PREDICTING BIOMASS OF SUBMERSED AQUATIC VEGETATION USING THE LTRMP AQUATIC 
VEGETATION RAKE METHOD

Kevin P. Kenow1, James E. Lyon1, Randy K. Hines1, and Abdulaziz Elfessi2
1U.S. Geological Survey, Upper Midwest Environmental Sciences Center, La Crosse, 
WI 54603 2Department of Mathematics, University of Wisconsin-La Crosse, La 
Crosse, WI 54602.

We assessed the distribution and biomass of submersed aquatic vegetation in the 
lower end of Navigation Pool 8 at about 200 randomly selected sites annually, 
during 1999 through 2001.  Trained personnel conducted vegetation sampling per 
the established Long Term Resources Monitoring Program procedures and at the 
same time harvested all vegetation present in four 1/3-m2 quadrats.  Rake index 
values and biomass data were summarized by site and the resulting data used to 
model the relationship between rake indices and vegetation biomass (both total 
and for individual species).  We constructed linear regression models using log-
transfomed biomass data for sites visited in 1999 and 2000.  Data collected in 
2001 was used to validate the resulting models by assessing the predictive 
capability of each model.  The coefficient of determinations (R2) for predicting 
total biomass was 0.84 and for individual species ranged from 0.60 for water 
milfoil (Myriophyllum spicatum) to 0.89 for coontail (Ceratophyllum demersum).  
The accuracy and precision of models tested using independent data indicate that 
the LTRMP aquatic vegetation rake method data can be used to reliably predict 
total vegetation biomass and biomass of selected species. 

Keywords: biomass, LTRMP, Navigation Pool 8, sampling, submersed aquatic 
vegetation