Eutrophication: How Fertilizer Suffocates a Lake

The Redfield ratio: why phosphorus is the master switch

In 1934 the oceanographer Alfred Redfield noticed that marine plankton, and the seawater they grow in, share a near-constant elemental recipe: 106 carbon : 16 nitrogen : 1 phosphorus, by atoms. Written as an idealized photosynthesis-plus-assimilation equation, building one "unit" of algal biomass and releasing oxygen looks like this:

• 106 CO₂ + 16 NO₃⁻ + HPO₄²⁻ + 122 H₂O + 18 H⁺ → C₁₀₆H₂₆₃O₁₁₀N₁₆P + 138 O₂

The ratio is the key to eutrophication. Because algae need N and P in a fixed 16:1 proportion, whichever nutrient runs out first caps growth — this is Liebig's law of the minimum. In most freshwater lakes that limiting nutrient is phosphorus, because many cyanobacteria can fix atmospheric N₂ to cover a nitrogen shortfall, but nothing fixes phosphorus from air. So a tiny addition of P — a few hundredths of a mg/L — is leveraged 106-fold into carbon fixed as new algae. In coastal marine systems the limit usually swings to nitrogen, because phosphate is resupplied from sediments and N is more readily lost. Manage the limiting nutrient and you manage the bloom; this single insight underlies every phosphate-detergent ban since the 1970s.

From fertilizer bag to bloom: the phosphorus pathway

Roughly half the reactive nitrogen and much of the phosphorus loading the world's waters traces back to the Haber–Bosch process and mined phosphate rock. Applied to fields as ammonium phosphate or superphosphate, the surplus that crops don't take up moves two ways. Nitrate (NO₃⁻) is highly soluble and mobile, so it leaches through soil into groundwater and tile drains. Phosphate behaves differently: the orthophosphate ion (as H₂PO₄⁻ / HPO₄²⁻ near neutral pH; note phosphoric acid pKa values 2.15, 7.20, 12.35) binds tightly to soil minerals and rides into streams stuck to eroding particles.

In the water column, phosphate partitions between dissolved and mineral-bound forms governed by solubility equilibria. Two reservoirs matter:

• Iron binding: under oxygenated conditions, phosphate adsorbs to and co-precipitates with ferric oxyhydroxide, FeOOH·(PO₄), locking P into sediment.
• Calcium binding: in hard, alkaline lakes it precipitates as hydroxyapatite, Ca₅(PO₄)₃OH, a sparingly soluble mineral.

As long as P stays mineral-bound, the lake stays clear. Eutrophication is fundamentally the story of that bound phosphorus being released — and, crucially, the release is self-amplifying.

Internal loading: the iron redox trap that keeps a lake sick

Here is the vicious cycle that makes eutrophication so stubborn. A bloom grows, dies, and sinks. Bacteria decompose the dead algae, consuming oxygen. When bottom-water dissolved oxygen falls below about 2 mg/L (hypoxia) and then to zero (anoxia), the sediment chemistry changes state. The controlling variable is redox potential (Eh).

Under oxygen, iron sits as insoluble ferric iron, Fe(III), holding phosphate captive. Once anoxia sets in, bacteria use Fe(III) as an electron acceptor and reduce it:

• Reduction: Fe(OH)₃ + 3 H⁺ + e⁻ → Fe²⁺ + 3 H₂O
• Consequence: the Fe(III)–PO₄ complex dissolves, so bound phosphate is set free: FeOOH·PO₄ + e⁻ + H⁺ → Fe²⁺ + PO₄³⁻ + H₂O (schematic)

The liberated phosphate diffuses up out of the sediment and fertilizes the next bloom — internal loading. This is why a lake can keep blooming for decades even after farmers stop adding P: the sediment holds a "legacy" phosphorus bank, and every anoxic summer withdraws from it. In many restored lakes, internal loading supplies more phosphorus than the entire inflowing river. Breaking the loop requires either re-oxygenating the hypolimnion or dosing the lake with iron, aluminum (alum, Al₂(SO₄)₃), or lanthanum-modified clay to re-lock the phosphate permanently.

How the water runs out of oxygen

The lethal step is the oxygen budget, and it is pure stoichiometry. Photosynthesis by the bloom produces O₂ in the sunlit surface — often to supersaturation, > 150%. But respiration and decay consume it, and the reverse of the Redfield equation shows the scale of the debt:

• C₁₀₆H₂₆₃O₁₁₀N₁₆P + 138 O₂ → 106 CO₂ + 16 NO₃⁻ + HPO₄²⁻ + 122 H₂O + 18 H⁺

Every mole of algal carbon oxidized burns roughly 1.3 mol of O₂. Because the dead cells sink, that oxygen demand is dumped into the cold, dense bottom layer (the hypolimnion), which is cut off from the atmosphere by thermal stratification. Warm surface water floats on cold deep water; the two don't mix through summer, so the deep layer cannot recharge its oxygen. Henry's-law solubility is also working against the lake — O₂ solubility falls from ~9 mg/L at 20 °C to ~7 mg/L at 30 °C, so warmer water starts with less oxygen to spend. Add the decay debt on top, and dissolved oxygen crashes toward zero. Fish that need > 4–5 mg/L flee or die; the bottom becomes a dead zone. When even nitrate and sulfate are exhausted, anaerobic microbes turn to methanogenesis and sulfate reduction, generating CH₄ and toxic H₂S (that rotten-egg smell of a dying pond).

The scale: from farm ditches to the Gulf of Mexico

Eutrophication is not a local nuisance — it is a planetary-scale rerouting of two element cycles. Humans now fix more nitrogen industrially (~120 Tg N/yr via Haber–Bosch) than all natural terrestrial processes combined, and we mine roughly 20 Tg P/yr of phosphate rock. A large fraction ends up in water.

• The Gulf of Mexico dead zone: nitrate carried by the Mississippi River — draining 41% of the contiguous U.S. — creates a summer hypoxic zone averaging 15,000–22,000 km² (the size of New Jersey), with a 5-year goal of shrinking it to 5,000 km².
• Lake Erie: declared "dead" in the 1970s, rescued by phosphate-detergent bans, then re-eutrophied after 1995 as dissolved reactive phosphorus from no-till agriculture climbed again — the 2011 bloom covered 5,000 km².
• The Baltic Sea: one of the largest human-caused hypoxic areas on Earth, up to ~70,000 km² of seafloor with < 2 mg/L O₂.

Worldwide, more than 400 coastal dead zones have been documented, collectively covering > 245,000 km². The economic cost in the U.S. alone — lost fisheries, tourism, and water treatment — runs into billions of dollars per year.

The toxins and the fix

Not all algae are equal. When phosphorus is abundant but the water is warm and calm, cyanobacteria (blue-green algae) outcompete diatoms and green algae — they regulate buoyancy with gas vesicles, fix their own nitrogen, and many produce cyanotoxins. Microcystis makes microcystin (a cyclic heptapeptide that inhibits protein phosphatases PP1/PP2A in the liver); Anabaena and Aphanizomenon can make the neurotoxins anatoxin-a and saxitoxin. This is what forced Toledo's shutdown and what makes eutrophication a public-health problem, not just an ecological one.

The chemistry also points to the cures, and they are all about the limiting nutrient:

• Cut external P: buffer strips, cover crops, and reduced fertilizer application — controlling P at the source has driven the recovery of hundreds of lakes.
• Lock internal P: alum or lanthanum-clay dosing precipitates sediment phosphate as insoluble AlPO₄ or LaPO₄ (Ksp ~10⁻²²), starving future blooms.
• Re-oxygenate: hypolimnetic aeration keeps Eh high enough to hold iron as Fe(III), re-arming the phosphate trap.

The encouraging news is that eutrophication is partly reversible — Lake Washington in Seattle went from near-collapse in the 1960s back to Secchi depths of ~7 m within about a decade once sewage phosphorus was diverted. But the sediment legacy means recovery is measured in decades, not seasons, and stopping the fertilizer flow is always cheaper than cleaning up the bloom.

Oligotrophic (nutrient-poor) versus eutrophic (nutrient-rich) water bodies

Property	Oligotrophic lake	Eutrophic lake
Total phosphorus	< 0.010 mg P/L	> 0.030 mg P/L
Chlorophyll-a	< 2.5 µg/L	> 8 µg/L (blooms > 100)
Secchi depth (clarity)	> 4 m	< 2 m
Bottom-water O₂	Saturated (~8–10 mg/L)	Hypoxic/anoxic (< 2 mg/L)
Dominant algae	Diatoms, green algae	Cyanobacteria (Microcystis, Anabaena)
Sediment P behavior	Bound to Fe(III), retained	Released as Fe(III)→Fe(II) reduces