Capacity limitation and potential phosphorus yield

While it is clear that (probably all) phytoplank-ton can take up and assimilate the entire measurable MRP resource base, without first experiencing rate limitation, and that, thereafter, some species at least are extremely effective in maintaining their biomass, it is no less clear that the maximum supportable biomass, Bmax, cannot exceed the capacity of the most scarce resource relative to demand (Box 4.1). It remains generally true that, in a large number of larger, deeper lakes in the higher latitudes, phosphorus is the nutrient that is exhausted first and, thus, the one that imposes the upper limit on the supportive capacity of the location. The generality is supported by the well-known 'Vollenweider model' and the impressive fit of the average phytoplank-ton biomass present in a selection lakes to the corresponding average phosphorus availability in the same lakes (Vollenweider, 1968; 1976; Organisation for Economic Co-operation and Development, 1982; see also Section 8.3.1). Several generi-cally similar regression models from the same era (Sakamoto, 1966; Dillon and Rigler, 1974; Oglesby and Schaffner, 1975) provide analogous findings. It has to be recognised that the datasets are dominated by information from just the well-studied, northern-hemisphere oligotrophic lakes in which later understanding confirms the maximum phytoplankton carrying capacity is determined by the availability of phosphorus. It has equally to be recognised that the condition does not apply everywhere - it is less likely to apply to large, continental lakes at low latitudes, especially to lakes in arid regions, and to smaller, shallower lakes at all latitudes. It certainly does not apply to the open oceans, although its relevance to coastal waters should not be dismissed. Nevertheless, there remains a danger in supposing that the Vollenweider-type equations can be used to predict phytoplankton biomass in a given individual lake. Plainly, the criterion of capacity limitation by phosphorus must be demonstrable. Moreover, the equations are statistical and not predictive, indicating no more than an order-of-magnitude probability of average biomass that may be supported. The trite circumsciption of lakes or seas as being 'phosphorus-limited' (or 'nitrogen', or 'anything else' limited') is to be avoided completely.

Several authors have tried to express the maximum yield of biomass as a function of nutrient availability (Lund, 1978; Reynolds, 1978c). Lund regressed maximum summer chlorophyll against total phosphorus in a small lake (Blelham Tarn, UK) in each of 23 consecutive years during which the lake underwent considerable eutrophication. Reynolds (1978c) chose to regress the chlorophyll concentrations measured in several contrasted lakes in north-west England at the times of their vernal maxima against the corresponding MRP concentrations at the start of the spring growth. Despite certain obvious drawbacks to this approach (no allowance was made for intermediate hydraulic exchange and nutrient supply, neither were any other loss processes computed), the regression comes close to expressing the notion of a direct yield of algal chlorophyll for a known resource availability. This same regression equation (4.13) has been shown to be applicable to the prediction of the maximum algal concentration in other British lakes in which the MRP falls to analytically undetectable levels (Reynolds, 1992a; Reynolds and Davies, 2001), enhancing the supposition that the resource-limited yield is predictable from the available resource. It has also been used to estimate the chlorophyll-carrying capacity of the MRP resource in lakes where the maximum crop is susceptible to other limitations (Reynolds and Bellinger, 1992) and is now incorporated into the capacity-solving model of Reynolds and Maberly (2002).

Whereas, it was originally estimated that:

where [MRP]max is the highest observed concentration of MRP and [chla]max is the predicted maximum chlorophyll (both units in |g l-1 or mg m-3), the later applications are used to predict an instantaneous yield against the supposed bioavailability of P. Thus,

Estimating exactly what is bioavailable, without enormous analytical effort, remains problematic. However, on the assumption that phosphatase activity will raise the supportive capacity only negligibly and that mixotrophic enhancement rarely applies outside the habitats in which it is recognised (see Section 4.3.3), then the resource currently available to the phytoplankton is represented by the unused MRP in solution plus the intracellular phosphorus already in the algae. The minimum estimate of the resource in intra-cellular store can be gauged simply by reversing Eq. (4.15) to solve the BAP invested in the standing crop.

BAP can be estimated by first solving Eq. (4.16), based on the existing chlorophyll concentration, then adding the equivalent intracellular cell P content thus predicted to the existing MRP concentration. Substituting this solution in Eq. (4.15) gives an instantaneous carrying capacity and potential chlorophyll yield.

Figure 4.6

(a) Observed maximum chlorophyll concentrations in lakes in north-west England as a function of bioavailable phosphorus, as detected by Reynolds (1992a) and the regression originally proposed by Reynolds (1978c) on the basis of a study of just three lakes (see Eqs. 4.14, 4.15). (b) A later, larger dataset of observed chlorophyll maxima from UK lakes plotted against the predtion of the Reynolds regression. As expected, a majority of points lie below the predicted maximum but a number are above it, sometimes substantially so. Graphs redrawn from Reynolds and Davies (2001) and Reynolds and Maberly (2002).

Figure 4.6

(a) Observed maximum chlorophyll concentrations in lakes in north-west England as a function of bioavailable phosphorus, as detected by Reynolds (1992a) and the regression originally proposed by Reynolds (1978c) on the basis of a study of just three lakes (see Eqs. 4.14, 4.15). (b) A later, larger dataset of observed chlorophyll maxima from UK lakes plotted against the predtion of the Reynolds regression. As expected, a majority of points lie below the predicted maximum but a number are above it, sometimes substantially so. Graphs redrawn from Reynolds and Davies (2001) and Reynolds and Maberly (2002).

L-1 (i.e. 6.32 |g chla (|g BAP)-1); 10 |g BAP L-1 will support <24 |g chla L-1 (i.e. 2.4 |g chla (|g BAP)-1), whereas the return on 100 |g BAP L-1 is <91.4 |g chla L-1 (i.e. 0.91 |g chla (|g BAP)-1). A small, biomass-limiting BAP has to be used very efficiently but a larger base, one that perhaps challenges the next potential capacity of the biomass, is used with more 'luxury', at least before the external resource is exhausted. In terms of biomass, the effect may be even more striking, bearing in mind the tendency towards relatively lower biomass-specific chlorophyll contents of phytoplankters in sparse, light-saturated, nutrient-limited populations. Supposing a constant quota of 0.02 |g chla (|g cell C)-1 (see Section 1.5.4), the phytoplankton carbon yields available from 1-100 |g BAP L-1 may be calculated to be from 316 down to 45.7 |g C (|g BAP)-1. The corresponding range of phosphorus quotas, 0.0012-0.0085 mol P (mol cell C)-1, neatly spans the condition of cells close to their minimum (q0) to being close to the Redfield ideal (C : P ratio >800 to 118). This outcome is possibly more realistic than the direct solution of Bmax = Ki/q0 (as proposed in Box 4.1), which may greatly exaggerate outcomes extrapolated from abundant resource bases.

Was this article helpful?

0 0

Post a comment