J. Great Lakes Res. 23(4):416-439 Internal. Assoc. Great Lakes Res., 1997 Regionalized Loads of Sediment and Phosphorus to Lakes Michigan and Superior-High Flow and Long-term Average Dale M. Robertson! u.s. Geological Survey, Water Resources Division 8505 Research Way Middleton, Wisconsin 53562 ABSTRACT. Daily loads of suspended sediment and total phosphorus for the 10-year, 1-day design high flow and average of the 16-year period (1975 to 1990) were computed for 18 well-monitored tribu- taries to Lake Michigan and Lake Superior by use of constituent-transport models. The loads from these 18 reference tributaries were used to estimate the loads from all the United States tributaries (with drainage basins greater than 325 km 2 ) to Lake Michigan and Lake Superior by selection of a reference tributary with the most similar physical characteristics and use of a unit-area yield. Statistical compar- isons between computed yields and environmental factors were used to determine the physical character- istics that were most influential in selecting a reference tributary. Suspended sediment yields were affected primarily by river gradient and secondarily by the texture of surficial deposits, whereas total phosphorus yields were affected primarily by the texture of suificial deposits and secondarily by river gradient. Average total phosphorus loads were greatest in rivers entering the middle to southern part of Lake Michigan, especially those draining clayey surficial deposits in agricultural areas. During high flow, loads of phosphorus and suspended sediment from tributaries entering the southwestern part of Lake Superior dominate the total input of these constituents because of the steep gradients of the rivers and the clayey surficial deposits that they drain. These loads were used to compute regional loads and to rank the tributaries on the basis of their respective loads during a specified high flow and over extended periods. INDEX WORDS: Suspended sediment, phosphorus, water quality, sediment load, yield. INTRODUCTION The Laurentian Great Lakes system is the largest body of fresh water in the world. Each of the Great Lakes receives water and accompanying nutrients and sediments from many tributaries. Nutrient load- ing from some tributaries has caused eutrophication in bays of most of the Great Lakes, as well as throughout Lake Erie (Sonzogni et al. 1979). Ex- cessive sediment loading has caused water-clarity problems in nearshore areas and excessive sedimen- tation in many harbors. Reliable estimates of nutri- ent and sediment loading into the Great Lakes are needed for individual tributaries and for large re- gional areas over extended periods, as well as dur- ing specific types of high flow, to provide resource managers and policymakers a quantitative basis for ranking the tributaries on the basis of their relative ICorresponding author. E-mail: dzrobert@usgs.gov 416 contributions and for making decisions regarding these problems. A few of the many tributaries to the Great Lakes have been intensively sampled to estimate the load- ing of specific constituents, such as during the Green Bay Mass Balance Study (House et af. 1993) and the Lake Michigan Mass Balance Study (Hur- ley et af. 1996), but constituent loadings from most tributaries are unknown. Therefore, regionalized es- timates of nutrient or contaminant loading are not possible without a method for extrapolating load es- timates from a few well-monitored tributaries to the remaining unmonitored areas. Various approaches have been developed to estimate loading from un- monitored or partially monitored areas (Richards 1989). These range from unit-area-yield methods of estimating the load of an unmonitored river to ap- plying process-driven watershed models to the un- monitored basins. As part of the International Joint Commission