Phosphorus-Biomass Relationships in Lakes
151
In the preceding sections, model output has been presented as function
of input concentration, although some emphasis has also been placed on
the interaction between water flow and input concentration in determining
the critical P loading level for stable grazer control of algal biomass. Thh
presentation form hides the fact that for different loading levels, leading to
the same average level of total particulate P (i.e., P + (JZ), the average
phyto- and zooplankton biomasses will be quite similar. Figure 5.18 shows,
as an example, the predicted relationships between grazer biomass and
total particulate P for two dilution rates differing by an order of magnitude.
The difference in the two curves is caused by the increasing importance of
algal sinking losses to the phosphorus retention with increasing water residence time. Figure 5.18 indicates that the phosphorus-biomass relationships predicted by the present model are insensitive to differences in the
flushing rate (D), and that they therefore should apply across lakes with
different hydraulical characteristics.
NIV A made an impressive survey of plankton biomass and water chemistry in 355 Norwegian lakes, ranging from ultraoligotrophic to hypereutrophic in temperate to arctic and subalpine areas (Faafeng et al. 1990).
One problem with relating the results from the NIV A study to the model
predictions is that only total phosphorus was measured. This means that
some decision must be made on how the forms of phosphorus considered
by the model relate to measured total phosphorus; or more specifically,
whether dissolved organic phosphorus (OOP) can be considered an active
part of the pelagic phosphorus cycle, or not. The low variability in the relative OOP fraction of total P (Fig. 2.4), could be taken as an argument for
both views. If measured OOP is biased by leakage of cell contents from
broken cells (Taylor and Lean 1991), it should certainly be considered as a
biologically active form. On the other hand, if the partitioning is the result
of a slow exchange process with the biologically active forms of phosphorus, nop should not be part of the phosphorus cycle represented by the
present model.
Figure 5.19 shows the relationship between zooplankton biomass and
total particulate P predicted by the present model. The main feature of the
model prediction is that the relationship appears to be hyperbolic rather
than allometric, as assumed by most regression models. The vertical
asymptote of the hyperbolic relationship corresponds to the critical input P
concentration P'L' below which the system is unable to support a second
trophic level, while the horizontal asymptote corresponds to the asymptotic
zooplankton biomass Z'. When data from the NIV A survey are recalculated
under the assumption that the OOP fraction of total P does not participate
directly in the pelagic phosphorus cycle, the zooplankton biomass is seen to
reach the predicted upper limit in some lakes, while remaining more than
two orders of magnitude below the potential yield in other lakes. Making
the opposite assumption with regards to availability of OOP would displace
the data points to the right in Fig. 5.19, but otherwise preserve this pattern.
151
In the preceding sections, model output has been presented as function
of input concentration, although some emphasis has also been placed on
the interaction between water flow and input concentration in determining
the critical P loading level for stable grazer control of algal biomass. Thh
presentation form hides the fact that for different loading levels, leading to
the same average level of total particulate P (i.e., P + (JZ), the average
phyto- and zooplankton biomasses will be quite similar. Figure 5.18 shows,
as an example, the predicted relationships between grazer biomass and
total particulate P for two dilution rates differing by an order of magnitude.
The difference in the two curves is caused by the increasing importance of
algal sinking losses to the phosphorus retention with increasing water residence time. Figure 5.18 indicates that the phosphorus-biomass relationships predicted by the present model are insensitive to differences in the
flushing rate (D), and that they therefore should apply across lakes with
different hydraulical characteristics.
NIV A made an impressive survey of plankton biomass and water chemistry in 355 Norwegian lakes, ranging from ultraoligotrophic to hypereutrophic in temperate to arctic and subalpine areas (Faafeng et al. 1990).
One problem with relating the results from the NIV A study to the model
predictions is that only total phosphorus was measured. This means that
some decision must be made on how the forms of phosphorus considered
by the model relate to measured total phosphorus; or more specifically,
whether dissolved organic phosphorus (OOP) can be considered an active
part of the pelagic phosphorus cycle, or not. The low variability in the relative OOP fraction of total P (Fig. 2.4), could be taken as an argument for
both views. If measured OOP is biased by leakage of cell contents from
broken cells (Taylor and Lean 1991), it should certainly be considered as a
biologically active form. On the other hand, if the partitioning is the result
of a slow exchange process with the biologically active forms of phosphorus, nop should not be part of the phosphorus cycle represented by the
present model.
Figure 5.19 shows the relationship between zooplankton biomass and
total particulate P predicted by the present model. The main feature of the
model prediction is that the relationship appears to be hyperbolic rather
than allometric, as assumed by most regression models. The vertical
asymptote of the hyperbolic relationship corresponds to the critical input P
concentration P'L' below which the system is unable to support a second
trophic level, while the horizontal asymptote corresponds to the asymptotic
zooplankton biomass Z'. When data from the NIV A survey are recalculated
under the assumption that the OOP fraction of total P does not participate
directly in the pelagic phosphorus cycle, the zooplankton biomass is seen to
reach the predicted upper limit in some lakes, while remaining more than
two orders of magnitude below the potential yield in other lakes. Making
the opposite assumption with regards to availability of OOP would displace
the data points to the right in Fig. 5.19, but otherwise preserve this pattern.