Measuring Oxygen Demand

Knowing how much oxygen is consumed per unit volume of water or area of sediment in a lake is important to understanding lake metabolism and in planning for the provision of adequate oxygen. Systems that mix the lake or add air or oxygen must counteract the oxygen demand to be successful. Measuring demand is tricky, however, since the rate at which oxygen is removed is hard to isolate from factors that put oxygen into the water or allow it to move within water, and consumption of oxygen is not linear over the range from about 2 mg/L down to 0 mg/L.

Laboratory tests can be run in which oxygen loss over a set amount of time is measured in a closed bottle at a standard temperature, and this is useful, but this won’t include sediment oxygen demand, which is often the dominant source. For deep lakes, one can use temperature and dissolved oxygen profiles from a few weeks apart at a time when oxygen is >2 mg/L everywhere to get a reasonable estimate. Oxygen addition from the atmosphere and downward mixing necessitates using values from deeper water, usually below the depth of the thermocline, even if it has not strongly formed yet. Timing is critical. In New England stratification sets up from April into June and is highly weather dependent, but once it sets up, loss of oxygen near the bottom can be rapid and oxygen concentrations can become too low to use in demand calculations. When the oxygen concentration is less than about 2 mg/L, further removal slows down and calculations using low values will underestimate actual demand, something to be avoided when planning oxygen additions.

For proper data, subtraction of later oxygen from earlier oxygen at each depth increment can be summed to yield the mass of oxygen lost over a square meter of lake over time (Table 1). If the temperature has risen between measurements, the water will naturally hold less oxygen, so a correction for temperature-induced oxygen loss must be applied, based on the difference in saturation concentration at the earlier and later temperature measurements. In the example in Table 1, the thermocline forms at about 6 m but was very weak at the time of the April measurements. Yet it was apparent that little oxygen from above was reaching deeper. There was a slight increase in temperature, so the differences first obtained were adjusted down slightly. The difference in total oxygen demand is not large, but can be if the time between measurements is longer.

Table 1. Calculation of Oxygen Demand in a Deep Lake

In shallow lakes it is harder to find a time or place where factors other than oxygen demand are not significantly influential, but we have found one approach that often works.  Photosynthesis ceases overnight in the absence of light, and at least some nights there is calm that limits mixing. Measurements made after dusk and before dawn can be used if oxygen remains >2 mg/L throughout the water column, and the results can be quite striking. Oxygen demand from respiring algae or vascular plants will add to demand from decay, but that is all part of the demand and worth including. The same calculation approach used for the deeper lake is applied to the shallow lake (Table 2), but the whole water column is included.

Table 2. Calculation of Oxygen Demand in a Shallow Lake

For reference, oxygen demand higher than about 0.5 g/m2/d can eventually lead to oxygen depletion in stratified lakes. Values >1.0 g/m2/d will cause depletion in the bottom layer by mid-summer, and values >2.0 g/m2/d will cause oxygen depletion much earlier in summer. Making multiple measurements is recommended to characterize the range of oxygen demand and get a sense for possible error induced by not being able to control all oxygen inputs. With enough measures a pattern usually emerges that allow one to make a fairly accurate estimate of oxygen demand.