Project: 2006-02523

Title: Whole Lake Experiments to Control Harmful Algal Blooms in Multi-Use Watersheds

Project Director: John Lehman (

Institution: University of Michigan , Ann Arbor , MI 48109

Project Period: September 1, 2006 to August 31, 2009

Amount: $383,355

This project addresses the source, fate, transport, and control of toxic cyanobacteria in a multi-use, multiple lake watershed that is used for agriculture, drinking water, recreation, sport fishing, irrigation, wastewater disposal, industrial processes, and hydroelectric power generation. It follows on research supported by the Waters and Watersheds program that was intended to develop a management plan for eliminating toxic cyanobacteria blooms in a chain of reservoirs along the Huron River in southeastern Michigan . Research results overturned some existing paradigms for management of these lakes and produced a rich data and public outreach resource accessible at . A management plan has now been developed that couples lake physics, chemistry, biological processes, and economics based on revenue from hydropower. Cooperation has been received from municipal governments and from state environmental regulators to conduct whole lake experiments guided by model predictions, using real time estimates of hydrology, power potential, and water quality, to learn if the new approach will fix the problem at acceptable cost.

New management theory will be tested by whole lake manipulations including selective depth withdrawal and oxygen addition. The specific objective is to halt toxic blooms without harming a substantial sport fishery or the water quality of downstream lakes. Hypotheses include the following:

This proposal aims to provide answers to a series of questions:

  1. Will rates of selective water withdrawal predicted by a numerical model prove sufficient to prevent anoxia and to reduce accumulation of phosphate and ammonium?
  2. Will rates of oxygenation predicted by the same model prevent anoxia and reduce accumulation of phosphate and ammonium?
  3. Will reductions in hypolimnetic concentrations of P and N lead to (a) reduced epilimnetic P, (b) reduced nutrient export to downstream lakes (transport), (c) reduced summertime algal biomass (fate), (d) reduced occurrence of cyanobacterial blooms?
  4. Can computer-generated projections be translated into operational decision making fast enough to control water quality?
  5. Can an algorithm be developed to improve management practices with attention to both environmental quality and economic considerations?


Groundwork has been laid through outreach, education, information sharing, and policy development in true partnerships with citizen advisory commissions, municipal governments and NGOs. This makes it possible to get approval and willing cooperation for manipulations to a heavily used lake four square kilometers in size. As with previous work, we will use public surveys to evaluate outreach and project visibility.

In order to achieve the objectives, we will proceed as follows:

  1. Conduct mass balance and in-lake measurements on 3 lakes: an upstream control, an experimental lake, and a lake downstream from the experimental basin. Each of these lakes already has 3 years of pre-manipulation data for reference.
  2. Apply a computer model developed in Excel™ and Visual Basic to real time water flow data and lake conditions to guide decisions about water discharge at a hydroelectric dam.
  3. Refine the model to improve treatment of weather effects on lake stratification and vertical mixing.
  4. Coordinate activities with outreach and stakeholder involvement led by a local NGO and municipal partners.
  5. Based on experiment results, and in consultation with government officials, propose long term sustainable structural or management changes that can be implemented by local governments.

Moored instruments will be deployed. Control lakes will have water samples retrieved weekly; the experimental lake will be sampled daily during critical periods to assess the longitudinal distribution of temperature, oxygen, nutrients, and algae. Routine field analyses include temperature, specific conductance, pH, dissolved oxygen (by luminescence), Secchi transparency depth, irradiance (lake profiles), and in vivo fluorescence of both chlorophyll and phycocyanin. Chemical determinations will include nitrite, nitrate, ammonium, DN, PN, silicate, SRP, DP, TP, chlorophyll, phycocyanin, CDOM, alkaline phosphatase, and microcystin by ELISA. Scientific collaborators will measure DOC (Ann Arbor Water Utilities) and specific algal toxins (USDA). AAS and ICP-MS are available as needed for metals and major cations. Algal species composition will be assessed by inverted microscope. We will measure meteorological variables relevant to heat balance and wind stress.

We will test whether it is possible to withdraw water selectively to maintain oxygen and suppress internal loading. Through separate experiments we will test model predictions about direct oxygen supply and dispersion. The critical question will be whether these manipulations suppress toxic cyanobacteria. Specifically, we predict:

EPA Grant Number: R830653-010

Title: Adaptive Management for Improved Water Quality in Multi-Use Watersheds

Investigator: John Lehman

Institution: University of Michigan

Project Period: February 1, 2003 to January 31, 2007

Project Amount: $745,883

Research Category: Nutrient Science for Improved Watershed Management Program

This project will develop a management plan for eliminating nuisance algal blooms in a chain of reservoirs along the Huron River in southeastern Michigan. The river-reservoir system is used for municipal drinking water, wastewater disposal, irrigation, industrial processes, hydroelectric generation, sport fishing, and recreation. The impoundments episodically develop surface scums of bluegreen algae, and emit foul odors including hydrogen sulfide. The project is highly relevant to local, regional, and State efforts that have been trying for years to improve water quality in this watershed. Partnerships have been formed to ensure that the scientific information gained through this project become translated into education, outreach, policy, and decision-making. Past management strategy has focused on phosphorus loading alone, but it has failed to prevent massive nuisance conditions as recently as summer 2001. Efforts based on new thinking successfully predicted the 2001 blooms; now this study will sharpen the predictions and develop management approaches to eliminate nuisance conditions in the future.

Objectives/Hypotheses: Project objectives include compartmentalizing the river system to pinpoint watershed segments responsible for inputs and internal processes that reduce stoichiometric ratios of nitrogen to phosphorus. Evaluating the role of redox transformations and assessing magnitudes of anaerobic nitrate respiration are key elements of the research plan. The project will evaluate interactions of river discharge volumes, internal transformations, and weather events as components of adaptive management theory. It will identify the places and conditions that account for significant amounts of denitrification, as well as reservoir management responses that can counteract the conditions that promote nuisance blooms.

Approach: The project emphasizes both a firm scientific foundation and translation of scientific knowledge into outreach, education, and policy development. Partnerships have been formed with municipal governments and NGOs to seek community input, disseminate information, and evaluate outreach activities. Scientific investigations will focus on nitrogen, phosphorus, trace metals, redox chemistry and hydrodynamic interactions with weather events and hydroelectric dam operations.

Expected Results: Successful management practices developed with cooperation from municipalities and community governments will be communicated through a well-developed outreach program. These best practices have application to similar problems in other watersheds.