2. Shallow lakes and deep lakes

Limnology, the science of freshwaters , is supremely an integrative environmental science [108]. To understand how a lake or river system works demands knowledge of the nature of the catchment from which the water is derived, its geology, land use, meteorology, hydrology and soil science. All these influence the amount of water flowing into a lake, and its quality, especially the amounts of phosphorus and nitrogen compounds that are key nutrients in determining the productivity of the lake. The greater the human use of the catchment, whether by urban development or agriculture, the greater will be the annual loads of N and P that reach the lake. Natural communities have evolved mechanisms of retaining what are scarce resources for them. Removal of natural land vegetation destroys mechanisms of conservation of these nutrients within the terrestrial system.

(Fig 3). Agricultural systems do not have these mechanisms and are comparatively 'leaky' of nutrients

Fig 3 Major pathways of nutrients in deep and shallow lakes. Shallow lakes are those in which aquatic plants are able to colonise at least 50% of the bottom area.

The behaviour of these nutrients in deep lakes can be described by simple models, for the mass of water in such a lake basin is relatively uniform. It is not absolutely uniform, of course, and contains a wealth of detailed variety that is important in its ecology, but much understanding can be gained by treating it as a completely mixed 'reactor' [153]

Water entering the lake will generally be relatively richer in nitrogen than phosphorus, simply because this planet's supply of phosphorus is relatively low compared with that of nitrogen and because phosphorus compounds tend to be rather more insoluble than nitrogen ones. Both of these elements will be present in inorganic and organic dissolved forms and in particulate forms. But because microorganisms can make most forms ultimately available for algal growth, we bulk them as the 'total phosphorus' and 'total nitrogen' for consideration

Though agricultural soils seem to be becoming saturated with phosphate fertilizers and more phosphate is running off the land than previously [55,56,69,70], nitrogen, particularly in the form of nitrates, is very soluble and thus the total N to total P ratio of runoff waters, tends to be higher (often >>10:1 by weight) than the ratio in which algae require these elements (around 7-10:1 by weight). This means that in deep lakes, as phytoplankton algae suspended in the water grow and take up these elements, phosphorus is likely to be the first of the two to become limiting and will determine the potential maximum crop or biomass that the water can support [128,129]

The total phosphorus concentration is usually lower in a deep lake than the concentration in the incoming stream water. This is because some of the phosphorus, having entered algal cells, is lost to the bottom sediments, where it is fixed and does not easily return to the illuminated surface layers where algal growth can take place. Phosphorus is combined with oxidised iron and manganese compounds in the surface sediment layers where oxygen can diffuse in and maintain high concentrations of Fe3+ and Mn3+

Phosphorus is also lost through the overflow of a lake, so that with these two major sinks, the concentration cannot normally build up in the water and must be maintained by continued inflow (Fig 3). A simple equation, the Vollenweider model [153] describes these consequences for mixed deep lakes:

[M]=…L / z (s+r)

In this model, [M] is the concentration of phosphorus (it can be applied to any element, however, as this is a general model); loads from all possible sources (soil, effluent, rain etc) are combined as …L, z is the mean depth of the lake, s is the net rate at which phosphorus is lost to the sediment (or strictly is exchanged with the sediment, as it can in certain circumstances be released, when s becomes negative) and r is the flushing rate of the lake. Typical units are mg l-1for [M], g m-2 of lake surface for L, metres for z and year-1 for s and r. There is generally a high correlation between total P and chlorophyll a, a usual measure of algal biomass, in deep lakes. The amount of algal biomass present in the water is what usually constitutes the eutrophication problem (for clarity of the water, or production of objectionable algal growths). Consequently much management has been directed towards reducing [M] by reducing L. Little can be done to control z and s and increasing the flushing rate is usually impossible as it requires an additional water supply

For deep lakes, the model has been valuable in indicating potentially important processes. The predictions it makes of phosphorus concentrations, however, do not always accord well with reality, because of difficulties in measuring some variables. Sedimentation rate is particularly difficult to measure, as can be loading rates, especially where the loads do not come in steadily but are pulsed in rainstorms and flood events

The model may give sound predictions for total phosphorus concentrations but the expected consequences for algal crops may not be realised. The scatter diagrams (Fig 4a) which appear to show so convincing a relationship between mean total p concentration and summer chlorophyll a [46] rely strongly on the effects of logarithmic transformation to mask variability in the relationship. An arithmetic plot of the same data (Fig 4b) will show a huge range of chlorophyll a concentrations for the same total P concentration. This is because in particular lakes, either phosphorus is not the limiting factor for accumulation of biomass (nitrogen may be) or because, although phosphorus may set a potential upper limit, this is not achieved because some other factor, like grazing of the crop, prevents this happening [111]. In many shallow lakes, one or both these reasons may prevail. Although it is valuable as a comparative basis for thought, the simple assumptions of the Vollenweider model crumble before the complexities of shallow lakes!