Abstract
The validity of the OECD annual mean in‐lake total phosphorus (TP) to chlorophyll (Chl.a) relationship as a predictive tool in eutrophication studies is questioned. Annual mean ratios of dissolved inorganic nitrogen / dissolved reactive phosphorus which are > 10 seem inadequate to identify “phosphorus‐limited lakes” as this yields a distribution of data points which can be covered by New Zealand lakes and some eutrophic lakes in the OECD data set which appear to be “nitrogen‐limited”. Evidence from New Zealand lakes supports the view that the distribution of supposedly “phosphorus‐limited lakes” of the OECD data set contains a systematic error, the probability of biomass control by phosphorus shortage being greater where the Chl.a / TP ratio is high for any specified annual mean Chi .a concentration. It is claimed that the predictive capability of the OECD data set will be improved if it is accepted that TP concentrations may have to be reduced without concomitant reduction of Chl.a until the Chl.a / TP ratio equates with a position near the upper left‐hand side of the data distribution. Chl.a reduction will then occur with further reduction of TP following the upper left‐hand boundary of the data distribution. This approach will accommodate situations where algal biomass is initially controlled by light or nutrients other than phosphorus.