98 



vegetation. Combined floating and submerged biomass can be 

 assessed historically using equation 3.6. Despite an R2 of 

 approximately 0.60, the adjusted R2 indicates that this model, which 

 is based on 7 diatom taxonomic groups, explains about 39% of the 

 variance in floating and submerged biomass. 



A pplying Predictive Models to Obtain Historical TSI Estimates 

 The predictive models shown above can be used to determine 

 the historical trophic state classification of lakes taking the nutrients 

 in macrophyte biomass into account (Canfield et al. 1983a). The 

 floating-leaved and submerged biomass model (equation 3.6) can be 

 used in conjunction with equations 3.4 or 3.8, predicting percent- 

 , ., ,.. area coverage, in order to estimate the total submerged and floating- 

 V'^I-'J leaved biomass for historical samples. Total floating-leaved and 

 submerged biomass (TFSB) is calculated as: 

 TFSB = SA X (PACi / 100) x Bi 



I H 





where SA = lake surface area (m^) 



PACi = inferred percent-area coverage 



Bi = inferred floating-leaved and submerged biomass (kg m-2). 



Canfield et al. (1983a) report different percent P values for the 

 dry weight of individual macrophyte taxa, though variation is shown 

 between lakes in those values. It is difficult to calculate accurately 

 the amount of P contained in the inferred total biomass because the 

 biomass of individual macrophyte taxa cannot be determined from 

 the diatom predictive models, and inferences are for wet weight of 

 macrophyte biomass. If an overall mean percent P value (0.234%) is 



