Fertiliser runoff and a reservoir fish kill
A reservoir in Malta's rural north supplies irrigation and is a backup drinking-water source. Upstream, a farmer applies 120 kg of nitrogen fertiliser per hectare to a field. After 35 mm of rain in two days, nitrate in the reservoir rises from 4 mg/L to 18 mg/L. Within ten days, chlorophyll a doubles, daytime pH reaches 9.2, dissolved oxygen is 12 mg/L at the surface but 2 mg/L at 4 m depth. Dead fish are found near the dam. The water utility can switch to a different source for 14 days at an extra cost of €18,000; after that it must use the reservoir or impose restrictions. The farmer says the algae were already present before the rain and that the fertiliser is needed to keep the farm viable. The regulator asks for a decision within 48 hours.
1. Explain, using the data, how the fertiliser could lead to low oxygen at 4 m even though surface oxygen is high.
Read the worked answer
The sequence is: fertiliser nitrate is washed into the reservoir after 35 mm of rain; nitrate rises from 4 to 18 mg/L; algae use the nutrient and grow, so chlorophyll a doubles. In daylight, photosynthesis raises pH to 9.2 and produces oxygen at the surface, which is why surface oxygen is 12 mg/L. When algae die, they sink. Bacteria decompose them and consume oxygen. If the water is stratified, oxygen at 4 m is not replaced from the surface, so it falls to 2 mg/L. Fish need more than that and die. The high surface oxygen therefore does not contradict the low deep oxygen; it is part of the same bloom. The data are consistent with eutrophication, but they are still correlation. I would check whether the algae were already present, whether other nutrient sources exist, and whether the deep low oxygen occurred before the rain.
2. Should the utility switch sources now? Give a decision and the evidence you would collect in the next 48 hours.
Read the worked answer
I would switch now if the reservoir is a drinking-water source and the treatment plant cannot reliably remove nitrate or algal toxins. Nitrate at 18 mg/L is a concern for drinking water, and the fish kill shows the ecosystem is already stressed; toxins from a bloom can pass through some filters. The extra €18,000 is significant, but it buys 14 days to test. In the next 48 hours I would measure nitrate at the intake, identify the algae and test for microcystin, profile oxygen and temperature with depth, check flow and stratification, and inspect upstream fertiliser and septic sources. If the reservoir is only for irrigation, the switch is less urgent, but the fish kill still requires action. The decision should be precautionary for drinking water and proportionate for irrigation. A 14-day switch is reversible, which is an advantage.
3. The farmer offers a 30% fertiliser cut next season. Evaluate whether this is a sufficient response, and suggest one additional measure with a trade-off.
Read the worked answer
A 30% cut is a start but probably not sufficient on its own. The bloom is already in the reservoir, and nutrients can be released from sediment, so the response will lag. The farm also needs soil testing, correct timing, a buffer strip along the watercourse, and a cover crop to reduce erosion. A 30% cut may be enough if the field is the main source and if it is combined with these measures; it is not enough if other farms or septic tanks also contribute. The trade-off is yield and income against water quality. I would make the cut conditional on monitoring and review it after one season. Regulation may be needed if voluntary cuts are not met, but a voluntary plan with advice and cost-sharing is more likely to be accepted. The decision should be based on the whole catchment, not one field.