume. The filamentous green Oedogonium accounted for 44 percent of the wood biovolume. Of these genera, only Cocconeis was prominent in the phytoplankton samples at this site. Periphyton at UM-872 was quite diverse with diatoms, greens, and blue-greens all being present. Rock and wood samples were fairly similar and both had high percentages of diatoms. Navicula, Amphora, Pinnularia, Cocconeis, and Gomphonema were among the most common diatoms. Greens, represented by Chlorococcales and Chlorococcum, were present at low densities. Non-motile blue-greens were the predominant blue-green form. The periphyton and phytoplankton had few genera in common in 2000. Fish and habitat assessments were conducted on five sites were located on the Mississippi River. The three upstream sites were surrounded by forest, while the two downstream locations were in residential/urban areas. Bank erosion was moderate at the most upstream location, which was dominated by fine substrate materials. Coarse substrates (boulder, cobble, and gravel) increased in dominance from upstream to downstream sites. Downstream locations also had a high degree of channel stability. Instream cover for fish, which was most prevalent at sites UM-953 and UM-895, consisted of woody debris, aquatic macrophytes, boulders, and overhanging vegetation. Deep pools and

29

shallows were present at all sites except for UM-1056, which lacked riffles. QHEI scores on the Mississippi River ranged from 50 (UM-1056, Aitkin) to 68 (UM-953, Royalton). The biological integrity rating for the Mississippi River ranged from fair to excellent, with the site located at Royalton (UM-953) having the highest value. Sites with the highest IBI had the greatest species richness and number of intolerant species. IBI scores increased through the first three sites (Aitkin to Royalton), but were generally lower at the downstream locations. The percent tolerant and percent omnivore species were highest in the two sampling events at UM-895, near Monticello.

Rum River With the exception of Lake Mille Lacs at its headwater, the Rum River primarily drains the NCHF ecoregion. Two downstream sites at Isanti, Minnesota (RUM-34) and St. Francis, Minnesota (RUM-18) were monitored in 1999 and 2000. No significant streams enter between these sites and about 170 square miles of watershed are added between the sites. Period of record annual and summer-mean flows average 630 and 585 cfs respectively at St. Francis. Summer-mean flow in 1999 was slightly higher at

596 cfs, while summer flow for 2000 was below average at 277 cfs. Residence time was on the order of 8.1 days in 1999 and 8.7 days in 2000 at RU-18. Residence time at the upper site (RU-34) was slightly less in each year, approximately 7.6 and 8.2 days respectively for 1999 and 2000 (Table 1). No significant difference in TP was noted between the upstream (RU-34) site and the downstream (RU-18) site in either year (Table 3). In general, TP concentrations fluctuated positively with flow (Fig. 5b), however the relationship between flow and TP was rather weak (Fig. 5e). Baseflow TP concentration was on the order of 75-100 µg/L in both years. Chlorophyll did, however, exhibit a downstream increase over this 16-mile reach (Table 3) in both years and the difference was most pronounced in 2000. Based on residence time alone, we would not anticipate significant increases in chlorophyll between the two sites. Chlorophyll did not vary consistently with flow in the Rum River, as chlorophyll increased both during declining (low) flows as well as with increasing flow (Fig. 6). Peak concentrations for 1999, 36 and 55 µg/L, were noted for RU-34 and RU-18 respectively on July 26. River flow had declined for several days prior to July 26 (Fig. 6), however, a significant increase in flow was noted on July 26 (125 cfs increase). This likely brought benthic algae into suspension by scouring substrates. The August and September samples were collected on the declining hydrograph (Fig. 6) and chlorophyll concentrations were lower than earlier in the summer. In 2000 chlorophyll-a concentrations were higher at both sites on average and some of the highest concentrations were associated with low flow in August and September.

30

Figure 7. Phytoplankton composition for 1999 and 2000.

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Alg

al c

ompo

sitio

n

8/17/1999 CWR-35.5 8/9/2000 CWR-35.5 9/21/1999 CWR-35.5 9/11/2000 CWR-35.5

Crow Wing River 1999 v 2000

MiscDinoflBGACryptosGreensDiatom Family

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Alg

al c

ompo

sitio

n

7/19

/199

9 R

U-1

8

7/19

/200

0 R

U-1

8

8/9/

1999

RU

-18

8/17

/200

0 R

U-1

8

9/16

/199

9 R

U-1

8

9/25

/200

0 R

U-1

8

7/12

/199

9 R

U-3

4

7/19

/200

0 R

U-3

4

8/9/

1999

RU

-34

8/17

/200

0 R

U-3

4

9/16

/199

9 R

U-3

4

9/25

/200

0 R

U-3

4

Rum River 1999 v 2000


(a)

(b)

31

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Alg

al c

ompo

sitio

n

7/26

/199

9 C

R-0

.2

7/19

/200

0 C

R-0

.2

8/9/

1999

CR

-0.2

8/17

/200

0 C

R-0

.2

9/16

/199

9 C

R-0

.2

9/25

/200

0 C

R-0

.2

7/26

/199

9 C

R-2

3

7/19

/200

0 C

R-2

3

8/9/

1999

CR

-23

8/17

/200

0 C

R-2

3

9/16

/199

9 C

R-2

3

9/25

/200

0 C

R-2

3

Crow River 1999 v 2000


0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Alg

al c

ompo

sitio

n

8/16

/199

9 U

M-8

72

8/10

/200

0 U

M-8

72

9/22

/199

9 U

M-8

72

9/12

/200

0 U

M-8

72

7/20

/199

9 U

M-9

56.2

7/13

/200

0 U

M-9

53

8/3/

1999

UM

-956

.2

8/24

/200

0 U

M-9

53

9/8/

1999

UM

-956

.2

9/12

/200

0 U

M-9

53

Upper Mississippi River


(c)

(d)

32

BOD5 concentrations in 1999 were generally 1.5 mg/L or less on each sample date, with the highest concentrations coinciding with the higher flows of July. As chlorophyll increases at the upstream site a corresponding increase in BOD5 is noted downstream. And as chlorophyll declines, a similar decline in BOD5 is noted too. Patterns were similar in 2000 but summer-mean BOD5 was slightly higher as compared to 1999. Phytoplankton was collected on three dates in 1999 at both sites. No single algal form remained dominant over the summer and on average, no substantial difference was noted in the composition of algae between the two sites (Fig. 7). Diatoms comprised from about 12 to 39 percent of the algal samples from the Rum. In the July samples the benthic Nitzschia was the predominant genera, with Cyclotella, Fragilaria, and Melosira becoming more prominent in the August and September samples. Cyclotella is considered planktonic while species of Fragilaria and Melosira may be either benthic or planktonic according to Kelly (1998). The Cryptomonad, Rhodomonas, was a significant portion (45 % of sample) of the July 12 sample. Cryptomonads were present but not abundant in samples from other study sites in 1999. Green algae typically represented from 10-25 percent of the algae and Scenedesmus. was the most common genera. “Non-motile” blue-green algae ranged from 10 to 63 percent of the algae samples and were a significant part of the seston on July 19, August 9, and September 16.

Phytoplankton in 2000 was dominated by the diatoms, Aulocoseira, Surirella, and Amphora, early in the summer. Blue-green blooms of Lyngbya, Anabaena, Aphanizomenon were common in July. Lyngbya dominated the sample in September along with the diatoms Navicula and Nitzschia.

Periphyton at RU-34 was dominated by diatoms (11 genera) including: Cyclotella, Navicula, Amphora, Nitzschia, Rhoicosphenia, and Fragilaria and greens (2 genera) including: Stigeoclonium and Chlorococcum. At RU-18 Cladophora, a filamentous algae, accounted for 53

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Alg

al c

ompo

sitio

n

8/10/1999BE-54

8/1/2000 BE-54

9/22/1999BE-54

9/27/2000BE-54

8/10/1999BE-94.3

8/30/2000BE-94.3

9/22/1999BE-94.3

9/18/2000BE-94.3

Blue Earth River 1999 v 2000


(e)

33

percent of the periphyton on the rock sample. Other prominent greens included Pediastrum and Chlorococcum. The diatoms Cocconeis, Navicula , and Rhoicosphenia were among the dominants as well. Diatoms accounted for about 75 percent of the periphyton on the wood sample at this site with many of the aforementioned genera represented. Overall the periphyton and phytoplankton communities at RU-18 had a few genera, such as the diatoms - Navicula, Nitzschia, and Fragilaria in common based on these collections. QHEI scores ranged from 60 at the upstream site (RU-34) to 78 at the downstream site (RU-18). Both sites were visited two times and QHEI scores between the sampling events varied by 5 units or less (Appendix II, Table 2). Residential development and forested land surrounded the upstream location, while the downstream location was forested. Both sites contained riffles with coarse substrate, and had a moderate amount of instream cover for fish. Both sampling events at each site yielded an IBI rating of good (Appendix II, Table 3). Species richness was slightly higher at the upstream location. The percent tolerant individuals, omnivores, and simple lithophils and the number of invertivore species were greater at the downstream site. Simple lithophils require clean gravel substrate to spawn, and the percent present is inversely correlated with habitat degradation due to excessive siltation (Berkman and Rabeni 1987). Invertivores are dependent upon a stable invertebrate food base; disruptions in this food base through human disturbance can lead to a decrease in the number of invertivore species (Niemela and Feist 2002).

Crow River

The Crow River has two very distinct subwatersheds. The North Fork comprises about 53 percent of the watershed, draining from the NCHF ecoregion with numerous lakes and wetlands throughout the watershed (Fig. 1). The South Fork, comprising about 47 percent, drains primarily from the agricultural WCBP ecoregion. Feedlots are common throughout both watersheds (MPCA, 2000). The highly agricultural land use, combined with numerous small wastewater discharges dotted throughout both watersheds,

contributes to the “nutrient-rich” conditions in the Crow. Long-term annual and summer mean flow at CR-23 are 816 (23.1 m3/s) and 839 (23.8 m3/s) cfs, respectively. Summer-mean flow in 1999 was slightly above average at 886 cfs, while 2000 was below average at 301 cfs. Total phosphorus concentrations were very high at both sites in 1999 and 2000, typically ranging between 250 – 400 µg/L (Fig. 5). In the Crow there was no significant difference in summer-mean TP between years at the two sites, however a slight downstream decrease in TP from the upstream (CR-23) to the downstream (CR-0.2) site occurred in both years (Table 3). This suggests that sedimentation of TP may be occurring over this reach and/or the possibility of dilution (intermixing) by the Mississippi River at the CR-0.2 site when the Mississippi is at a high stage, as it was throughout the summer of 1999. Meals et al. (1999) note seasonal retention of nutrients in short reaches of streams is common and given low flows and low gradient of the Crow, over this reach, seasonal retention seems a distinct possibility. No consistent pattern in TP concentrations was evident between the two sites in 1999 nor 2000 (Fig. 5). TP concentrations

34

remained relatively high throughout the summer and no direct correspondence to flow was evident (Fig. 5e); if anything, concentrations tended to increase with decreasing flow (Fig. 5c). Baseflow TP concentrations were on the order of 300-350 µg/L in both years. This is most likely a reflection of the numerous point source discharges in the watershed that serve to keep concentrations elevated during low flow. Chlorophyll concentrations were high throughout 1999 and ranged from about 50 to 150 ug/L. No consistent upstream – downstream pattern was evident, but on average, concentrations were slightly higher at the upstream site (Fig. 6). Again as noted above, sedimentation and/or dilution from the Mississippi River may influence concentrations at CR-0.2. Chlorophyll concentrations at both sites varied inversely with flow (Fig. 6). As flow decreased from mid July to mid-August chlorophyll concentrations increased at Rockford and peaked at 154 µg/L on August 9. An abrupt increase in flow (~ three-fold increase) from August 9 to August 24, 1999 resulted in lower chlorophyll concentrations on August 25. Summer-mean concentrations in 2000 were higher at both sites as compared to 1999. Again chlorophyll-a tended to vary inversely with flow (Fig. 6). Travel time between CR-23 to CR-0.2 varied from about 20 hours at 1,500 cfs up to about 50 hours at 200 cfs (Fig. 21). Considering average flows of about 900 cfs and 300 cfs respectively for 1999 and 2000 this would translate to travel (residence) time over this reach of about 25 to 35 hours – likely insufficient to allow for significant production of algae over this reach. Chlorophyll-a concentrations were significantly higher in 2000 as compared to 1999 as residence time increased from about 11.3 to 12.6 days (1.3 day or 12 percent increase). BOD5 concentrations ranged from about 2 to 7 mg/L at both sites in 1999. Concentrations were slightly higher at the upstream site averaging 4.5 mg/L as compared to 4 mg/L at the downstream site. A comparison of chlorophyll at upstream CR-23 to BOD5 at CR-0.2 shows good correspondence between these two measures. As chlorophyll increases at the upstream site a concurrent downstream increase in BOD5 is evident on most dates. Both chlorophyll and BOD5 peaked on the August 9 sampling, which corresponded to low flow conditions. Summer-mean BOD5 was higher in 2000 as compared to 1999. Phytoplankton at CR-23 in 1999 was dominated by the filamentous blue-greens, Oscillatoria, Anabaena, Aphanizomenon, and non-motile blue-greens. Blue-greens, on average, comprised 79 percent of the algae in the three samples from CR-23 and ranged from 78 to 81 percent (Fig. 7). Diatoms were common, but at lower levels, and consisted primarily of Melosira, Cyclotella, and Stephanodiscus. At CR-0.2 blue-greens including Oscillatoria and Aphanizomenon were also prominent in the 1999 samples. Phytoplankton in 2000 consisted primarily of diatoms, greens and Cryptomonads. Diatoms ranged from 8 to 57 percent of the algae assessed at CR-23 in 2000 and Cyclotella, Stephanodiscus, and Nitzschia were the most common genera. The green algae were present at low levels (5-11 percent) and the planktonic Scenedesmus and Ankistrodesmus were the predominant genera. Diatoms, including Stephanodiscus, Cyclotella, and Aulocoseira, were also prominent at CR-0.2. Scenedesmus, Pediastrum, and non-motile Chlorococales were among the dominant greens. Euglena was particularly abundant in the September sample. Blue-green algae were not prominent in the 2000 samples.

35

Periphyton at CR-23 was dominated by several genera of diatoms including: Navicula, Melosira, Cocconeis, Epithemia, and Nitzschia. Dominant greens were represented by Oedogonium, and Chlorococcum. Blue-greens were not very prominent in the periphyton samples. At CR-0.2 the periphyton samples were dominated by diatoms and represented by about seven primary genera including - Cyclotella, Melosira, Navicula, Nitzschia, Cocconeis, Gomphonema, and Amphora. In general there were some similar genera (diatoms principally) found in the periphyton and phytoplankton in 2000. The two sites on the Crow River had both coarse and fine substrates (cobble, gravel, and sand) (Appendix II, Table 2). The upstream location had a riparian area 10 to 50 meters wide on either bank followed by residential development. The downstream location had a similar riparian area, but was bordered by row crop. Shallows, deep pools, boulders, and woody debris provided instream cover. QHEI scores for the Crow River sites were 68 and 60 respectively for CR-23 and CR-0.2. The biotic integrity rating for the Crow River sites was good at the upstream site (CR-23) and fair at the downstream site (CR-0.2), consistent with the QHEI scores. The total taxa, number of intolerant species, percent omnivores, number of piscivore species, and the number of invertivore species declined between these sites.

Blue Earth River The Blue Earth River in south central Minnesota has its headwaters in Iowa and drains an area of about 3,542 square miles before entering the Minnesota River. The main stem of the Blue Earth above the confluence of the LeSueur and Watonwan Rivers was the focus for this study. This reach drains about 1,527 square miles, above the gauge at Rapidan, Minnesota. The annual and summer-mean flows at Rapidan are 1,071 and 1,184 cfs respectively. At 2,110 cfs summer-mean flow in 1999

was above the long-term mean, while at 1,073 cfs summer-mean 2000 flow was below average. Five sites arrayed from near the Iowa border (BE-94) to a site upstream of the Rapidan Dam (BE-18) were included in 1999. The Blue Earth River exhibited high flow in 1999 and average flow in 2000 – though August and September low flows were evident in both years. Total phosphorus concentrations were fairly similar among sites and averaged between 233 to 248 µg/L in 1999 (Table 3). Concentrations ranged from about 250 µg/L in early July, peaked at 400 – 500 µg/L on July 21 and declined thereafter to about 100 µg/L or less by September. TP concentrations were highly correlated with flow based on comparisons at BE-18 (Fig. 5d, e), which suggests the strong influence of NPS runoff in regulating the TP concentration at this site. A major runoff event on July 20 caused flows to double over a two-day period from July 19-20. This accounted for the peak TP on July 21 (Fig. 5). Flow peaked on July 25 and generally declined thereafter. TSS concentrations peaked at 440 mg/L on July 21 at the upstream (BE-94.3) site. In 2000 summer-mean concentrations were slightly lower than 1999, however concentrations tended to increase in a downstream direction. Baseflow concentrations at BE-18 were on the order of 50 – 150 µg/L (Fig. 5).

36

Chlorophyll-a concentrations generally increased in a downstream direction in 1999 with the highest concentrations typically occurring at the two most downstream sites (BE-54 and BE-18). Concentrations in 2000 increased sharply between BE-94.3 and BE-73.2 and remained relatively stable between the three downstream sites (Table 3). While TP concentration did not increase over this reach watershed area increased from 804 mi2 at BE-94.3 to 1,367 mi2 at BE-73.2. The larger watershed area (70 percent increase) would provide increased flow and longer residence time at the downstream site. Based on flow and chlorophyll-a comparisons at BE-18 for 1999, chlorophyll-a increased during stable to low flow conditions, though some high flow related increases in chlorophyll-a were noted as well. Perhaps the best pairing of sites for examining downstream changes in chlorophyll-a would be the sites at BE-73.2 and BE-54. Over this 19-mile reach no significant tributaries enter and only about 24 square miles of watershed are added. BE-73 averaged 75 µg/L while BE-54 averaged 98 µg/L (an increase of 31%) and on most dates concentrations were higher at the downstream site. Chlorophyll concentrations did not vary consistently with flow. The major runoff event on July 20-21, 1999 served to increase concentrations – presumably by scouring and bringing benthic algae into suspension (Fig. 6). Chlorophyll was low on July 27, 1999 following this event and as flow declined. Stable flows through much of August contributed to the elevated seston chlorophyll concentrations noted at most sites for the remainder of the summer. Summer-mean and maxima were higher for all sites in 2000 as compared to 1999 and concentrations tended to be highest during the low flows of August and September. BOD5 concentrations were similar between BE-73.2, BE-54 and BE-18.2, and averaged between 3.0 to 3.5 mg/L (Table 3). A comparison of upstream chlorophyll-a at the upstream BE-54 site to the downstream BOD5 at the downstream BE-18 site showed good correspondence between the two measures. As chlorophyll-a concentrations rose (or fell) at the upstream site a corresponding change was noted at BE-18. The peak BOD5 of 6.5 mg/L coincided with the peak chlorophyll-a of 145 µg/L at BE-54. Summer-mean BOD5 was higher at all sites in 2000 as compared to 1999. In 1999 phytoplankton samples were analyzed from both the upper site at BE-94.3 and the downstream BE-54. On average blue-greens were prominent at both sites followed by diatoms and greens at BE-94.3 and greens and diatoms at BE-54 (Fig. 7). Diatoms were most prominent on July 21 with the genera Nitzschia, Navicula, Achnanthes, Cyclotella, and Synedra among the dominant genera. With the exception of Cyclotella all others are considered benthic forms (Kelly, 1998). Green algae represented by Ankistrodesmus, Phacus, and Scenedesmus increased from nine percent up to 27 percent of the algae at BE-94.3 from July to September. At BE-54 green algae were insignificant with the exception of the low flow August 10 sample where they comprised 44 percent of the sample and were dominated by Ankistrodesmus, Scenedesmus, and Monoraphidium. The relative composition of the phytoplankton and the noted downstream increases in chlorophyll-a suggests that active growth of phytoplankton may be taking place over the reach from BE-94.3 to BE-54. Blue-green algae including Oscillatoria, Aphanocapsa, and Aphanizomenon were common at BE-54, while benthic diatoms were relatively uncommon. In contrast, no significant difference in chlorophyll-a was noted between the two most downstream sites (BE-54 and BE-18) in 1999 nor was there a significant difference between all three sites for

37

2000 (Table 3). Blue-greens comprised from 37 to 50 percent of the sample at BE-94.3 with Oscillatoria, Aphanocapsa, and “non-motile” blue-greens being the most common in July and August and Aphanizomenon becoming prominent in September. A similar pattern was noted at BE-54 with the blue-greens comprising from 47 to 81 percent of the algae samples. The presence of the slow growing Oscillatoria suggests long residence time for this reach (Reynolds, 1991). Phytoplankton samples at the most upstream site, BE-94.3, from 2000 exhibited similar dominant species as with the downstream sites, e.g., Oscillatoria and Cyclotella. In addition a huge blue-green bloom comprised primarily of Anabaena and Aphanizomenon was noted as well. Phytoplankton at BE-73.2 was dominated by a blue-green - Oscillatoria and diatom -Cyclotella. At BE-54 phytoplankton composition was similar to 1999 with Oscillatoria, Cyclotella , Nitzschia and Melosira among the dominants. Periphyton samples were taken at two sites: BE-73.2 and BE-54 in 2000. At BE-73.2 about seven genera, mostly diatoms, comprised 80% of periphyton samples. Nitzschia, Navicula and Stephanodiscus were the most common diatom genera. Euglena and non-motile blue-greens were the other two common forms. At BE-54 periphyton was dominated by four genera primarily and included the diatoms - Cyclotella and Nitzschia and the blue-green Oscillatoria and non-motile blue-greens. Oscillatoria, Cyclotella, and Cryptomonas dominated the phytoplankton in 2000. Fish IBI and habitat assessments were conducted at five sites on the Blue Earth River. The Blue Earth River had a considerable amount of fine substrates (silt, sand) with a moderate to normal amount of silt (Appendix II, Table 2). Most sites had heavy to severe bank erosion and minimal instream cover for fish. Riffles were present at two sites, BE-73 and BE-54; however, only BE-73 contained coarse substrates (cobble and gravel). The land use within the floodplain of the Blue Earth River is dominated by row crop agriculture. An undisturbed riparian area 5 to 50 meters in width was present on each bank along the sampling locations. Riparian habitat and instream features remained fairly constant throughout the length of the Blue Earth River with QHEI scores for the five sites ranging from 42 to 52. BE-73 exhibited the highest score, while both the most upstream (BE-100) and downstream (BE-18) sites exhibited the lowest scores. Thirty species of fish were collected from the Blue Earth River. The IBI rating varied from very poor to good. In sites BE-100, BE-73 and BE-54 species common in degraded environments (tolerant species) comprised over two-thirds of the collected fish community (Appendix II, Table 3). These sites also had a high percentage of omnivore species. Omnivorous species are those that have the physiological ability to digest both plants and animals (Karr et al. 1986). The ability to utilize multiple food sources allows the omnivore species to switch to another food source when one type of food is disrupted; dominance by omnivorous species indicates an unstable food base. At the most downstream site (BE-18) tolerant species comprised only seven percent of the total fish community. The number of intolerant, piscivore, and invertivore species at this site were higher than in other sites. The fish community at this site was rated as good, while other sites were designated as poor or very poor.

38

Red River of the North

The Red River was added to this study in 2000 as a part of a continued effort to understand relationships in nutrients, chlorophyll-a and BOD in large rivers across Minnesota’s varied ecoregions. Four Red River stations ranging from RE-536 at Brushvale to RE-298 at Grand Forks were monitored during the summer of 2000. All four sites had USGS flow gages at or near the site that allow for description of flow regime (Table 1). The most downstream site, RE-298, is immediately above the confluence with the Red Lake River while the USGS site is immediately below the Red Lake River. Flows were adjusted accordingly (Red Lake River subtracted) when paired with data from RE-298 and as reported in Table 1. While the Red River lies totally within the Red River ecoregion, there are substantial portions (tributaries) that drain from the adjacent NCHF and NMW ecoregions to the east in Minnesota and NGP ecoregion to the west in the Dakotas. Runoff from these various subwatersheds will influence the overall quality of the Red River in addition to runoff that arises from within the ecoregion.

Over the course of the summer, discharge at the two upstream stations decreased slightly, and discharge at the two downstream stations showed two distinct peaks at the end of June and then leveled off to flows slightly lower than flows recorded at the beginning of summer (Fig. 8). TP varied inversely with flow at the four sites (Fig. 8). During a high flow event in late June, TP concentrations were high at all sites and exceeded 1 mg/L at RE-403. Summer-mean TP and nitrate-N concentrations were on the order of 200 – 300 µg/L at the two upstream sites but increased to 500 – 600 µg/L at the two downstream sites (Table 3). Between-site differences in TP declined somewhat as river flow declined (Fig. 8a.). TSS concentrations increased as well in a downstream fashion with summer-mean concentrations of 55 mg/L at RE-536 and peaking at 374 mg/L at RE-403 (Table 3). Of these amounts, inorganic suspended sediment typically comprised 85-90 percent of the TSS.

Total chlorophyll-a concentrations did not increase consistently from upstream to downstream as mean concentrations peaked at 38 µg/L at RE-452 and RE-403 and actually declined to 26 µg/L at RE-298. At RE-403 and RE-298 summer-mean pheophytin exceeded chlorophyll-a, which was not found elsewhere (Table 3). Chlorophyll-a did not vary consistently with flow though there was a general tendency toward declining chlorophyll with declining flow at the three upstream sites (Fig. 8a and b). At RE-298, there was a mixed response to flow. In general total chlorophyll-a decreased over the course of the summer consistent with declining flow. On average the chlorophyll-a:TP ratio was similar between the two upstream sites but was much lower at the two downstream sites (Fig. 8a). Summer-mean pH was somewhat low as compared to the other rivers and tended to decrease from upstream (RE-536) to downstream (RE-298) (Table 3). This is likely related to the low chlorophyll-a (Table 3) and overall low algal activity (e.g., Fig. 4) at the downstream sites. Transparency generally decreased as discharge increased. This was likely due to particulates brought into suspension as a result of higher discharges since TSS was also positively correlated with discharge. The relationship between BOD5 and discharge was similar to that of chlorophyll-a and discharge. The periphyton at the two Red River sites was dominated by diatoms. Many of the dominant forms were common to other sites in this study including Cocconeis, Cymbella, Navicula,

39

Rhoicosphenia and Nitzschia. However, there were forms such as Gyrosigma that were quite uncommon elsewhere. Blue-green algae, principally non-motile blue-greens and Oscillatoria, were among the dominant at both Red River sites. One algal form, Batrachospermum vagum (a filamentous red alga), which comprised 31 percent by biovolume on the rock sample at RE-536, was not found elsewhere in the study. No phytoplankton samples were available for comparison. Figure 8. Red River (a) TP, (b) chlorophyll-a, (c) Chla:TP and (d) chlorophyll-a and flow comparisons among sites. (a) (b) (c)

TP over time, by site

0.00

0.40

0.80

1.20

1.60

3-Ju

n

17-J

un

1-Ju

l

15-J

ul

29-J

ul

12-A

ug

26-A

ug

9-Se

p

23-S

ep

Time

TP (m

g/L)

298 403 452 536

Red River 2000 Chl-T by site

0.0

20.0

40.0

60.0

80.0

100.0

120.0

3-Ju

n

17-J

un

1-Ju

l

15-J

ul

29-J

ul

12-A

ug

26-A

ug

9-Se

p

23-S

ep

Time

Chl

-T (u

g/L)

298 403 452 536

Chl a:TP

0.000.020.040.060.080.100.12

RE-536 RE-452 RE-403 RE-298

Chl

a:T

P

40

RE-536 Chlorophyll-a & Flow

0

4 0 0

8 0 0

1, 2 0 0

1, 6 0 0

Da t e

0

5

10

15

2 0

2 5

3 0

Discharge Chl a

RE-403 Chlorophyl-a & flow

0

10,000

20,000

30,000

40,000

13-A

pr

30-J

un

15-S

ep

1-D

ec

Date

Dis

char

ge, c

fs

0

5

10

15

20

25

30

Chl

a (u

g/L)

Discharge Chl a

RE-452 Chlorophyll-a & flow

0

1,000

2,000

3,000

4,000

5,000

6,000

13-A

pr

30-J

un

15-S

ep

1-D

ec

Date

Dis

char

ge, c

fs

0102030405060

Chl

a (u

g/L)

Discharge Chl a

RE-298 Chlorophyll-a & flow

0

10,000

20,000

30,000

40,00013

-Apr

30-J

un

15-S

ep

1-D

ec

Date

Dis

char

ge, c

fs

0

5

10

15

20

25

30

Chl

a (u

g/L)

Discharge Chl a

d)

41

III. DISCUSSION

A. Water Quality and Aquatic Community Patterns and Relationships Among Streams: 1999-2001

This section of the report will address relationships in various parameters across the entire set of rivers and sites. It will include not only those relationships among nutrients and chlorophyll-a but also linkages among other biotic and physical data. Examples of how these relationships might be used to develop ecoregional nutrient criteria for rivers will follow. Relationships among nutrients, chlorophyll-a and BOD5 Using regression analysis we were able to demonstrate significant (F-test < 0.001), consistent and positive relationships between total phosphorus (TP) and sestonic total chlorophyll-a (ChlT) in Minnesota rivers (Fig. 9; Table 4, Eq 3 & 4) based on sites monitored in 1999 and 2000 (excludes the Red River). The slope for the 2000 data appeared steeper, as compared to the 1999 regression, but was not significantly different (95% confidence level). The significant relationship between TP and ChlT is consistent with a worldwide study conducted by Van Nieuwenhuyse and Jones (1996) and a Canadian study by Basu and Pick (1996). In each of these studies significant linear regressions (log-log) of TP and total chlorophyll-a exhibited significant R2 values of 0.72 and 0.76 respectively. In our relationships, sites from the different ecoregions fall into distinct regions of the regression (Fig. 9). The predominately NLF ecoregion sites (Crow Wing and UM-1065, 1029, and 1004) were clustered near the lower end and transitioned to the NCHF sites. The Blue Earth (WCBP) and Crow River (WCBP-NCHF transition) sites were on the upper end. Variability in the relationship was evident for the Blue Earth sites in both years with the downstream (higher order) sites above the regression line and the lower order (BE-100 and BE-94) below the line. Table 4. Regression equations derived based on river-nutrient study. Regressions based on summer-mean concentrations for sites sampled in 1999 and 2000 unless otherwise noted. Red River data are not included in these regressions. Year Regression Equation R2 Equ # 1999: ChlT (µg.L-1) = 0.34 TP (µg.L-1) – 13.2 (R2=0.89) (3) 2000: ChlT (µg.L-1) = 0.44 TP (µg.L-1) – 20.0 (R2=0.91) (4) 1999: ChlT (µg.L-1) = 7.98 TN (mg.L-1) + 11.45 (R2= 0.49) (5) 2000: ChlT (µg.L-1) = 8.17 TN (mg.L-1) + 21.86 (R2= 0.25) (6) 1999: ChlT (µg.L-1) = 75.53 TKN (mg.L-1) – 48.39 (R2 = 0.94) (7) 2000: ChlT (µg.L-1) = 100.84 TKN (mg.L-1) – 61.39 (R2 = 0.96) (8) 1999: TP (µg.L-1) = 231.23 TKN (mg.L-1) – 105.32 (R2 = 0.95) (9) 2000: TP (µg.L-1) = 188.67 TKN (mg.L-1) – 63.35 (R2 = 0.95) (10) 1999: BOD5 (mg.L-1) = 0.031 ChlT (µg.L-1) + 0.89 (R2 = 0.97) (11) 2000: BOD5 (mg.L-1) = 0.043 ChlT (µg.L-1) + 0.88 (R2 = 0.95) (12) 1999: BOD5 (mg.L-1) = 0.011 TP (µg.L-1) + 0.42 (R2 = 0.92) (13) 2000: BOD5 (mg.L-1) = 0.019 TP (µg.L-1) - 0.14 (R2 = 0.87) (14)

42

1999: BOD5= 0.043 Chl-a + 0.95 (R2=0.98) (15) 2000: BOD5= 0.053 Chl-a + 0.87 (R2=0.98) (16) 1999: BOD5= 0.18 Pheo +0.48 (R2=0.80) (17) 2000: BOD5= 0.26 Pheo +0.64 (R2=0.80) (18) 1999: Turbidity (NTU) = 0.50 TSS (mg.L-1) + 3.63 (R2=0.96) (19) Likewise we found a positive, but weaker, relationship between ChlT and total nitrogen (TN) (Table 4 Eq 5 & 6). The consistently high nitrate-N concentrations in the Crow and Blue Earth Rivers (Table 3) contributed to the poor relationship. TKN exhibited much stronger and significant relationships (Fig. 10, F-test, p< 0.001) with total chlorophyll in both years (Fig. 3; Table 4, Eq. 7 & 8). This was anticipated because TKN is largely organically bound nitrogen and much of the measured N is likely tied up in the algal cells (similar to TP). TKN is highly correlated with TP as well in both years (Table 4, Eq. 9 & 10). Basu and Pick (1996) noted a significant relationship between log TN and log total chlorophyll-a in their streams (R2 = 0.66). They noted, however, that it is unlikely that TN concentration regulated chlorophyll in their study rivers. The four reaches of the Red River that were monitored during the summer of 2000 differ from the other rivers in that light appears to limit primary productivity in the Red River as opposed to phosphorus. Whereas summer- mean TP was a significant predictor of mean chlorophyll-a when data from the other five rivers are combined (P < 0.001), it is not a significant predictor when the Red River data are included in the regression analysis (P = 0.18). Average TP in the Red River in 2000 was 400 µg/L (Table 3), which was higher than the individual station means from the other rivers (Table 3). However, the higher phosphorus concentrations in the Red River did not translate into higher productivity. There was a trend of increasing TP in the downstream stations, yet the chlorophyll-a:TP ratio decreased in these reaches (Fig. 8 a,b,c). In contrast, in the Blue Earth and Mississippi rivers, there was a trend of higher total chlorophyll-a:TP with increasing watershed size (Table 3, Fig. 11). Unless specifically noted, the discussion below does not include the Red River information.

Watershed area, flow and residence time are highly related (Table 1) and contribute to between-year and between-site differences in water quality in these rivers. This finding is consistent with Reynolds (2000), who notes that the biomass of river phytoplankton is proportional to the residence time (age) of the water it is in. While this holds in general, it does not apply to all sites in this study. In the nutrient-poor Crow Wing, TP concentrations were consistently higher at the downstream site in 1999 and 2000 (Table 3), however there was no significant difference in ChlT between these sites in either year (Fig. 11). Overall, Crow Wing ChlT concentrations were quite low and varied positively with flow (Fig. 6). Here the productivity tends to be in the periphyton rather than the phytoplankton. In contrast, the nutrient-rich Blue Earth River exhibited a downstream increase in ChlT that peaked near BE-73 or BE-54 and then declined at BE-18. The largest between-site increases were noted among the third order site BE-100, BE-94 and the three downstream fourth order sites (Table 3). Increased residence time (watershed area) and TP (Tables 1 and 3) between BE-100 and BE-94 contributed to the increased ChlT between these two sites, while increased residence time (and not increased TP) between BE-94 and BE-73 was likely the reason for the downstream ChlT increase in both years (Fig. 11). However, ChlT was stable or declining over the last two sites. There was no appreciable change in TP or watershed area between BE-73 and BE-18 and other factors are likely influencing ChlT. The

43

Crow did not exhibit a downstream increase in ChlT in either year however, as with the Blue Earth River, it exhibited peak chlorophyll during periods of low or stable flow (Fig. 7). Also, as noted previously the downstream site (CR-0.2) may be overly influenced (diluted) by the Mississippi River. The Rum River, intermediate between these extremes, exhibited a downstream increase in ChlT but did not exhibit a consistent relationship between chlorophyll and flow (Fig. 6). The fifth and sixth order Mississippi River sites exhibited minimal change in ChlT relative to watershed area in both years (Fig. 11). Stable TP concentrations in 1999 (Table 3) and short time-of-travel (about three to four days) between UM-1004 and UM-895 contribute to this lack of change in ChlT. TP concentrations again were relatively stable over this reach (Table 3) in 2000 but time-of-travel was on the order of seven to eight days and a gradual increase in ChlT from the upstream sites to UM-895 was noted. In both years, however, a marked increase in TP and ChlT was noted between UM-895 and UM-872. A lack of a relationship between ChlT and watershed area for the high order Mississippi sites was consistent with Basu and Pick (1996) who noted that the biomass of phytoplankton in large (greater than fifth order) rivers was strongly correlated with TP concentrations but not with water residence time.

44

Figure 9. Summer-mean total phosphorus (TP) and chlorophyll-a (Chl-T) for (a) 1999 and 2000 sites and (b) including Red river sites.

0

25

50

75

100

125

150

0 50 100 150 200 250 300 350 400TP ppb

Chl

T pp

b

1999 2000 Linear (2000) Linear (1999)

0

25

50

75

100

125

150

0 100 200 300 400 500 600 700

TP ppb

Chl

T pp

b

CW,MI,RU,CR,BE Red Linear (CW,MI,RU,CR,BE)

(a)

(b)

45

Figure 10. Summer-mean total Kjeldahl nitrogen vs. total chlorophyll-a for sites sampled in 1999 and 2000 (excludes Red River).

A second primary goal of this study was to establish the relationship between chlorophyll-a and BOD5. ChlT and BOD5 exhibited strong and significant (F-test, p <0.001) relationships in the rivers studied (Fig. 12 and Table 4 Eq 11 & 12). The slopes for the 2000 data were not significantly different. As with TP, the NLF sites exhibited low chlorophyll and BOD5, while values were much higher for both parameters in the Blue Earth and Crow Rivers (Table 3, Fig. 12). The transitional NCHF sites were intermediate between these two extremes. These between-year comparisons suggested that the combination of less dilution, increased algal growth (respiration), and algal decomposition played a more important role in the production of BOD during the lower flows of 2000 as compared to the higher flows of 1999. As anticipated based on Figs. 10 and 12, the relationship between TP and BOD5 was equally strong and significant (F-test <0.001) for both years (Fig. 13) (Table 4, Eq. 13 & 14). The slope for the 2000 data was steeper and significantly different (95% confidence interval) than the slope for 1999.

0

25

50

75

100

125

150

0.00 0.50 1.00 1.50 2.00 2.50TKN ppm

Chl

T pp

b

2000 1999 Linear (2000) Linear (1999)

46

Figure 11. Summer mean total chlorophyll-a and chl-a: TP versus watershed area for sites sampled in 1999 and 2000.

1999

0

25

50

75

100

125

150

0 10,000 20,000 30,000 40,000 50,000

Watershed km 2

Chl

T pp

b

CWR UM RUM CR BE

2000

0

25

50

75

100

125

150

0 10,000 20,000 30,000 40,000 50,000

Watershed km2

Chl

T pp

b

CWR UM RUM CR BE

(a)

(b)

47

Figure 12. Summer-mean total chlorophyll-a and BOD5 for sites sampled in 1999 and 2000.

Figure 13. Summer-mean total phosphorus vs. BOD5 for sites sampled in 1999 and 2000.

0.01.02.03.04.05.06.07.08.0

0 25 50 75 100 125 150

ChlT ppb

BO

D 5 p

pm

2000 1999 Linear (2000) Linear (1999)

0.01.0

2.03.0

4.05.0

6.07.0

8.0

0 50 100 150 200 250 300 350 400

TP ppb

BO

D 5 ppm

1999 2000 Linear (2000) Linear (1999)

48

Chlorophyll-a (living cells) and pheophytin (degradation product) data from individual sites and dates can further describe the influence of algae on BOD. The living cells (at the sample site) contribute to respiration, while the dead cells transported from upstream reaches contribute via bacterial decomposition. Individual regression of mean chlorophyll-a and BOD5 (Table 4, Eq. 15 & 16) and pheophytin-a and BOD5 (Table 4, Eq. 17 & 18) showed significant relationships (F-test, p <0.01) for each. Chlorophyll-a, however, explained a higher percentage of the variation in BOD5 in both years. Viewing individual data can further reinforce the relationships. A comparison of chlorophyll-a and pheophytin-a to BOD5 at CR-23 and BE-18 showed good correspondence if the two values are combined (i.e., total chlorophyll) with the exception of June 12, 2000 (Crow) and June 28 (BE-18) when BOD5 was high relative to the chlorophyll-a (Fig. 14). The June 12th sample was taken at the peak of the hydrograph (Fig. 6). High runoff brought organic and inorganic material into suspension, as evidenced by the elevated TSS (120 mg.L-1) on that date, which would contribute to the BOD load in the river (in addition to algal respiration and decomposition). All other samples were taken on the declining limb of the hydrograph or during low flow (Fig. 6) when TSS was about 60-70 mg.L-1. For BE-18, the one date with poor correspondence was June 28, 2000 when BOD5 was elevated relative to chlorophyll-a. Again, as with the Crow River, increased flow (Fig. 6) and peak TSS (270 mg.L-1) were noted on this date. In contrast, peak BOD5 and chlorophyll-a co-occurred on August 11, 2000, under low flow condition (Fig. 6). TSS was 86 mg.L-1 on that date which is a very low value as compared to the peak concentration of 270 mg.L-1. In each of these two examples there was very good correspondence between BOD5 and total chlorophyll based on individual sample dates, which further supported the excellent correlations previously noted. These data (BE-18 and CR-23) also suggested that flow and TSS periodically affect this relationship as well.

49

Figure 14. BOD5, chlorophyll-a and pheophytin comparisons for Crow and Blue Earth River sites.

Crow River (CR-23) 2000

0255075

100125150175200225

12-Jun 29-Jun 19-Jul 2-Aug 17-Aug 5-Sep 13-Sep 25-Sep

Chl

-a/P

heo

ppb

0

2

4

6

8

10

12

BO

D 5 p

pm

Chl-a Pheo BOD5

Blue Earth (BE-18) 2000

0255075

100125150175200225

7-Jun 28-Jun 18-Jul 11-Aug 16-Aug 30-Aug 13-Sep 27-Sep

Chl

/Phe

o pp

b

0

2

4

6

8

10

12

BO

D 5 p

pm

Chl-a Pheo BOD5

50

Monitoring at additional river sites in 2001 provided an opportunity to “validate” or check on the performance of the various predictive models (Table 4). For this purpose several medium to high order stream sites in the Minnesota, St. Louis, and Red River basins were sampled on 6-8 occasions in 2001. Summer-mean TP, chlorophyll-a and BOD5 from these sites were compared to data from our core sites used in this study and the regression equations that were developed. Summer-mean data for these sites are included in Appendix III. In general, many of the rivers monitored in 2001 fall on or near the regression line based on the 1999/2000 data (Fig. 15) and includes sites from the St. Louis, Otter Tail, Cottonwood, Minnesota, Redwood, and East Des Moines Rivers. There were, however several rivers that produced less chlorophyll-a per unit TP. Of these the Straight, Mustinka, and Watonwan are small streams (watersheds < 1,000 mi2) as compared to the original data set. Others, like the Wild Rice and Buffalo, in the Red River basin, had rather high TSS with summer-means of 94 and 112 mg/L respectively. The one site with extremely high chlorophyll-a (175 µg/L, Fig. 15) relative to its TP concentration was on the West Des Moines River. This site had a small but highly eutrophic reservoir upstream of it – which may have contributed to the high chlorophyll-a concentration (exported algal downstream). Though there was some variability in the relationship between TP and chlorophyll-a for the 2001 data, as compared to the original data set, the relationship between chlorophyll-a and BOD5 was quite consistent with the original equations (Fig. 16). This suggests the equations (Eq. 11 & 12, Table 4) should be applicable to a broad range of streams. Figure 15. Summer-mean total phosphorus vs total chlorophyll-a for core (1999 & 2000) sites compared to new 2001 sites.

Summer-mean TP vs Chl-T: 1999, 2000, 2001

0

25

50

75

100

125

150

175

200

0 50 100 150 200 250 300 350 400 450 500

TP ppb

Chl

-T p

pb

99/00 2001 Linear (99/00)

low orderhigh TSS

51

Figure 16. Summer-mean total chlorophyll-a versus BOD5 for the (1999&2000) sites (Table 1) compared to new 2001 sites. Algal Community Composition A third primary goal of this study was to relate algal community composition to nutrient status and watershed size for medium to large Minnesota rivers. The dominant forms of algae generally included diatoms, blue-greens, and greens – based on measures of abundance (units/mL) and biovolume (µm3/cm2). Other forms, such as Cryptomonads and Chrysophytes, were represented to a much lesser degree. Phytoplankton (seston) composition varied not only in terms of the origin of the seston, i.e., benthic vs. sestonic, but also along a gradient of nutrient enrichment (Figs.17a & b). Diatoms were a significant proportion (by abundance and biovolume) in the Crow Wing River but tended to decline in significance in the larger (e.g., Mississippi) or more nutrient-rich rivers (e.g., Rum). In the highly nutrient-rich Crow and Blue Earth Rivers blue-green algae were the dominant forms of algae in 1999. In general total biovolume tended to increase with increased TP concentration, though the results for the Crow River sites were not completely consistent with this general pattern. These patterns were not quite as pronounced in the 2000 data (Fig. 17). Based on abundance, miscellaneous genera (e.g., Chrysophytes) were more prominent in 2000 at several sites. As a result, the relative abundance of diatoms and greens were reduced at the nutrient-poor sites (e.g. Crow Wing) and blue-greens were not as abundant at nutrient-rich sites (e.g. Crow). Biovolume-based comparisons, however, suggested that diatoms were prominent at several of the sites and that the miscellaneous genera were not overly abundant. Also, as with chlorophyll-a, total biovolume at most sites was higher in 2000 as compared to 1999. Blue-green algae figured

Summer-mean ChlT vs. BOD5: 1999, 2000, 2001

0.0

2.0

4.0

6.0

8.0

10.0

0 50 100 150 200CHl-T ppb

BO

D5

ppm

99/00 2001 Linear (99/00)

WDM-3

52

Figure 17. Mean phytoplankton composition based on abundance (a) and biovolume (b) for selected sites in 1999 and 2000. Based on three samples per summer. Sites sorted by summer-mean TP. Worldwide percentage drawn from Rojo et al. (1994). (a)

Mean algal abundance (units/mL) for 1999. Sorted by mean TP

0%

20%

40%

60%

80%

100%

CWR-35(59)

UM-953(62)

UM-1004(71)

UM-872(92)

RU-18(131)

RU-34(137)

BE-94(247)

BE-54(248)

CR-0(329)

CR-23(359)

World%

% c

ompo

sitio

n

Diatoms Greens BlueGr Crypto Misc

Mean Algal Abundance (units/mL) for 2000. Sorted by TP.

0%

20%

40%

60%

80%

100%

CWR-35 (59)

UM-1056(59)

UM-953(62)

UM-872(92)

RU-18(131)

RU-34(137)

BE-73(205)

BE-94(247)

BE-54(248)

CR-0(329)

CR-23(359)

% c

ompo

sitio

n

Diatoms Greens BlueGr Crypto Misc

53

(b)

% Biovolume for 2000. Sorted by TP

0%

20%

40%

60%

80%

100%

CWR-35.5 UM-953 UM-1056 UM-872 RU-18 RU-34 BE-94.3 BE-54 CR-0.2 CR-23 World%

% B

iovo

lum

e

Diatom Family Greens BGA Cryptos Misc

% Biovolume for 1999. Sorted by TP

0%

20%

40%

60%

80%

100%

CWR-35.5 UM-956.2 UM-1004 UM-872 RU-18 RU-34 BE-94.3 BE-54 CR-0.2 CR-23 World%

% B

iovo

lum

e

Diatom Family Greens BGA Cryptos Misc

54

prominently in the increased biovolume (1999 vs. 2000) for UM-953, RU-18, RU-34, BE-94.3, and BE-54. Somewhat surprisingly though blue-greens were not a significant form at the Crow River sites in 2000. Again there is a general relationship between TP concentration and total biovolume across the various sites, with the exception of CR-23. UM-872 exhibited the largest increase (four-fold increase) in biovolume between 1999 and 2000. Rojo et al. (1994), in their assessment of algal communities in temperate versus tropical streams, provide one basis for describing the average percent composition of algae (based on biovolume) in temperate rivers (Fig. 17 a & c). Based on their summary a “typical” temperate river would exhibit a fairly balanced algal community where diatoms and greens would be the most prominent forms while blue-greens would be a minor component. In general, the Crow Wing, Rum, and Upper Mississippi sites: (UM-1004 and UM-953) exhibited compositions somewhat similar to a “typical” composition for temperate rivers based on Rojo et al. (1994). Common diatoms at these sites included less pollution tolerant genera such as Cocconeis and Acananthus (Palmer 1969). Prygiel and Leiato (1994) noted these genera as well in the headwater stream to the reservoir Val Joly in northern France. In the shallow and clear Crow Wing River, periphyton scoured from the substrate of the stream was an important component of the measured seston based on the predominance of benthic diatoms at CWR-35 in 1999 (Fig. 7) and comparisons between periphyton and phytoplankton samples in 2000. Lohman and Jones (1999) observed this in their work on Missouri third and fourth order streams and Descy et al. (1987) described this as well. In the Rum River, UM-953, and UM-872 the more pollution tolerant (Palmer 1969) diatoms, such as Nitzschia, Melosira, and Cyclotella were quite common. Cyclotella and Stephanodiscus, which were particularly abundant at UM-872 in 1999, are considered planktonic rather than benthic diatoms based on Kelly (1998). The TP optima is high for several species within these genera, e.g., Nitzschia ranges from 35 – 260 µg/L, Cyclotella range from 70 – 400 µg/L and Stephanodiscus range from 200 – 300 µg/L. The presence of the Cryptomonads, Rhodomonas minutas and Cryptomonas, further attest to the nutrient richness of these sites based on Swale’s (1969) observations in the nutrient-rich Stour River in England. The most common green algae in the Rum, UM-953, and UM-872 were Ankistrodesmus and Scenedesmus, which Swale (1969) noted as the two most common green algae genera in the nutrient-rich Severn River. The blue-green, Lyngbya, was prominent in the Rum and UM-953, while Aphanizomenon and Oscillatoria were common at UM-872 in 2000. The combination of higher TP and longer residence time favor the production of planktonic algae in this reach (UM-872) of the Mississippi River. This would be consistent with other studies on downstream reaches (e.g., Baker and Baker, 1979). Though there were a few dominants in common (e.g. Rum sites), the periphyton dominants were distinctly different from the phytoplankton at these sites. In the highly nutrient-rich Crow and Blue Earth Rivers pollution-tolerant blue-green algae (including Oscillatoria, Aphanizomenon, and non-motile blue-greens) were dominant in 1999. Prygiel and Leitao (1994) noted Oscillatoria blooms in the highly nutrient-rich river below the reservoir Val Joly as well. They note that while the reservoir “seeded” the Oscillatoria and other blue-greens, the river sustained them and allowed for bloom concentrations. While blue-greens remained dominant at the Blue Earth sites in 2000, they were a minor portion of the phytoplankton at the Crow River sites. Rather, diatoms like the planktonic Aulocoseira,

55

Cyclotella, and Stephanodiscus were quite abundant along with the greens Ankistrodesmus and Scenedesmus. In general, for the Crow and Blue Earth Rivers chlorophyll-a increased with declining flow (Fig. 7) suggesting active growth of phytoplankton in the water column, rather than a dependence on scoured periphyton as the source of the measured sestonic chlorophyll-a. This is further supported by the prominence of planktonic diatoms, greens, and slow-growing Oscillatoria in the algae samples. To further understand the relationship between phytoplankton and periphyton for these rivers comparisons were made between periphyton communities on rock and wood substrate vs. phytoplankton. Based on Fig. 18 there was no significant difference between rock or wood substrates for all three algal types. Diatoms exhibited a distinct preference for substrates, while greens were more abundant in the water samples. Blue-greens exhibited no distinct preference and generally comprised a small percentage of the overall samples from all media. Figure 18. Comparison of algal composition of periphyton rock and wood samples and phytoplankton samples for summer 2000.

rock water woodMEDIA

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

PC

TDIA

TOM

S


0.0

0.1

0.2

0.3

0.4

0.5

0.6

PC

TGR

EE

NS


0.0

0.1

0.2

0.3

0.4

PC

TBG

A

56

Relationships among total suspended solids, turbidity, and transparency

Strong relationships between TSS, turbidity, and transparency tube measurements have previously been demonstrated (Sovell, 2000). This study provided an opportunity to review relationships among TP, total suspended solids, turbidity, chlorophyll-a, and transparency tube measurements across several rivers and ecoregions. Figure 19. Summer-mean total suspended solids vs. turbidity Summer-mean turbidity and TSS are highly correlated (Fig. 19; “outlier” values are for RE-403 and RE-298). In general transparency declines as turbidity, TSS and TP increase (Fig. 20a, b, &d). These relationships are strong for transparency vs. turbidity (R2 = 0.95) and transparency vs. TSS (R2 = 0.94); but somewhat weaker for transparency vs. TP (R2=0.72). As transparency declines from 60 to 20 cm there is a rather sharp (linear) increase in turbidity and TP. Based on these data, a 20-cm transparency-tube reading is about equivalent to the water quality standard for turbidity in most Minnesota streams of 25 Nephalometric Turbidity Units (NTUs). For TSS, the sharp increase is much more pronounced and does not occur until transparency reaches about 10-15 cm (Fig. 20a). Although flows were much higher in 1999 than 2000, differences in transparency and its relationship to other water quality variables were minimal. Some patterns are evident when the transparency and chlorophyll-a tube data are compared. Chlorophyll-a levels are low at transparencies below 10 cm, presumably because of light limitation, as is the case for the Red River sites (Fig. 20e). Chlorophyll-a peaks between 10 and 20-cm transparency – a range over which TP is relatively high (Fig. 20e) and TSS and inorganic turbidity (Fig. 20a & b) does not severely limit light. Chlorophyll-a declines again at around 25-cm transparency, where TP is often low and may be limiting algal productivity and/or the algal activity is dominated by benthic algae that do not affect water transparency.

TSS vs. Turbidity for 1999, 2000, & 2001

y = 0.50x + 3.63R2 = 0.96

0

20

40

6080

100

120

140

160

0 50 100 150 200 250 300 350 400

TSS mg/L

Turb

. NTU

1999 2000 2001 Linear (1999)

57

The various rivers/ecoregions array themselves along various positions in the tranparency tube relationships (Fig. 20c). NLF sites, such as the Crow Wing and Upper Mississippi, are characterized by low TSS and turbidity and high transparency and exhibit a linear relationship. NCHF rivers/sites such as the Rum are intermediate. The transitional NCHF-WCBP sites (e.g., Crow) and WCBP sites (e.g., Blue Earth) are characterized by high TSS and turbidity and very low transparency measurements. At this end of the scale small changes in transparency correspond to larger changes in TSS and turbidity. Lastly, the highly turbid Red River sites occupy the low transparency portion of the graph and the transparency range overlaps with the Blue Earth, yet extends even lower. When turbidity measurements for the Red River are plotted against transparency tube readings, the best fit is an exponential curve, with an R2 of 0.93 (Fig. 20e), as opposed to a linear relationship with an R2 of only 0.85. These data appear to fit in well with other similar data from Minnesota streams, as published in the MPCA’s 1999 Report of the Water Quality of Minnesota streams (Sovell, 2000). As transparency decreases, turbidity increases exponentially. When data from individual streams are plotted separately, a linear fit is appropriate (Blue Earth and Mississippi in the 1999 report). However, the ranges in turbidity from these two rivers barely overlap one another; hence the data points occupy different portions of the curve that contains all of the data. Given these relationships the transparency data should provide a good basis for estimating the other variables. However, the individual relationships (curvilinear vs. linear) and variability in the relationship, at different positions on the transparency scale, should be considered in these predictions.

Relationships among nutrient enrichment, dissolved oxygen and fish IBI The periphyton and diurnal DO studies of the USGS and the MPCA fish IBI work, at selected sites in 2000, provide an opportunity to look further at linkages among the various biological and chemical aspects of the rivers. Daily-mean DO flux (based on 4 days of measurement) varied positively with TP and seston chlorophyll-a with R2 values of 0.55 and 0.60 respectively based on paired data from the Crow Wing, Mississippi, Rum, Crow, and Blue Earth River sites (Fig. 21). The Red River sites deviated substantially from this relationship. An even stronger relationship was noted for TP and community (phyto and periphyton) respiration (R2=0.78, Fig. 22). Likewise seston chlorophyll-a and community respiration exhibited a strong relationship (R2=0.76, Fig 22). In each of these comparisons the correlation would be strengthened if values for UM-1056 were omitted. As discussed previously the production of chlorophyll-a per unit TP at this site was quite low relative to the other river-sites in this study. Likewise DO and DO flux (Fig. 20) were quite low as well. Similar comparisons were made with DO production, community respiration, and periphyton chlorophyll-a. In contrast to the strong correlations TP and phytoplankton chlorophyll-a exhibited (Fig. 21), periphyton chlorophyll-a exhibited no relationship with community DO production or respiration (Fig. 23, R2=0.06). Likewise, there was no correlation between phytoplankton chlorophyll-a and periphyton chlorophyll-a.

58

Figure 20. Summer-mean transparency tube measurements as compared to TSS, turbidity, TP and chlorophyll-a. (a) (b) (c)

T-tube vs. TSS for 1999, 2000, 2001

y = 6017x-1.5365

R2 = 0.9432

0

50

100

150

200

250

300

350

400

0.0 10.0 20.0 30.0 40.0 50.0 60.0

T-tube cm

TSS

mg/

L

1999 2000 2001 Power (2000)

T-tube vs turbidity for 1999, 2000 & 2001

y = 1972x-1.3801

R2 = 0.95

0

20

40

60

80

100

120

140

160

0.0 10.0 20.0 30.0 40.0 50.0 60.0

T-tube cm

Turb

NTU

1999 2000 2001 Pow er (2000)

Summer-mean T-tube vs Turbidity

R2 = 0.72

R2 = 0.95

020406080

100120140160

0.0 10.0 20.0 30.0 40.0 50.0 60.0

T-tube cm

Turb

NTU

UM, RU CR BE Red Linear (UM, RU) Power (BE)

59

(d) (e)

Summer-mean T-tube vs. Chl-T

0

40

80

120

160

0.0 10.0 20.0 30.0 40.0 50.0 60.0

T-tube cm

Chl-T

ppb

UM, RU CR BE Red

Summer-mean TP and T-tube

0100200300400500600700

0.0 10.0 20.0 30.0 40.0 50.0 60.0

T-tube cm

TP p

pb

UM, RU CR BE Red

60

Figure 21. Diurnal DO flux compared to TP and seston chlorophyll-a

TP vs. DO Flux

y = 0.012x + 2.43R2 = 0.47

01234

5678

0 50 100 150 200 250 300 350 400

TP ug/L

DO

Flu

x m

g/L

CW,UM,RU,CR,BE Red Linear (CW,UM,RU,CR,BE)

Daily mean DO Flux (4-6 days) vs. summer-mean chlorophyll-a

y = 0.03x + 2.8R2 = 0.60

0

1

2

34

5

6

7

8

0 20 40 60 80 100 120 140

Chl-a ug/L

DO

Flu

x m

g/L


61

Figure 22. Community DO production and respiration vs. TP and seston chlorophyll-a.

Summer-mean TP vs. Community respiration (96 hr mean)y = 0.0013x + 0.059

R2 = 0.78

0.00

0.10

0.20

0.30

0.40

0.50

0.60

0 50 100 150 200 250 300 350 400

TP ug/L

DO

gO

2/m

3/hr


Phytoplankton Chl-a vs. Community respirationy = 0.0029x + 0.1187

R2 = 0.76

0.00

0.10

0.20

0.30

0.40

0.50

0.0 50.0 100.0 150.0

Chl-a ug/L

DO

gO

2/m

3/hr


Phytoplankton Chl-a vs 02 Productiony = 0.0042x + 0.3352

R2 = 0.61

0.000.100.200.300.400.500.600.700.800.901.00

0.0 20.0 40.0 60.0 80.0 100.0 120.0 140.0

Chl-a ug/L

DO

gO

2/m

3/hr


62

Figure 23. Community DO production and respiration as compared to periphyton

chlorophyll-a.

Periphyton Chl-a vs DO Production

0.00

0.20

0.40

0.60

0.80

1.00

0.0 20.0 40.0 60.0 80.0 100.0 120.0 140.0 160.0

Peri chl-a ug/L

DO

g O

2/m

3/hr

CW,UM,RU,CR,BE Red

Periphyton Chl-a vs Community respiration

0.00

0.10

0.20

0.30

0.40

0.50

0.0 20.0 40.0 60.0 80.0 100.0 120.0 140.0 160.0

Chl-a ug/L

DO

gO

2/m

3/hr

CW,UM,RU,CR,BE Red

63

Fish IBI and macroinvertebrate EPT indices Forty-six species of fish were collected from sites in the Upper Mississippi River Basin and thirty species from the Blue Earth River (Appendix II). Biotic integrity ratings varied from very poor to excellent. The highest IBI scores were from the Crow Wing River and the upstream Mississippi River sites (Table 5). Macroinvertebrates were collected from ten sites as well (Table 5). In-stream habitat quality, as measured by QHEI (Appendix II), varied among the sites as did landuse in the watershed of the streams. For all sites, a strong relationship was noted between IBI scores, EPT values and the QHEI (Fig. 24). Habitat quality ranged from high values in the Crow Wing (QHEI = 79-80) to very low values in the Blue Earth River (QHEI = 42-52). Watershed landuse varied significantly over the range of sites with percent “disturbed” landuse (urban and agricultural) ranging from 34-49% for the Crow Wing sites to 95-96% for the Blue Earth sites. In general high IBI and EPT scores were associated with high QHEI and sites with lower percentages of disturbed landuse (Appendix II). Overall based on the six sites with both IBI and EPT measures excellent correspondence was found between the two biological measures (R2=0.80; Fig. 23). Correlations among these two biological measures and various water quality variables were explored based on data collected in summer 2000 (unless otherwise noted comparisons were made with summer-mean values). The correlation with TP was rather low but similar between the two indices (Fig 25a). The correlation between IBI and TN was quite high (R2=0.88) but non-existent for EPT (Fig. 25a). This high correlation was principally a function of the nitrate-N concentration, which exhibited an R2=0.89. Based on Fig. 24a the data from the Blue Earth sites (high nitrate-N and low IBI) figure prominently in defining this relationship. It would be desirable to collect data from sites intermediate between the Crow River (nitrate-N ~ 3 - 4 mg/L) and the Blue Earth (nitrate-N ~ 5.5 mg/L) to see if this relationship holds up over the entire range. The nutrient “response” variables chlorophyll-a and BOD5 exhibited higher correlation with IBI and EPT as compared to TP. For these two parameters correlations were higher for EPT as compared to IBI (Fig. 25b). Also the differences between the nutrient-rich Crow and Blue Earth sites were much less pronounced for the EPT as compared to the IBI, which was characterized by a large difference in IBI values for the two rivers. It was felt that the relatively high QHEI for the Crow accounted in part for the relatively high IBI score as compared to the Blue Earth that had a low QHEI. In both cases BOD5 exhibited a higher correlation with IB and EPT than did chlorophyll-a. Dissolved oxygen (DO) is a frequently measured parameter that is often used as one basis for determining whether waters support a fishery (e.g. 305(b) or 303(d) assessments). The DO standard for warmwater rivers such as these is 5 mg/L. Using data from the USGS diurnal study we were able to demonstrate relationships among TP, chlorophyll-a, and diurnal DO flux (Fig. 21). No relationship was evident among IBI and EPT and minimum DO (Fig 25c). This is most likely due to the small range in values and that the minimum value in all but one instance was 5.0 mg/L or greater. Maximum DO, a reflection of algal productivity, did however exhibit high correlation with IBI and EPT, with R2 of 0.65 and 0.56 respectively (Fig. 25c). Fish IBI values exhibited a significant correlation with diurnal DO flux with a R2=0.49 and an even higher R2

64

with EPT (R2=0.84). The correlation (R2) with fish IBI would be increased to 0.76 if the value from UM-1056, a site with poor habitat and where diurnal measurements were made following a storm event, is omitted. Suspended sediment can impact aquatic life directly by clogging gills, reducing light, or indirectly by burying habitat. Based on this limited data set high correlation was found among two surrogate measures of suspended sediment - TSS and turbidity and IBI and EPT (Fig. 25d). The highest correlation was between TSS and EPT (R2=0.94). Data such as these may be useful for evaluating the 25 NTU turbidity water quality standard. For example, based on this limited data set there did not appear to be a marked change in IBI or EPT over a range in turbidity values from about 5 - 40 NTUs. Additional data would be required to more completely define these relationships. Poor habitat, nutrient enrichment, and other water quality factors combine to yield a low IBI and EPT for the Blue Earth River. In contrast the Crow Wing had good habitat, low nutrients, low diurnal flux and a high IBI and EPT. The other sites were arrayed between these two extremes. The Crow while being extremely nutrient-rich exhibited a “good” IBI score largely because of good habitat (CR-23). As habitat quality declined at CR-0.2 a decline in IBI was noted (Table 5). Table 5. Fish IBI and macroinvertebrate EPT index values for 2000. Site CWR

-72 CWR

-35 UM

-1056 UM -953

UM -872

RU -34

RU -18

CR -23

CR -0.2

BE -100

BE -94

BE-73

BE-54

BE-18

IBI 80 86 68 69 69 63 65 55 40 25 21 EPT 21 14 19 14 14 22 16 15 11 10 QHEI 79 80 50 63 65 65 78 68 60 43 46 52 48 42

65

Figure 24. Fish IBI and invertebrate EPT compared to QHEI.

EPT vs. Fish IBI

R2 = 0.80

0

5

10

15

20

25

0 20 40 60 80 100

IBI

EPT

QHEI vs. EPT

R2 = 0.76

05

10152025

0 20 40 60 80 100QHEI

EPT

QHEI vs IBI

R2 = 0.87

0

20

40

60

80

100

0 20 40 60 80 100

QHEI

IBI

66

Figure 25. Fish IBI and invertebrate EPT compared to: a) nutrients b) chlorophyll-a and BOD, and c) dissolved oxygen, and d) turbidity.

TN vs. EPT

R2 = 0.075

0

5

10

15

20

25

0.0 2.0 4.0 6.0 8.0 10.0

TN mg/L

EPT

TN vs. IBI

R2 = 0.88

0

20

40

60

80

100

0.00 2.00 4.00 6.00 8.00

TN mg/L

IBI

TP vs EPT

R2 = 0.31

05

10152025

0 100 200 300 400

TP ug/L

EPT

TP vs. IBI

R2 = 0.25

0

20

40

60

80

100

0 100 200 300 400

TP ug/L

IBI

(a

NO3 vs. EPT

R2 = 0.02

0

5

10

15

20

25

0 2 4 6 8

NO3 mg/L

EPT

NO3 vs. IBI

R2 = 0.89

0

20

40

60

80

100

0.00 1.00 2.00 3.00 4.00 5.00 6.00

NO3-N mg/L

IBI

67

Chl-a vs EPTR2 = 0.50

05

10152025

0 50 100 150Chl-a ug/L

EPT

BOD vs. EPT

R2 = 0.65

0

510

1520

25

0.0 2.0 4.0 6.0 8.0

BOD mg/L

EPT

BOD5 vs. IBIR2 = 0.46

0

20

40

60

80

100

0.0 2.0 4.0 6.0 8.0

BOD5 mg/L

IBI

Chl-a vs. IBI

R2 = 0.37

0

20

40

60

80

100

0 20 40 60 80 100 120 140 160

Chl-a ug/L

IBI

(b)

68

DO Flux vs. EPT

R2 = 0.84

05

10152025

0 2 4 6 8

DO Flux mg/L

EPT

DO max. (96 hr.) vs IBI

R2 = 0.65

0

20

40

60

80

100

0.0 5.0 10.0 15.0

DO max. mg/L

IBI

DO min. (96 hr.) vs. IBI

R2 = 0.09

0

20

40

60

80

100

0.0 2.0 4.0 6.0 8.0

DO min. mg/L

IBI

DO max. (96 hr.) vs. EPTR2 = 0.56

0

5

1015

20

25

30

0.0 5.0 10.0 15.0DO max. mg/L

EPT

DO min. (96 hr.) vs. EPT

R2 = 0.06

0

5

10

15

20

25

0.0 2.0 4.0 6.0 8.0

DO min. mg/L

EPT

DO flux (96 hr.) vs. IBI

R2 = 0.49

0

20

40

60

80

100

0 2 4 6 8

DO flux mg/L

IBI

(c)

69

Turbidity vs. EPTR2 = 0.62

0

5

10

15

20

25

0 10 20 30 40 50 60

Turbidity NTU

EPT

TSS vs. IBI

R2 = 0.69

0

20

40

60

80

100

0 20 40 60 80 100

TSS mg/L

IBI

Turbidity vs. IBI

R2 = 0.68

0

20

40

60

80

100

0 10 20 30 40 50 60

Turbidity NTU

IBI

TSS vs. EPT

R2 = 0.94

0

5

10

15

20

25

0 20 40 60 80 100

TSS mg/L

EPT

(d)

70

B. Approaches for Nutrient Criteria Development Given these measurements, observations, and relationships (our study and the scientific literature) it is important to show how this information might be used to identify quantifiable thresholds for establishing nutrient criteria for the protection or improvement of stream health. Further, it will be valuable to frame this in a regional context as per USEPA (2000). A regional approach recognizes that rivers in different ecoregions may not be able to achieve the same endpoints (TP, TN, etc.) because of inherent differences in watershed soil fertility, land use, land-form, riparian cover and related factors. This study provides a variety of relationships and thresholds that can be used for developing nutrient criteria or goals for rivers. This includes relationships among TP, chlorophyll-a, BOD5, algal composition, diurnal DO flux, macroinvertebrate index and fish IBI scores. Other factors that may modify the relationships (and need be considered) include stream order (watershed size), river flow (residence time), TSS, and suspended sediment to name a few. Consistent with USEPA (2000), the ecoregions from which the rivers drain must be considered as well since the landscape factors that are used to delineate the ecoregions (e.g., landform, soils, and land use) may influence or dictate the potentials of streams. Since we are not moving directly into nutrient criteria development at this point it may be useful to demonstrate how these relationships and considerations might be used to develop nutrient goals for rivers. We will draw on the rivers (reaches) monitored in this study to serve as examples of various approaches for setting goals or ultimately developing criteria. These approaches will include the various considerations brought forth in USEPA (2000); such as: a) consideration of reference site data, b) use of predictive relationships, c) biocriteria considerations, and d) protection of downstream waters.

Low nutrient concentrations, protection-oriented approach: Crow Wing River

The Crow Wing River at Nimrod (CWR-72) exhibited TP concentrations that were quite typical for streams in the NLF ecoregion based on comparisons with data from minimally-impacted streams (McCollor and Heiskary, 1993) and well within the interquartile range based on USEPA data summary (Table 6). TP concentrations were slightly higher at the downstream site near Staples (CWR-35). However, the watershed is about twice as large (Table 1), with much of the drainage arising from the NCHF ecoregion (~42 % by area), and hence we would expect somewhat higher concentrations. Other parameters such as BOD5 and

turbidity were quite typical for the ecoregion as well (Table 5). A more complete analysis would consider flow and TP to allow for calculation of area-weighted yields and TP load. Other observations include: • algae and chlorophyll-a – Total chlorophyll-a about 4-5 µg/L at both sites in both summers.

Phytoplankton comprised predominately of benthic forms. Low diversity in periphyton. Abundant Cladophora noted at CWR-72 but periphyton chlorophyll-a was low.

• DO flux – DO remained above 5 mg/L with a low diurnal flux. • Fish IBI and QHEI – IBI was excellent at both sites and good habitat indicators. • Macroinvertebrate – high EPT value showed good diversity.

71

Based on these comparisons and related data for algae, diurnal DO and fish IBI, protection of current conditions seems appropriate for the upper portion of the Crow Wing River at Nimrod. The lower portion of the river at Staples appears to be relatively close to the typical ecoregion concentrations as compared to both the NLF and NCHF ranges and there were no algal or BOD-related problems that were evident based on the 1999 or 2000 data. Goals for this reach and the lower 35 miles to the mouth would require a closer review of data from some of the majority tributaries that drain to the Crow Wing, such as the Long Prairie River. Table 6. Crow Wing River data summary and comparison. Ecoregion interquartile ranges (25th-75th percentiles) drawn from Appendix I.

Site (ecoregion) Parameter MPCA EPA TP (ppb) IQ range IQ range 1999 2000 CWR-72 (NLF) 32 34 30-50 15-60 CWR-35 (NCHF) 59 49 70-170 40-200 Turbidity

(NTU)

CWR-72 3.0 2.5 2-4 CWR-35 4.0 3.3 5-10 2.6-5.8 BOD (ppm) CWR-72 1.0 1.2 0.9-1.6 CWR-35 1.0 1.2 1.2-1.9

Moderate nutrient concentrations, slight reduction approach: Rum River

TP concentrations were quite similar between sites and years for the Rum River (Table 6). The concentrations were within the typical range for the NCHF ecoregion (Table 6). Total chlorophyll-a, however, showed a distinct downstream increase in both years and was higher during the low flow summer of 2000. Slight downstream increases in BOD5 were noted as well (Table 7). The BOD5 values were near the upper end of the IQ range for the NCHF ecoregion. Related observations include: • Algae and chlorophyll-a – Phytoplankton were

comprised of a mix of benthic and planktonic forms. Blue-green forms were common in both years. Periphyton was fairly diverse. Abundant Cladophora noted at RU-18. Periphyton chlorophyll-a was high as compared to most other sites in this study.

• Turbidity – higher during high flow summer but comparable to IQ range (Table 6) • DO flux – DO remained above 5 mg/L with a moderate diurnal DO flux (6-13 mg/L). • Fish IBI and QHEI – IBI was good at both sites, though there was an increase in pollution

tolerant forms at the downstream site. Habitat was good at both sites and when combined with the moderate nutrient concentrations, adequate light and shallowness of the river makes it very favorable for periphyton growth

• Macroinvertebrate – high EPT value at RU-18 (Table 5).

72

While current (1999 & 2000) TP concentrations were not overly excessive, the Rum River at RU-34 and RU-18 would nonetheless benefit from a reduction in TP. At 100 µg/L, which is near the median for the MPCA and EPA data sets (Appendix I), chlorophyll-a would be reduced to about 20-25 µg/L (Fig. 9) and BOD5 would be near 1.5 mg/L (Fig. 14). Slight reductions in diurnal DO flux and slight improvements in fish IBI are possible based on Figs. 20 & 24. Table 7. Rum River data summary and comparison. Ecoregion interquartile ranges derived from Appendix I. Site (ecoregion) Parameter MPCA EPA TP (ppb) IQ range IQ range 1999 2000 RU-34 (NCHF) 137 143 70-170 40-200 RU-18 (NCHF) 131 133 70-170 40-200 Turbidity

(NTU)

RU-34 (NCHF) 10 6.2 5-10 2.6-5.8 RU-18 (NCHF) 8 6.3 5-10 2.6-5.8 Chl-a (ppb) RU-34 (NCHF) 20 29 RU-18 (NCHF) 27 44 BOD (ppm) RU-34 (NCHF) 1.6 1.8 1.2-1.9 RU-18 (NCHF) 1.8 2.3 1.2-1.9

Nutrient–rich river, BOD reduction approach: Crow River

The Crow River at the reach from CR-23 to the mouth of the river at CR-0.2 receives drainage from both the NCHF ecoregion via the North Fork (about 53% of watershed) and the WCBP ecoregion via the South Fork. For goal setting purposes we will focus on CR-23, as it is unaffected by any back-watering from the Mississippi and is co-located with the USGS gage, and the two upstream sites that reflect the relative inputs from the North and South Forks of the Crow.

Distinct differences in water quality are evident between the two subwatersheds based on data from 2001 for CR-44 (South Fork) and CRN-2.33 (North Fork) (Appendix III) with summer-mean TP concentrations of 436 and 262 µg/L respectively. As expected CR-23 is somewhat intermediate between these two values (Table 8) and had relatively consistent mean concentrations for the two years. High baseflow TP concentrations were evident as well (Fig. 5). TP concentrations are extremely high as compared to NCHF data and are at or above the 75th percentile for the WCBP ecoregion (Table 8). Total chlorophyll-a was high in both years and increased under the low flows of 2000. Turbidity was higher in 1999, consistent with high flow

73

conditions, and was very high as compared to both NCHF and WCBP ecoregion data. BOD5 concentrations were high compared to NCHF data and near the 75th percentile for WCBP ecoregion data (Table 8). Similar comparisons can be made between site CRN-2.33 (NCHF) and CR-44 (WCBP) based on data in Appendix I and III and in both cases these sites perform poorly as compared to minimally impacted rivers in their respective ecoregions. Other observations for CR-23 are as follows: • Algae and chlorophyll-a – Phytoplankton was comprised mostly of planktonic forms. Blue-

green forms were dominant in 1999 but were much less common in 2000. Many pollution tolerant forms were noted in both years. Periphyton was fairly diverse but dominated by forms that favor nutrient-rich conditions. Periphyton chlorophyll-a was quite low as compared to the other sites in 2000, which may have been related to high turbidity.

• DO flux – DO remained above 5 mg/L with a high diurnal DO flux (ranged from 5.5-13 mg/L).

• Fish IBI and QHEI – IBI was good at CR-23 but only fair at CR-0.2, however this may be more a function of the habitat at the site rather than the water quality.

• Macroinvertebrate – EPT index values were low, implying low diversity. Large river plankton biomass cannot be created without a substantial nutrient supply (Reynolds, 2000). By most measures the water quality of the Crow River at CR-23 and the two upstream sites on the North and South Fork is poor. Establishing a BOD5 goal, which could be readily translated into TP and chlorophyll-a goals, may be one approach for goal setting for the Crow. The rationale here would be to reduce oxygen demand generated by algae in the river, thus improving conditions for fish and other aquatic life and reducing impacts on downstream waters such as the Mississippi. In this instance a BOD5 goal on the order of 2 – 3 mg/L may be appropriate – this is in the range of the 50th –75th percentile for the NCHF and 25th – 50th percentile for the WCBP ecoregion (Appendix I). At 2.0 mg/L this corresponds to a TP concentration in the 100-150 µg/L range and at 3.0 mg/L this corresponds to a TP range of 140 – 225 µg/L (Fig. 14). The range in response is related to the differences in algal production between a high flow (1999) and low flow (2000) summer.

74

Table 8. Crow River data summary and comparison. Ecoregion interquartile ranges derived from Appendix I.

Site (ecoregion) Parameter MPCA EPA TP (ppb) IQ range IQ range 1999 2000 CR-23 (NCHF) 359 349 70-170 40-200 (WCBP) 210-350 130-359 Turbidity

(NTU)

CR-23 (NCHF) 53 41 5-10 2.6-5.8 (WCBP) 14-27 15.0-55.0 Chl-a (ppb) CR-23 (NCHF) 105 143 (WCBP) BOD (ppm) CR-23 (NCHF) 4.5 6.6 1.2-1.9 (WCBP) 2.2-6.6

Nutrient-rich river and high algal response: Blue Earth River The Blue Earth River is one of the largest tributaries to the Minnesota River. The Minnesota River has documented dissolved oxygen impairment on its lower reach and a TMDL is being developed to address sources of the problem. Previously linkages were made among TP, chlorophyll-a and BOD and as a result upstream control of TP is a priority in the TMDL (VanNieuwenhuyse and Jones, 1996). The Blue Earth River TP concentrations, though high relative to other rivers, are fairly typical for the WCBP

ecoregion (Table 9). Likewise there was minimal downstream increase in TP across the sites monitored. TP concentrations were slightly lower during the lower flow summer of 2000 as compared to 1999 and baseflow concentrations at the most downstream site (BE-18) were rather low, on the order of 50-150 µg/L. However, the river is quite efficient in producing algae (especially during low flow) as evidenced by the high summer-mean chlorophyll-a concentrations and maxima ranging to 195 µg/L (Table 3). The high chlorophyll-a contributes to elevated BOD concentrations (Table 9). Other observations for the Blue Earth are as follows: • Algae and chlorophyll-a – Phytoplankton were comprised mostly of planktonic forms. Blue-

greens and pollution tolerant diatoms and greens were common in both years. In particular slow-growing blue-greens such as Oscillatoria attest to the nutrient richness and long residence time in the river. Many pollution tolerant forms were noted in both years. Relatively few genera dominated the periphyton collections at the two sites monitored.

75

Periphyton chlorophyll-a was quite low as compared to the other sites in 2000 (Fig. 3), which may have been related to high turbidity and poor substrate quality.

• DO flux – DO remained above 5 mg/L with a very high diurnal DO flux (6.5-16 mg/L) noted at BE-73.2.

• Fish IBI and QHEI – IBI ranged from poor to very poor at four of the sites to good at the most downstream site. Habitat was generally poor at most sites.

• Macroinvertebrate – EPT values were high at the most upstream site (BE-100), but declined downstream and were quite low at BE-54 and BE-18 (Table 5).

While Blue Earth TP and BOD concentrations are fairly typical for WCBP streams, this does not imply that the “status quo” is an acceptable condition here as the high TP yields high chlorophyll-a and contributes to the high BOD5 in this river and downstream as well in the Minnesota River. Biotic data (Fish IBI, macroinvertebrate EPT, and algal composition) reflect impaired conditions. Rather, appropriate endpoints (criteria) for this river, while needing to be lower than current conditions (because of documented impacts on the Minnesota River), must be established within a regional framework. For instance, if the lower quartile BOD5 (2.2 mg/L for the WCBP) was used as a desirable endpoint (e.g., background BOD) for this ecoregion, this would correspond to a summer-mean TP of about 120 µg/L (low flow) to about 170 µg/L (high flow) (Fig. 13). This value would be below the lower quartile for summer TP in the WCBP ecoregion (MPCA - 210 µg/L) and begin to approach the 25th percentile from the EPA region-wide data (Table 9). A slightly higher BOD goal of 3.0 mg/L (between 25th-50th percentile for Minnesota data) would equate to TP concentrations between 150 µg/L (low flow) and 225 µg/L (high flow) based on Fig. 13. Reductions in TP will result in reductions in chlorophyll-a and BOD and these reductions will be most pronounced during low flow summers. These reductions should in turn reduce the diurnal DO flux and could ultimately benefit the biota of the river as well, though lack of habitat may constrain this somewhat. Reductions in TP, chlorophyll-a and BOD will be beneficial to, and necessary to meet the Minnesota River TMDL as well. Allocations from modeling associated with that TMDL provides another approach for establishing goals for the Blue Earth River. Table 9. Blue Earth data summary and comparison. Ecoregion interquartile ranges derived from Appendix I. Site (ecoregion) Parameter MPCA EPA TP (ppb) IQ range IQ range 1999 2000 BE-94 (WCBP) 247 192 210-350 130-359 BE-73 243 205 BE-54 248 207 BE-18 240 223 Turbidity

(NTU)

BE-94 (WCBP) 59 31 14-27 15.0-55.0 BE-73 57 41 BE-54 68 46 BE-18 68 57

76

Chl-a (ppb) BE-94 (WCBP) 41 48 BE-73 62 101 BE-54 81 97 BE-18 74 82 BOD (ppm) BE-94 (WCBP) 2.1 2.7 2.2-6.6 BE-73 3.6 5.1 BE-54 3.4 6.3 BE-18 3.4 5.3

Large river draining multiple ecoregions -- mass balance approach: Mississippi River

In our study the Mississippi River represents an example of a large river that drains multiple ecoregions (Fig. 26). The various sites on the river range from the predominately forested NLF ecoregion in the upper portion to a dominance by the NCHF ecoregion near the end of the study reach (Table 10). As such, goals or criteria could not reasonably be set at the same levels for the entire reach from UM-1056 – UM-872. Here it would seem reasonable to identify two or more index sites where the river transitions from one ecoregion to the

next and use data from these and sites as a basis for establishing goals. TP concentrations were relatively uniform at the NLF to NLF-NCHF transition sites and increased at UM-872 (Table 11). TP concentrations were generally just above the IQ range for the NLF but well within the NCHF range based on both the MPCA and EPA data sets. Chlorophyll-a was likewise fairly low at the upstream sites but increased at the downstream UM-872 site (Table 11). BOD values were well within the IQ range for the NLF and NCHF ecoregions. As with the other rivers in the study chlorophyll-a was higher in 2000 as compared to 1999 (Table 10). Related observations include: • Algae and chlorophyll-a – Phytoplankton were comprised primarily of planktonic forms.

Diatoms were more abundant at the upper sites while blue-green and green forms were more abundant at the lower sites (UM-953 and UM-872). Periphyton diversity was quite low at UM-1056 but high at UM-872. Periphyton chlorophyll-a at UM-872 was high as compared to all other sites in this study and exceeded 100 mg/m2 a threshold often associated with nuisance conditions (Dodds et al. 1997).

• Turbidity – Turbidity was slightly higher in 1999 as compared to 2000 and were slightly high as compared to the IQ range but well below the 25 NTU state standard (Table 10)

• DO flux – DO was low at UM-1056 and had a very small flux as compared to the other sites (Fig. 4). At UM-872 DO fell below 5 mg/L but a typical diurnal DO flux (4.5-10 mg/L) was noted.

• Fish IBI and QHEI – IBI ranged from good to excellent with a distinct increase from UM-1056 to UM-953 and a slight decline thereafter. Habitat was good at most sites with the exception of UM-1056, which lacked riffles.

77

Based on the aforementioned analysis, slight reductions in the TP concentration of the Mississippi over the reach from UM-1056 – UM-872 may be appropriate. The focus would be to minimize negative impacts from excess algal growth on the Mississippi itself as well as reducing the downstream transport of TP, algae, and BOD. A BOD goal of 1.0 mg/L (near 25th percentile) for the upper sites (UM-1056 – UM-953) would require a TP concentration of approximately 50 µg/L (corresponds to the 75th percentile for the NLF). A slightly higher BOD goal of about 1.5 mg/L may be appropriate for the downstream sites (UM-895 - UM-872) that are increasingly influenced by NCHF ecoregion tributaries such as the Crow, Sauk, and Elk Rivers. This translates to a TP on the order of 60-70 µg/L and chlorophyll-a of about 15 µg/L. A slight reduction in diurnal DO flux would likely result (Fig. 20) and this may minimize low DO events at the UM-872 site. Since habitat is good over the reach from UM-895 to UM-872 this could result in some improvement in IBI over this reach. Table 10. Ecoregion composition of Mississippi River sites Rivers Nearest

city River Mile Wshed

Area (mi2)NLF %

NCHF%

WCBP %

Ecoregion Class.

Year

Mississippi Aitkin UM-1056 5,885 100% NLF ‘00Mississippi Crosby UM-1029 6,088 100% NLF ‘00Mississippi Brainerd UM-1004 7,320 100% NLF ‘99Mississippi Little Falls UM-965.4 11,185 83% 17% NLF ‘99Mississippi Royalton UM-953.7 11,600 80% 20% NLF-NCHF Mississippi Monticello UM-895 13,400 51% 49% NCHF-NLF Mississippi Anoka UM-872 17,100 30% 65% 5% NCHF

78

Table 11. Mississippi River data summary and comparison. Ecoregion interquartile ranges derived from Appendix I. Site (ecoregion) Parameter MPCA EPA TP (ppb) IQ range IQ range 1999 2000 UM-1056 (NLF) 59 30-50 15-60 UM-953 62 54 UM-895 (NCHF) 67 77 70-170 40-200 UM-872 92 84 Turbidity (NTU) UM-1056 (NLF) 12 2-4 UM-953 10 8 UM-895 (NCHF) 9 7 5-10 2.6-5.8 UM-872 12 9 Chl-a (ppb) UM-1056 (NLF) 7 UM-953 7 12 UM-895 (NCHF) 10 17 UM-872 22 29 BOD (ppm) UM-1056 (NLF) 1.1 0.9-1.6 UM-953 1.1 1.3 UM-895 (NCHF) 1.2 1.6 1.6-3.3 UM-872 1.5 2.1 Determining relative contributions to TP flux at various reaches of the river may be useful for targeting reductions in TP and provides a complimentary basis for setting goals for some of the rivers in this study, e.g. Crow and Rum. If UM-872 is used as a focal point the relative contributions to summer TP concentration can be approximated by estimating flow-weighted means and loading rates (flux) for select upstream tributaries. The Crow River was a rather substantial contributor of the TP loading measured at UM-872 (Fig. 26 and Table 12). In 1999 it contributed about 29% of the TP flux at UM-872 and about 7% of the flow. In 2000, a lower flow summer, it contributed about 22% of the TP flux at UM-872 and about 6% of the flow. For purposes of illustration, using a simple mass-balance approach, we can estimate the reductions in TP concentration and flux at UM-872 that may arise from upstream TP reductions. For example, if TP concentrations in the Crow (CR-23) were reduced to 200 µg/L in 1999 and 2000 this would result in mean concentrations of 80 and 86 µg/L respectively for 1999 and 2000 at UM-872 (assuming conservation of mass). Reductions to 150 µg/L (as discussed in Crow River example) would yield concentrations of 77 and 80 µg/L respectively for 1999 and 2000. This rather simple technique can help determine the reasonableness of achieving goals over a range of flow conditions and provide some preliminary estimates of required reductions to achieve downstream goals. The accuracy of these predictions hinges on the availability of paired concentration and flow data that allow reasonable and representative estimates of flux for the index period – in this case summer. Ideally these data should be available over varying flow

79

regimes and with respect to nutrient impacts on rivers average to low flow conditions may be most important. Table 12. Mass-balance comparison for Mississippi and Crow Rivers. Miss. (UM-872) Crow (CR-23) 1999 2000 1999 2000 TP - mean 92 µg/L 84 µg/L 359 µg/L 349 µg/LTP- flow-weighted 92 µg/L 96 µg/L 359 µg/L 375 µg/LFlow (cfs) 12,021 5,399 886 301Flux (kgP/yr) 987,597 462,845 284,040 100,797% flow @UM-872 7% 6%% flux @UM-872 29% 22%

Figure 26.

UM-953.763 ppb57%

CR-23359 ppb29%

UM-100472 ppb44%

RUM-18138 ppb7%

UM-87292 ppb

987,597 Kg P/yr

CWR-150 ppb10%

CWR-72.333 ppb2%

St. Paul

Little Falls

St. Cloud

Minneapolis

BrainerdWadena

Aitkin

Anoka

Upper Mississippi River BasinSummer - mean TP and Percent of TP Flux at UM-872 for 1999

N

0 20 40 60 80 MilesJanuary 2001

Ecoregions:Northern Minnesota WetlandsNorthern Lakes and ForestsNorth Central Hardwood ForestsWestern Corn Belt Plains

County BoundariesWatershed BoundariesMajor River

80

High nutrients and low algal response, focus on a downstream receiving water approach: Red River

While the mainstem of the Red River lies within the Red River ecoregion, much of its watershed drains from the NCHF and NMW ecoregion to the east in Minnesota and the NGP ecoregion to the west in North Dakota. The upstream site at RE-536 was among the four sites used to characterize “minimally impacted” conditions for the RRV ecoregion (McCollor and Heiskary, 1993). Data from the NGP and RRV likely provide the best basis for placing the water quality of these mainstem sites in perspective (Table 11). The Red River, in particular the downstream reaches, had the highest summer-mean TP of any river in this study (Table 3) and exhibited individual concentrations in excess of 1 mg/L during high flow conditions (Fig. 8). Concentrations remained high at baseflow with TP on the order of 150 – 300 µg/L at the upper sites and 400 – 500 µg/L at the lower sites (Fig. 8). Summer-mean TP at RE-536 is within the typical range for the NGP and RRV ecoregions, while concentrations at RE-403 and RE-536 are far in excess of the 75th percentile for either ecoregion based on MPCA or EPA data sets (Table 13). Algal response to these high concentrations was quite low as compared to other rivers in this study, with the two downstream sites deviating greatly from the established relationships (Fig. 9). These two sites (RE-403 and RE-298) exhibited chlorophyll-a: TP ratios that were two – three times lower than the two upstream sites (Fig. 8). A principal reason for the lack of response was the extremely high turbidity at the lower sites – values that far exceed the 25 NTU turbidity standard and are far in excess of the typical range for streams in the NGP or RRV ecoregions (Table 13). Related observations are as follows: • DO concentrations remained above 5 mg/L at RE-536 and RE-498 but overall DO flux was

quite low (~1.5 – 2.0 mg/L, Fig. 4) and there was essentially no relationship between DO flux and chlorophyll-a or TP (Fig. 21).

• Periphyton samples at these two sites were dominated by diatoms and blue-green forms. Periphyton chlorophyll-a was high as compared to other nutrient-rich rivers like the Crow and Blue Earth (Fig. 2). No phytoplankton or fish IBI data were available for comparison.

Given this lack of response to TP how should nutrient–related goal setting be approached for the Red River? For the two upstream sites, RE-536 in particular, there may be some in-stream response to lowered TP concentrations since chlorophyll-a was near predicted levels (Fig. 9), transparency tube – chlorophyll-a relationship was similar to other streams (Fig. 20), and turbidity and TSS were not as excessive as the downstream sites. At RE-536 TP concentrations on the order of 160 (25th percentile NGP, MPCA) to 170 µg/L (25th percentile RRV, EPA) may be a reasonable goal. This should yield some reduction in chlorophyll-a (Fig. 9) and lower BOD (Fig.12) to near the 25th percentiles for the NGP and RRV ecoregions (Table 11). Reductions in TP at the lower two sites would likely not result in measurable reductions in chlorophyll-a (Fig. 9). In this instance goal setting should focus on other uses that might be impaired by excess nutrients (e.g., drinking water supplies, fisheries, etc.) and/or nutrient impacts on downstream waters, e.g., Lake Winnipeg in Canada. This large and shallow lake routinely experiences nuisance blooms of algae. A recent Manitoba Conservation estimate places the U.S. portion of the Red River TP contribution to Lake Winnipeg at 43 percent (Bourne et al. 2002). The actual load reductions needed to improve the condition of the lake would require a TMDL-like process and subsequent load allocation. This in turn would provide a basis for establishing

81

an overall TP goal, likely expressed as an annual flow-weighted mean (rather than a summer-mean). As this goal is apportioned upstream it will be important to consider the ecoregions that comprise the Red River drainage and the water quality characteristics associated with streams (subwatersheds) draining those ecoregions (e.g. Appendix I). This will aid in the establishment of reasonable and hopefully achievable goals. Table 13. Red River data summary and comparison. Ecoregion interquartile ranges derived from Appendix I. Site (ecoregion) Parameter MPCA EPA TP (ppb) IQ range IQ range 1999 2000 RE-536 (NGP) 208 160 - 290 210 - 448 RE-452 312 RE-403 (RRV) 602 140 - 330 170 - 285 RE-298 502 Turbidity

(NTU)

RE-536 (NGP) 27 20 - 37 RE-452 69 RE-403 (RRV) 151 13 - 28 RE-298 148 Chl-a (ppb) RE-536 (NGP) 27 RE-452 38 RE-403 (RRV) 38 RE-298 26 BOD (ppm) RE-536 (NGP) 2.8 2.6 - 5.6 RE-452 2.1 RE-403 (RRV) 4.7 2.0 - 4.5 RE-298 2.9 IV. SUMMARY AND RECOMMENDATIONS Strong relationships were evident among in-stream nutrients (TP and TKN in particular) and algae (expressed as total chlorophyll) for medium to high order streams during the summer growing season. These findings are similar to that found in other studies (Van Nieuwenhuyse and Jones 1996; Basu and Pick 1996). We have also shown that BOD5 is highly correlated with phosphorus and chlorophyll and that the living cells (as expressed by chlorophyll-a) and the dead or dying cells (as expressed by pheophytin-a) both contribute to the BOD. The relationships among nutrients, algae and BOD also varied (between sites and years) as a function of watershed size, flow and residence time. Overall, these relationships should prove useful for nutrient criteria development and TMDL-type load allocations (where excess nutrients and dissolved oxygen are a primary concern).

82

Our regression equations should be applicable elsewhere for rivers with similar characteristics as those in this study as indicated by data collected from several new rivers in 2001. Among the more important considerations are watershed size (which influences flow, residence time, mean depth) and inorganic turbidity. Our sites ranged from ~ 800 km2 up to ~ 44,000 km2 and stream order ranged from 4th to 6th order at most sites. Sites with watershed areas less than about 2,590 km2 (1,000 mi2) exhibited slightly lower chlorophyll per unit TP than sites with larger watersheds with the exception of the Red River. An upper threshold (watershed size or stream order) likely exists as well since, as watershed size increases, stream depth increases leading to increased depth of mixing and increased light limitation – as is common in main-stem reservoirs. Also, systems with excessive and continuously high TSS or turbidity, such as the lower segments of the Red River, will yield less chlorophyll-a per unit TP than our regressions would estimate. Our rivers exhibited a wide range in TSS and some, such as the Blue Earth and Crow, routinely exhibit very high TSS, in conjunction with runoff events. However, as these events subside, TSS declines as well and based on the chlorophyll concentrations recorded in this study, light limitation did not appear to be a major factor at most study sites (with the exception of the Red) during the summer index period. Shifts in seston composition were evident as a combined function of nutrient status and watershed size (residence time). In the shallower streams with low to moderate nutrient concentrations, such as the Crow Wing and Rum Rivers, benthic diatoms were a common component of the seston. In these systems there was an abundance of colonizeable habitat and adequate light reaching these substrates, which encouraged the growth of periphyton. These periphyton become part of the seston via sloughing or as a result of flood-event scouring. In deeper streams with larger watersheds and longer residence time, such as the Mississippi, planktonic greens and blue-greens were increasingly common. In highly nutrient-rich rivers, such as the Crow and Blue Earth, periodic light limitation from high inorganic turbidity and short residence time may combine to limit algal growth, however these systems often respond quickly as flow and turbidity decline and planktonic pollution-tolerant greens and blue-greens dominate. Regardless of the “origin” of the sestonic chlorophyll-a, rivers with high nutrients exhibited high chlorophyll-a and high BOD5 while those with low nutrients exhibited the inverse. Diurnal DO flux, calculated as the range between the minimum and maximum DO (as measured by continuous in-stream monitors over 96 hours in this study) can be used as a measure of stress on the aquatic biological community in the streams. Diurnal DO flux exhibited significant correlation with stream TP and chlorophyll-a whereby, for most sites, as TP and chlorophyll-a increased DO flux increased. Estimates of DO production and consumption showed similar relationships. Diurnal DO flux also exhibited significant relationships with fish IBI and macroinvertebrate EPT scores across the five rivers and ten sites that were assessed. In this case as diurnal DO flux increased fish IBI and macroinvertebrate EPT decreased. This relationship was much stronger than that based simply on minimum DO. However, maximum DO, which is related to algal productivity, exhibited strong negative correlations to fish IBI and EPT index values. In general, sites that deviated from the DO flux vs. IBI relationship tended to have very poor habitat (e.g. Mississippi River at Aitkin), which likely served to reduce IBI in addition to the water quality.

83

Strong correlation between transparency-tube measures, TSS, and turbidity were evident from this study, which is consistent with previous reports. Relationships range from linear at high transparencies (typical for Mississippi, Rum and Crow Wing) to more curvilinear as transparency measures fall below about 20 cm (typical for Crow, Blue Earth and Red). Further correlations were noted between transparency, TP, and chlorophyll-a. The relationship with TP was similar to that for TSS and is a function of the high association of TSS and TP in streams. When transparency was compared to chlorophyll-a some distinct “responses” were noted over the range of transparency, whereby at high transparency chlorophyll-a was low (similar to TP), as transparency decreased from about 20 cm to 12 cm distinct increases in chlorophyll-a were evident (abundant nutrients and adequate light) followed by a distinct decline as transparency fell below 10 cm (light limitation). Overall this analysis, combined with previous work (e.g. Sovell, 2000), demonstrated that transparency tube measurement provides a basis for estimating other parameters and an additional basis for describing stream health. In comparing water quality parameters from the Red River with those from the five other rivers in this study, the Red River is on the upper end in terms of turbidity and flow. In the lower turbidity rivers, nutrients can be used to predict primary productivity. However, in the Red River (in particular the lower reaches) these relationships break down. The high turbidity blocks light and less light is therefore available for algal photosynthesis. With increasing levels of turbidity, systems move from a biological control of primary production where nutrient availability controls production, to a physical control, where light availability is the limiting factor. High levels of turbidity in the Red River are likely a combined function of ecoregion (e.g., fine clay soils) and land use characteristics, as well as alterations in the drainage network that encourage high runoff following storm events. Establishing nutrient goals or criteria within an ecoregional framework is reasonable and desirable (in contrast to setting a single criteria value). At this point we have not proposed ecoregion-specific criteria but rather have described potential approaches for setting nutrient-related goals (Table 14). As a part of this effort other factors are considered including: index period, flow regime, stream order (implies watershed size and residence time), whether algal productivity is benthic or sestonic, the composition of the watershed (in terms of contributing ecoregion) upstream of the reach in question. Fish IBI, macroinvertebrate indices, and habitat assessment are important to consider as well. Ecoregion-based data summaries from representative sites, not impacted by point sources, can help in this exercise and should be combined with interrelationships such as those established in this paper (or others that preceded it) to derive reasonable and hopefully achievable criteria. In systems that do not exhibit distinct responses to nutrients (e.g., downstream reaches of the Red River) downstream receiving water concerns may dictate goal setting or criteria development.

84

Table 14. Summary of approaches for setting nutrient criteria. River Nutrients Algae & DO flux Fish IBI Approach Crow Wing Low Low, benthic dominated.

Low DO flux. good (good habitat)

Protection

Rum Moderate Moderate, mix of benthic & sestonic. Moderate DO flux.

good (good habitat)

Slight TP reduction

Crow High High, primarily sestonic, periodic blue-green dominance. High DO flux

good-fair (good habitat)

Establish BOD goal that can be translated into TP and chlorophyll-a goals.

Blue Earth High High, primarily sestonic, blue-greens common. Very high DO flux.

poor (poor habitat)

TP & BOD reduction

Mississippi Low-Moderate

Primarily sestonic, some blue-greens. Moderate DO flux.

generally good (habitat variable)

Mass balance -- seek TP reductions to minimize mainstem and receiving water nutrient-related impacts.

Red High Low algal response in lower reaches because of high turbidity. Low DO flux.

no data this study

Minimize impact on downstream uses and receiving waters

Following are some recommendations based on this work to date and ongoing efforts:

• Paired water chemistry and biological (fish and invertebrate) data should be collected at additional sites. These sites would ideally complement those used in this study and reflect a range in stream order, nutrient enrichment and related factors. Standard biological techniques that allow for calculation of appropriate indices should be employed. Water chemistry samples should be collected on a minimum of 6-8 occasions during the summer index period. Monitoring outside of the summer index period, in particular during spring high flows, is quite valuable as well and would be important if there is a need to calculate downstream mass loadings or flow-weighted means of nutrients for TMDL-type assessments.

• Measurement of flow is critical to understanding response of river to nutrients and calculation of residence time. Whenever possible sample sites (rivers) should have flow gauges at or near the sample site to allow for characterization of flow regime over the index period and placing the response of the river to nutrient enrichment in perspective.

• High correlations (R2) were noted among total N and fish IBI and macroinvertebrate EPT and nitrate-N and fish IBI. The N-rich (nitrate-N ~5-7 mg/L) Blue Earth River sites figured prominently in these relationships. It would be useful to collect data from several sites with slightly lower nitrate-N (e.g., 3-4 mg/L) to determine if the relationships identified in this study hold up across the entire range of nitrate-N (~0.1-7.6 mg/L) encountered in this study and for that matter much of Minnesota. This would provide useful information to help determine the necessity for developing N criteria for rivers.

• Algal analysis, including measurement of chlorophyll-a, pheophytin-a, and algal identification, in higher order streams (focus of this study) should focus on seston (phytoplankton). In low order, shallow, streams emphasis should be placed on periphyton, as seston will generally be a reflection of the benthic algae. Based on the moderate-high order streams in this study benthic algae did not exhibit strong relationships with nutrient enrichment, DO flux, or most other factors tested.

85

• Further measurement of diurnal DO flux should be conducted – preferably at sites with water chemistry, biological and flow data, to allow for a more comprehensive understanding of how these factors may be interrelated. If long-term sites cannot be established for this purpose (which would provide information on seasonal and year-to-year variability) monitoring should be targeted to mid to late summer when base flow or low flow conditions are likely to be encountered.

• Information from this study can be used to help develop river nutrient criteria in Minnesota and may be useful in other states as well. In the interim before criteria are promulgated into standards, as required by U.S.EPA, the data and relationships identified in this report can be used as a basis for assessing the impact of nutrients on individual river systems, setting water quality goals for those systems and contribute to the overall management of rivers.

86

REFERENCES Admiraal, W., L. Breehaart, G. M. J. Tubbing, B. van Zanten, E. D. de Ruijter van Steveninck and R. Bijkerk. 1994. Seasonal variation in composition and production of planktonic communities in the lower River Rhine. Freshwat. Biol. 32:519-531. American Public Health Association. 1998. Standard Methods for the Examination of Water and Wastewater. 20th Edition. Anderson, W. and J. Klang. 1997. Water pollution trading: from policy to reality. Environ. Engineer Baker, A. and K. K. Baker. 1979. Effects of temperature and current discharge on the concentration and photosynthetic activity of the phytoplankton in the upper Mississippi River. Freshwat. Biol. 9:191-198. Baker, K. K. and A. Baker. 1981. Seasonal succession of the phytoplankton in the upper Mississippi River. Hydrobiologia 83:295-301. Bailey, P.A., J.W. Enblom, S.R. Hanson, P.A. Renard, and K. Schmidt. 1994. A fish community analysis of the Minnesota River Basin. In Minnesota River Assessment Project Report, Volume III, Biological and Toxicological Assessment. 212 p. Basu, B. K. and F. R. Pick. 1995. Longitudinal and seasonal development of planktonic chlorophyll a in the Rideau River, Ontario. Can. J. Fish. Aquat. Sci. 52:804-815. Basu, B. K. and F. R. Pick. 1996. Factors regulating phytoplankton and zooplankton biomass in temperate rivers. Limnol. Oceanogr. 41(7):1572-1577. Berkman, H.E., and C.F. Rabeni. 1987. Effect of siltation on stream fish communities. Environmental Biology of Fishes. 18:285-294. Billen, G., J. Garnier and P. Hanset. 1994. Modeling phytoplankton development in whole drainage networks: the RIVERSTRAHLER Model applied to the Seine river system. Hydrobiologia 289:119-137. Bourne, A., N. Armstrong, and G. Jones. 2002. A preliminary estimate of total nitrogen and total phosphorus loading to streams in Manitoba, Canada. Manitoba Conservation Report No. 2002-04 De Ruijter van Steveninck, E. D., W. Admiraal, L. Breebaart, G. M. J. Tubbing and B. van Zanten. 1992. Plankton in the River Rhine: structural and functional changes observed during downstream transport. J. Plank. Res. 14(10):1351-1368. DeLorme. 1994. Minnesota atlas and gazetteer. Yarmouth ME 96 pp. Descy, J-P. P. Servais, J. Smitz, G. Billen, and E. Everbecq. 1987. Phytoplankton biomass and production in the River Meuse (Belgium). Wat. Res. 21:1557-1566.

87

Dodds, W. E., V. H. Smith and B. Zander. 1997. Developing nutrient targets to control benthic chlorophyll levels in streams: A case study of the Clark Fork River. Wat. Res. 7:1738-1750. Dodds, W.E. and E. Welch. 2000. Establishing nutrient criteria in streams. J. N. Amer. Benthol. Soc. 19(1):186-196. Heiskary, S. A. and C. B. Wilson. 1989. The regional nature of lake water quality across Minnesota: An analysis for improving resource management. J. MN. Acad. Sci. 55(1):71-77. Heiskary, S. A. and W. W. Walker, Jr. 1995. Establishing a chlorophyll a goal for a run-of-the-river reservoir. Lake and Reserv. Manage. 11(1):67-76. Heiskary, Steve, and Howard Markus. 2001. Establishing relationships among nutrient concentrations, phytoplankton abundance, and biochemical oxygen demand in Minnesota, USA, Rivers. Journal of Lake and Reservoir Management. 17(4): 1-12. Karr, J.R. 1981. Assessment of biotic integrity using fish communities. Fisheries. 6(6):21-27. Karr, J.R., K.D. Fausch, P.L. Angermeier, and P.R. Yant. 1986. Assessing biological integrity in running waters: a method and its rationale. Special Publications 5. Champaign, IL: Illinois Natural History Survey. 28 p. Kelly, M. G. 1998. Use of community-based indices to monitor eutrophication in European rivers. Environ. Conserv. 25(1):22-29. Kromer-Baker, K. and A. Baker. 1981. Seasonal succession of the phytoplankton in the upper Mississippi River. Hydrobiologia 83:295-301. Lee, K. E. 2002. Water-quality parameters and benthic algal communities at selected streams in Minnesota, August 2000 – study design, methods and data. USGS Open-file Report 02-43. Moundsview, MN. http://mn.water.usgs.gov. Leopold, L. B., G. Wolman and J. P. Miller. 1964. Fluvial processes in geomorphology. W. H. Freeman, San Francisco, CA. Lohman, K. and J. R. Jones. 1999. Nutrient – sestonic chlorophyll relationships in northern Ozark streams. Can. J. Fish. Aquat. Sci. 56:124-130. Lohman, K. and J. R. Jones. 1999. Nutrient – sestonic chlorophyll relationships in northern Ozark streams. Can. J. Fish. Aquat. Sci. 56:124-130. Lyons, J. 1992. Using the Index of Biological Integrity (IBI) to Measure Environmental Quality in Warmwater Streams of Wisconsin. Gen. Tech. Rep. NC-149. St. Paul, MN: U.S. Department of Agriculture, Forest Service, North Central Experiment Station. 51 p.

88

Mid-Atlantic Coastal Streams Workgroup (MACS). 1996. Standard operating procedures and technical basis: Macroinvertebrate collection and habitat assessment for low-gradient nontidal streams sites. Environ. Toxicology and Chemistry 11: 437-449. McCollor, S. and S. Heiskary. 1993. Selected water quality characteristics of minimally impacted streams from Minnesota’s seven ecoregions. Addendum to: Descriptive characteristics of the seven Ecoregions of Minnesota. Minnesota Pollution Control Agency. Meals, D. W., S. N. Levine, D. Wang, J. P. Hoffmann, E. A. Cassell, J. C. Drake, D. K. Pelton, H. M. Galarneau and A. B. Brown. 1999. Retention of spike additions of soluble phosphorus in a northern eutrophic stream. J. N. Am. Benthol. Soc. 18(2):185-198. Microsoft Corporation. 1997. Microsoft Excel: Version 97. USA Miltner, R. J. and E. T. Rankin. 1998. Primary nutrients and the biotic integrity of rivers and streams. Freshwat. Biol. 40:145-158. Minnesota Pollution Control Agency. 2000. Upper Mississippi River Basin Information Document. St. Paul, Minnesota: Minnesota Pollution Control Agency. 281 p. Niemela, S.L., and M.D. Feist. 2000. Index of Biotic Integrity Guidance for Coolwater Rivers and Streams of the St. Croix River Basin. St. Paul, MN: Minnesota Pollution Control Agency. 47 p. Niemela, S.L., and M.D. Feist. 2002. Index of Biotic Integrity Guidance for Coolwater Rivers and Streams of the Upper Mississippi River Basin. St. Paul, MN: Minnesota Pollution Control Agency. 56.p. Omernik, J.M. and A.L. Gallant. 1988. Ecoregions of the Upper Midwest States. EPA/600/3-88/037. Corvallis, OR: United States Environmental Protection Agency. 56 p. Palmer, M.C. 1969. A composite rating of algae tolerating organic pollution. J. Phycol. 5: 78-82. Prygiel, J. and M. Leitao. 1994. Cyanophycean blooms in the reservoir of Val Joly (northern France) and the development in downstream rivers. Hydrobiologia 289: 85-96. Rankin, E.T. 1989. The Qualitative Habitat Evaluation Index (QHEI). Rational, Methods, and Applications. Ohio EPA, Division of Water Quality Planning and Assessment, Ecological Analysis Section, Columbus, Ohio. Reynolds, C.S. 2000. Hydroecology of river plankton: the role of variability in channel flow. Hydrology, Process. 14:3119-3132. Reynolds, C. S., P. A. Carling and K. J. Beven. 1991. Flow in river channels: new insights into hydraulic retention. Arch. Hydrobiol. 121(2):171-179.

89

Rojo, C., M. A. Cobelas and M. Arauzo. 1994. An elementary, structural analysis of river phytoplankton. Hydrobiologia 289:43-55. Simonson, T.D., J.Lyons, and P.D. Kanehl. 1994. Guidelines for Evaluating Fish Habitat in Wisconsin Streams. Gen. Tech. Rep. NC-164. St. Paul, MN: U.S. Department of Agriculture, Forest Service, North Central Experiment Station. 36 p. Søballe, D. M. and B. L. Kimmel. 1987. A large-scale comparison of factors influencing phytoplankton abundance in rivers, lakes, and impoundments. Ecology 68(6):1943-1954. Søballe, D. M. and B. L. Kimmel. 1987. A large-scale comparison of factors influencing phytoplankton abundance in rivers, lakes, and impoundments. Ecology 68(6):1943-1954. Sovell, L., S. Heiskary, and J. Anderson. 2000. Using the transparency tube in Minnesota’s new Citizen Stream Monitoring Program. Proceedings of the 1999 National Water Quality Monitoring Conference. Stoner, J. 1991. National Water-Quality Assessment Program – Red River of the North. USGS Open file report 91-151. Strahler A.N. 1957. Quantitative analysis of watershed geomorphology. Trans. Amer. Geophysical Union. 38: 913-920 Swale, E. M. F. 1969. Phytoplankton in two English rivers. Jour of Ecology. 57: 1-23. U.S. Environmental Protection Agency. 1997. Technical Guidance Manual for Developing Total Maximum Daily Loads. Book 2: Streams and Rivers. Part 1: Biochemical Oxygen Demand / Dissolved Oxygen and Nutrients / Eutrophication. Office of Water, Washington D.C. EPA 823-B-97-002 1997 U.S. Environmental Protection Agency. 1998. Clean Water Action Plan: Restoring and protecting America’s waters. ISBN 0-16-049536-9. U.S. Environmental Protection Agency. 1999a. Methods for analysis of water. V2.0. CD-ROM Office of Water, Washington D.C. EPA 821-c-99-004 U.S. Environmental Protection Agency. 1999b. Protocol for Developing Nutrient TMDLs. Office of Water, Washington D.C. EPA 841-B-99-007 1999 U.S. Environmental Protection Agency. 2000a. Nutrient Criteria Technical Guidance Manual. Lakes and Reservoirs. Office of Water, Washington D.C. EPA-822-B00-001 U.S. Environmental Protection Agency. 2000b. Nutrient Criteria Technical Guidance Manual. Rivers and Streams. Office of Water, Washington D.C. EPA-822-B00-002 United States Geological Survey. 1999. Red River NAWQA study.

90

Van Nieuwenhuyse, E. E. and J. R. Jones. 1996. Phosphorus-chlorophyll relationship in temperate streams and its variation with stream catchment size. Can. J. Fish. Aquat. Sci. 53:99-105.

91

List of Appendices

I. Ecoregion-based data summaries for streams II. Ecoregion land use, fish IBI, and habitat data III. Metric and IBI scores for nutrient sites by river IV. Water quality data from 2001 river sites V. Precipitation departure from normal for 1999, 2000, and 2001

Appendix I. Ecoregion-based data summaries for streams Table 1. Interquartile range of summer-mean concentrations for minimally impacted streams in Minnesota, by ecoregion. Data from 1970-1992. TP = total phosphorus, TSS = total suspended solids. (McCollor and Heiskary 1993)

TP (mg/l) Turbidity (NTU) TSS (mg/l) BOD (mg/L) Region 25% 50% 75% 25% 50% 75% 25% 50% 75% 25% 50% 75%

NLF 30 40 50 2 2 4 2 4 6 0.9 1.2 1.6NMW 50 60 90 5 7 12 7 11 20 1.2 1.5 1.9NCHF 70 100 170 5 7 10 8 10 18 1.6 2.2 3.3

NGP 160 220 290 20 23 37 37 55 89 2.6 3.8 5.6RRV 140 220 330 13 19 28 28 50 74 2.0 2.8 4.5

WCBP 210 270 350 14 19 27 26 47 76 2.2 4.3 6.6 Table 2. Summer IQ range from USEPA (2000) nutrient criteria guidance documents

TP (µg/l) Turbidity (NTU)

Region 25% 50% 75% 25% 50% 75% NLF 15 30 60 NMW 50 60 80 NCHF (51) 40 95 200 2.6 3.9 5.8 NGP (46) 210 314 448 - - - RRV (48) 170 230 285 WCBP (47) 130 240 359 15.0 40.0 55.0

92

Appendix II. Ecoregion land use, fish IBI and habitat data Table 1: Site information and land use categories for nutrient sites Stream Name Field

Number Visit Date Drain

SqMi % Ag

% Range

% Urban

% Forest

% Wetland

% Water

% Disturbed

Fish IBI

Blue Earth River 00MN001 8/15/00 255 91 2 2 4 0 0 96 14 Blue Earth River 00MN002 8/15/00 804 91 3 2 3 0 0 96 24 Blue Earth River 00MN003 8/30/00 1371 89 4 2 3 0 2 95 24 Blue Earth River 00MN004 8/29/00 1395 89 4 2 3 0 2 95 22 Blue Earth River 00MN005 8/30/00 1528 89 3 2 4 0 1 95 40 Crow Wing River 00UM026 7/31/00 939 15 17 2 51 9 7 34 80 Crow Wing River 00UM024 6/13/00 2232 28 18 2 39 8 4 49 86 Mississippi River 00UM087 8/22/00 5838 2 7 1 52 24 13 11 68 Mississippi River 00UM088 8/22/00 6060 2 8 1 52 24 13 12 78 Mississippi River 00UM091 8/24/00 11730 11 12 2 46 16 10 25 88 Mississippi River 00UM092 8/29/00 14032 18 14 2 42 14 9 35 54 Mississippi River 00UM092 9/14/00 14032 18 14 2 42 14 9 35 62 Mississippi River 00UM098 9/13/00 19041 27 15 3 35 12 8 44 69 Rum River 00UM044 9/12/00 1273 17 23 2 31 9 18 42 69 Rum River 00UM044 9/18/00 1273 17 23 2 31 9 18 42 69 Rum River 00UM066 7/27/00 1325 17 23 2 31 9 18 42 77 Rum River 00UM066 8/28/00 1325 17 23 2 31 9 18 42 63 Crow River 00UM080 7/26/00 2633 69 12 4 7 3 5 85 65 Crow River 00UM081 8/14/00 2751 68 12 4 8 3 5 84 55

93

Table 2: Qualitative Habitat Evaluation Index (QHEI) by river Stream Name Field

Number DrainSqMi

Visit Date

Fish IBI

QHEI Dominant substrate (s)

Substrate Quality

Amount Instream Cover1

Channel stability

Riparian width2

Bank erosion Riffle/run substrate3

Blue Earth River 00MN001 255 8/15/00 14 43 Silt, gravel, sand

Silt moderate Sparse Low Moderate Moderate Unstable

Blue Earth River 00MN002 804 8/15/00 24 46 Gravel, sand Silt moderate Sparse Low Moderate Heavy/severe Unstable Blue Earth River 00MN003 1371 8/30/00 24 52 Cobble,

gravel, sand Silt normal Sparse Moderate R-narrow, L-

moderate Heavy/severe Mod.

stable Blue Earth River 00MN004 1395 8/29/00 22 48 Silt, sand Silt moderate Sparse Moderate Moderate Moderate Unstable Blue Earth River 00MN005 1528 8/30/00 40 42 Gravel, sand Silt normal Nearly absent Low Moderate Heavy/severe Unstable Crow Wing River 00UM026 939 7/31/00 80 79 Boulder,

cobble, gravel, sand

Silt free Moderate High Wide None/little Stable

Crow Wing River 00UM024 2232 6/13/00 86 80 Boulder, cobble, sand

Silt free Moderate High R-Wide, L-moderate

None/little Stable, unstable

Mississippi River 00UM087 5838 8/22/00 68 50 Silt Silt normal Sparse, nearly absent

Moderate R-moderate, L-wide

Moderate Unstable

Mississippi River 00UM088 6060 8/22/00 78 59 Silt, sand Silt normal Sparse High Wide None/little Mod. stable

Mississippi River 00UM091 11730 8/24/00 88 68 Cobble, sand Silt normal Moderate High Wide None/little Mod. stable

Mississippi River 00UM092 14032 8/29/00 54 63 Cobble, sand Silt normal Moderate High R-narrow, L-very narrow

None/little Stable

Mississippi River 00UM092 14032 9/14/00 62 67 Boulder, sand

Silt normal Moderate, sparse High Moderate None/little Stable

Mississippi River 00UM098 19041 9/13/00 69 65 Boulder, gravel, sand

Silt normal Sparse High Narrow None/little Stable

Rum River 00UM044 1273 9/12/00 69 65 Cobble, sand Silt normal Moderate High Moderate None/little Stable Rum River 00UM044 1273 9/18/00 69 60 Cobble, sand Silt normal Sparse High R-wide, L-

moderate None/little Stable,

unstable Rum River 00UM066 1325 7/27/00 77 78 Cobble,

gravel, sand Silt normal Moderate High Wide None/little Stable

Rum River 00UM066 1325 8/28/00 63 77 Cobble, gravel

Silt normal Moderate High Wide None/little Stable

Crow River 00UM080 2633 7/26/00 65 68 Cobble, gravel, sand

Silt normal Moderate High Moderate None/little Stable

Crow River 00UM081 2751 8/14/00 55 60 Cobble, gravel, sand

Silt normal Sparse High Moderate R-moderate, L-none/little

Mod. stable

1Extensive >75%, Moderate 25-85%, Sparse 5-25%, Nearly absent <5% 2Wide >50m, Moderate 10-50m, Narrow 5-10m, Very narrow <5m, None3Stable (Cobble, boulder), Moderate stable (Large gravel), Unstable (Fine gravel, sand)

94

Table 3: Metric and IBI scores for nutrient sites by river

Stream Name Field Number

DrainMi2

Fish IBI IBI

Rating

Total Taxa1 (Total native)

# Intolerant

% Tolerant

# DSN2

% Omnivore

# Piscivore (%Piscivore)3

% SL4

#Invert (%Benthic in.)5

% DELT

Blue Earth River 00MN001 255 14 very poor 12(11) 0 68 0 53 0 3 4 (2) 3 Blue Earth River 00MN002 804 24 poor 19(18) 1 69 2 62 5 (2) 9 9 (11) 1 Blue Earth River 00MN003 1371 24 poor 16(15) 1 39 3 (2) 23 6 (5) 19 9 (22) 2 Blue Earth River 00MN004 1395 22 poor 13(12) 0 64 1 60 0 3 7 (4) 0 Blue Earth River 00MN005 1528 40 good 19(18) 2 7 3 (2) 6 18 (5) 9 11 (9) 0 Crow Wing River 00UM026 939 80 excellent 18 2 12 4 6 4 43 10 0 Crow Wing River 00UM024 2232 86 excellent 27 4 12 4 10 5 68 16 2 Mississippi River 00UM087 5838 68 good 16 1 1 1 0 4 20 11 0 Mississippi River 00UM088 6060 78 good 23 2 3 2 2 6 28 13 1 Mississippi River 00UM091 11730 88 excellent 26 2 6 3 4 9 64 11 1 Mississippi River 00UM092 14032 54 fair 15 1 19 1 18 5 39 7 0 Mississippi River 00UM092 14032 62 good 15 1 13 1 13 6 42 6 1 Mississippi River 00UM098 19041 69 good 17 1 6 2 5 4 65 10 1 Rum River 00UM044 1273 69 good 22 2 23 3 4 5 13 10 0 Rum River 00UM044 1273 69 good 24 2 31 2 4 4 16 13 1 Rum River 00UM066 1325 77 good 19 2 13 4 12 4 58 10 0 Rum River 00UM066 1325 63 good 13 1 14 2 13 3 54 7 0 Crow River 00UM080 2633 65 good 23 1 30 3 23 5 19 13 0 Crow River 00UM081 2751 55 fair 16 0 37 2 14 3 26 8 2 1Total taxa used as metric for Upper Mississippi River Basin IBI; total native species used as metric for Minnesota River Basin IBI 2Number Darter Sculpin Notorus (DSN) used in Upper Mississippi River Basin IBI; number darter species used as metric for Minnesota River Basin IBI 3Number piscivore used as metric in Upper Mississippi River Basin IBI; percent piscivore used as metric in Minnesota River Basin IBI 4SL = Simple lithophils 5# Invertivore used as metric for Upper Mississippi River Basin IBI; % Benthic invertivore used as metric for Minnesota River Basin IBI

95

Table 4: Narrative guidelines for interpreting overall IBI (modified from Karr 1981; Karr et al. 1986, and Lyons 1992).

Biological Integrity Rating

Overall Upper Mississippi River

Basin

IBI Score

Overall Minnesota River Basin IBI

Score

Fish Community Attributes

Excellent 100-80 60-50 Comparable to the best situations with minimal human disturbance; a full array of age and size classes were represented.

Good 79-60 49-40 Species richness somewhat below expectations; size/age distributions may show signs of imbalance.

Fair 59-40 39-30 Decreased species richness; size/age distributions may show signs of imbalance.

Poor 39-20 29-20 Decreased species richness; size/age distributions may show signs of imbalance; growth rates and condition factors sometimes depressed; hybrids sometimes common.

Very poor 19-0 20-12 The community is indicative of an environment that is severely modified by human disturbance; few species present. Age/size distributions are abnormal; DELT fish (fish with deformities, eroded fins, lesions, or tumors) may be present in the most severely degraded environments.

No score No fish Thorough sampling finds few or no fish; impossible to calculate IBI.

96

Appendix III. Water quality data from 2001 river sites

River-sites monitored in 2001

River Location River mile Wshed mi2

St. Louis Scanlon SL-21 3,430 Crow - Main Rockford CR-23 2,520

N. Fork Rockford CRN-6 1,478 S. Fork Mayer CR-44 1,277

Rum St. Francis RU-18 1,360 Mississippi Anoka UM-872 17,100 Cannon Welch CA-13 1,472 Blue Earth Amboy BE-54 1,055 Mankato BE-0 1,551 Straight Fairbault ST-18 435 Watonwan Garden City WA-6 851 Cottonwood New Ulm CO-0.5 1,300 Redwood Redwood Falls RWR-1 629 Des Moines -- West Fork

Petersburg WDM-3 1,250

-- East Fork Ceylon EDM-6 202 Minnesota Delhi MI-212 ? Mustinka Wheaton MU-0 879 Otter Tail Fergus Falls OT-49 1,740 Buffalo Georgetown BR-3 1,108 Wild Rice Hendrum WI-3 1,628 Crow Wing Nimrod CWR-72 1,030 Crow Wing Staples CWR-35.5 2,130

97

Summer-mean concentrations for 2001 river sites

Site TEMP COND DO pH TURB Col BOD5 TSS TSV TKN NO3-N TP Chl-a Pheo Tube

C umhos mg/L SU NTU PtCo mg/L mg/L µg/L ug/L cm

BE-0 21 618 10.2 8.4 50 3.5 98 14 1.3 6.69 0.229 50.3 5.9 16

BE-54 21 616 7.9 8.2 47 4.2 82 16 1.6 4.42 0.393 69.9 15

BR-3 23 754 6.5 8.4 73 40 2.4 112 16 1.2 0.30 0.245 11.4 4.7 12

CA-13 20 571 8.7 8.4 6 2.0 13 3 4.38 0.155 14.3 5.5 55

CO-0.5 20 793 9.4 8.5 43 3.9 90 16 1.3 6.07 0.205 56.9 10.8 18

CR-23 22 629 8.3 8.4 29 4.3 53 13 1.6 1.04 0.296 58.8 17.6 17

CR-44 23 871 9.0 8.6 30 6.2 54 15 2.0 2.68 0.436 79.2 22.7 18

CRN-2.33 22 521 8.1 8.3 34 3.8 61 13 1.4 0.38 0.262 53.0 10.9 17

CWR-72.3 17 359 8.5 8.2 3 2 0.6 0.24 0.033 2.2 60

EDM-6 18 698 6.4 7.9 57 3.2 58 15 9.17 0.201 32.3 8.6 30

MI-212 23 827 8.5 8.5 23 3.0 52 10 1.5 0.55 0.206 36.5 10.2 26

MU-0 23 1138 5.4 8.2 24 49 3.0 27 6 1.6 0.41 0.355 17.4 7.0 23

OT-49 22 368 7.8 8.5 7 20 2.4 10 4 1.0 0.05 0.075 18.6 4.7 72

RUM-18 20 261 7.6 8.0 6 1.5 9 3 1.0 0.21 0.105 9.9 3.9 55

RWR-1 22 1200 8.9 8.5 27 2.3 40 9 1.1 3.42 0.253 41.2 12.9 21

SL-21 22 143 7.6 7.5 5 119 1.0 6 2 0.6 0.06 0.027 4.3 1.7 89

ST-18 18 694 8.3 8.0 11 1.2 19 4 6.90 0.382 8.7 2.7 45

UM-872 22 340 8.0 8.4 8 1.7 14 4 0.9 0.32 0.076 15.6 4.4 47

WA-6 19 723 9.5 8.5 26 2.9 55 10 1.3 7.98 0.373 35.0 3.5 27

WDM-3 21 664 8.8 8.6 35 9.3 55 19 3.25 0.235 154.4 19.0 19

WI-3 24 459 6.8 8.5 56 40 1.7 94 11 1.0 0.09 0.166 9.4 1.9 14

98

APPENDIX IV. Precipitation-departure from normal maps.

establishing relationships among in-stream nutrient ... · macroinvertebrate indices and...

Documents