
Fish pond volume: measure the average water depth
A 1,000 m² fish pond that is 2 m deep beside the outlet does not necessarily hold 2,000 m³ of water. The useful estimate depends on average depth across the pond and the current water level. FAO guidance offers a practical starting point for recording the volume and its uncertainty.
- Content type
- Practical guide
- Sector
- Aquaculture
- Animal group
- Fish
- Keywords
- OxygenTemperatureStress
The deepest point is a poor substitute for average depth
A depth reading beside a pond outlet is often deeper than most of the pond. Applying that measurement to the entire surface effectively assumes vertical walls and a flat bottom at the outlet level. Production ponds commonly have sloping margins, uneven bottoms and areas of accumulated sediment. An old capacity figure can therefore become unreliable even when the outer boundary has not moved.
Coche and van der Wal’s 1981 FAO manual sets out a geometric approach: establish the water surface area, estimate a representative average depth and multiply the two. This practical guidance concerns water volume. It does not establish a universal biological carrying capacity or a stocking limit for farmed fish.
Farm teams need a volume estimate whose basis can be checked. Recording the water level, survey layout and measurement date makes the estimate easier to check and reuse. A highly specific capacity entered in a spreadsheet is much less helpful if nobody knows which water level or survey it represents.
Give the whole pond a place in the survey
Bathymetry describes the depth and shape of the submerged bottom. A simple survey still needs measurements that represent the pond’s different areas. Taking several convenient readings close to an accessible bank can miss deeper water or silty sections. The water surface also needs to be distinguished from the total land area shown on a property plan.
FAO varies the survey with the geometry: a few distributed readings for a small, regular pond with a consistent slope, more readings for a larger pond, and a grid for irregular bottoms. The transferable principle is spatial coverage. Copying a particular number of survey points without checking the bottom shape does not make the resulting mean representative.
On a regular grid, the areas represented by points are broadly comparable. Where sections have different areas, their mean depths need to be weighted accordingly. A narrow deep sump should not contribute as much as a large shallow margin merely because each has one reading. Unsurveyed sections should remain visible in the record rather than silently inheriting a nearby measurement.
The vertical reference matters too. Water depth is measured from the bottom to the water surface at the time of the survey. Recording that level allows later surveys to be compared. Access and working arrangements must suit the site; an older manual’s suggestion to enter a pond is not, by itself, an adequate assessment of safe access.
A worked example shows the consequence
Consider an illustrative pond rather than a reported farm case. Its wetted area is 1,000 m² and its properly estimated average depth is 1.2 m. The resulting volume is 1,200 m³, equivalent to 1,200,000 litres. If its maximum depth of 2 m were used instead, the result would be 2,000 m³: an excess of 800 m³, or approximately 66.7% above the average-depth estimate.
This example does not quantify instrument accuracy. It isolates the effect of choosing the wrong depth. Uncertainty in area, missing survey points and the water-level reading still needs consideration. Adding decimal places would not make those uncertainties disappear.
If fish biomass has been estimated separately, the volume error carries into a stocking density expressed in kg/m³. Keeping the fish mass unchanged while reporting 2,000 m³ makes that density appear lower than when using 1,200 m³. The ratio describes mass per unit volume; it does not establish whether the environment meets the animals’ needs.
Keep the estimate tied to the operating level
A survey made at the highest operating level cannot automatically describe a pond after drawdown. Sloping banks also change the wetted area. For ponds operated at several levels, a documented level–volume relationship avoids treating every fall in water height as a uniform layer removed from an unchanging surface.
A practical record can retain the point layout, raw depth readings, averaging method, water level and excluded areas. A separate units check helps catch centimetre-to-metre or cubic-metre-to-litre mistakes. After dredging or structural changes, the team should revisit whether the old calculation still represents the bottom.
Daily management requires additional observations. Dissolved oxygen, temperature, feeding and fish behaviour cannot be calculated from pond volume. Conditions can differ within a large pond, and a total volume does not show that all of its water is equally suitable for the stock. A falling level accompanied by altered behaviour warrants assessment of both the water and the animals.
Make the limits easy to find
The record should state the intended use of the estimate. A standing volume is different from a measured inflow or outflow: flow describes the amount of water passing a point over time. Both may be needed for management, but one cannot replace the other.
The practical decision is to use representative average depth at the actual water level and retain enough information to repeat the calculation. An explicit uncertainty helps identify when a fresh survey is needed. It also prevents an old nominal capacity from being treated as a current measurement.
Vetofish can help plan a survey, check the consistency of units and interpret the volume alongside health observations through its advisory and support service, tailored to aquaculture operations.


