How water temperature and dissolved oxygen change feeding

Fish take on the temperature of the water around them. Appetite, gut transit, growth and oxygen demand all move with it, and each species has a band where these run at their best inside a wider band it can tolerate. Outside the optimum, performance falls off quickly, and a ration that converted well one month can end up uneaten the next.

Dissolved oxygen sets a second ceiling. Digesting a meal and turning it into tissue costs oxygen, so fish that are short of it eat less, and feed offered beyond that reduced appetite shows up as waste and a worse FCR. A ration fixed as a percentage of biomass, set without reading temperature and oxygen, overfeeds exactly when fish can use feed least.

Temperature sets the pace of feeding

Because fish are ectotherms, their metabolic rate follows water temperature. A review in the Global Seafood Alliance's Responsible Seafood Advocate notes that oxygen consumption roughly doubles with a 10 C rise within a species' tolerance range. Appetite climbs with temperature toward an optimum and then falls sharply. FAO's manual on small-scale rainbow trout farming puts trout appetite at its best between about 7 and 18 C, with intake dropping steeply above 18 C until feeding stops.

Intake and feed efficiency do not peak at the same temperature. The same FAO manual warns that trout near 18 C feed very intensively but digest the feed less completely, and places the best growth from consumed feed at roughly 13 to 15 C. In a tank trial with Atlantic salmon post-smolts, maximum daily intake rose from 0.47% of biomass at 7 C to 0.88% at 19 C when oxygen was not limiting. A ration sized for warm water overfeeds fish in cold water, and near the top of the range extra intake does not buy extra growth.

Each species has its own band

FeedOptima's reference data puts the typical optimum for Atlantic salmon at 8 to 14 C, within a tolerable range of about 2 to 20 C. Nile tilapia sit at the opposite end, with a typical optimum of 26 to 30 C and a tolerable range of about 16 to 35 C. Ontario's aquaculture species profile for tilapia describes the best results near 28 C, loss of appetite and growth below 20 C, no feeding below 16 C and possible mortality below 12 C. It also notes that tilapia in northern climates such as Canada are grown in insulated, heated recirculating systems.

The table gives FeedOptima's typical ranges for three species farmed in Canada.

Species	Typical optimum	Tolerable range	DO target	DO minimum
Atlantic salmon	8 to 14 C	2 to 20 C	8 mg/L	7 mg/L
Rainbow trout	12 to 17 C	1 to 24 C	7.5 mg/L	5.5 mg/L
Nile tilapia	26 to 30 C	16 to 35 C	5 mg/L	3 mg/L

The chart shows FeedOptima's temperature adjustment for salmon and tilapia. Inside each optimum band the factor holds at 1.0, meaning no penalty; outside it the factor falls steadily toward about 0.6 at the edge of the tolerable range and drops to 0.5 beyond. The two optimum bands do not overlap, so a feeding table built for one species is no guide to the other.

Oxygen is the price of digestion

Every meal raises oxygen demand. FAO's trout manual notes that oxygen consumption increases considerably during and after feeding, and the Southern Regional Aquaculture Center (SRAC) guide to measuring dissolved oxygen observes that DO falls shortly after fish are fed, even in well-mixed, aerated tanks. In an example given in the Advocate review, channel catfish used 680 mg of oxygen per kg per hour 1 hour after feeding, compared with 380 mg/kg/h after an overnight fast. This post-meal rise, known as specific dynamic action, is the oxygen cost of digesting food and building tissue from it.

When oxygen runs short, fish cut intake first. In the post-smolt trial, appetite held steady as DO fell until a threshold, then declined; that threshold rose from 42% of air saturation at 7 C to 76% at 19 C. A Journal of Fish Biology study on juvenile Atlantic salmon found that acute hypoxia at 50% air saturation roughly halved aerobic scope, that digestion claimed a larger share of that scope at 21 C than at 15 C, and that gut transit slowed when digesting fish met low oxygen. Put together, warmer water raises the oxygen level needed for full appetite, and a meal eaten in low oxygen is processed more slowly.

Low oxygen also worsens FCR through overfeeding. A second Advocate article, on pond oxygen, summarizes channel catfish studies in which survival, production and FCR were better in ponds where the average daily minimum DO did not fall below 3.5 mg/L. The fish were fed from a feeding table with no adjustment for poor appetite, and part of the FCR penalty came from feed the oxygen-limited fish did not eat.

Warm water holds less oxygen

Oxygen solubility falls as water warms. The solubility table in the Advocate review gives about 11.3 mg/L at saturation in freshwater at 10 C, 9.1 mg/L at 20 C and 7.8 mg/L at 28 C, at sea-level pressure. Across the step from 10 to 20 C, the fish's oxygen demand roughly doubles while the water's capacity to hold oxygen falls by about 20%. Research on farmed salmon describes the result: periods of warm, low-oxygen water in summer and autumn, accompanied by a drop in feed intake.

Species differ in how much oxygen they need to feed well. The Advocate review explains that warm-water fish typically unload oxygen from hemoglobin to their tissues more readily than cold-water fish, a major reason cold-water species need higher dissolved oxygen. FeedOptima's typical DO target of 8 mg/L for Atlantic salmon is well above the 5 mg/L used for Nile tilapia. Tilapia tolerate lower oxygen but live in water that holds less, and at 28 C a 5 mg/L target already sits at about 64% of saturation.

The oxygen chart shows FeedOptima's adjustment for the same two species. Below each species' minimum the factor sits at 0.5, and between the minimum and the target it rises until feeding performance is unconstrained at 1.0. Salmon need far more dissolved oxygen than tilapia before that point, so a reading that is comfortable in a tilapia system can be a feeding limit in a salmon tank or cage.

Oxygen also moves through the day and across the unit. SRAC notes that pond DO is generally lowest at dawn and near the bottom and highest at dusk, with critically low levels in warm months usually occurring at night; it also describes pronounced gradients along raceways. The Advocate review recommends checking oxygen especially between midnight and dawn. A single morning reading at the inlet says little about the oxygen fish have while a meal is being digested.

In practice

1. Record water temperature at feeding depth daily and set the ration from a feeding table indexed to temperature and fish size, not a fixed percentage of biomass.
2. Measure DO at the tank outlet, the downstream end of a raceway or inside the pen during and just after the main meal, not only at the inlet.
3. In ponds, log DO at dawn and dusk to find the daily low and high, check between midnight and dawn in warm months, and keep meals away from the overnight low.
4. When DO at feeding is below the species target, cut meal size; below the minimum, hold feed and restore the ration in steps as oxygen recovers.
5. In warm water, raise aeration or supplemental oxygen before the meal so the post-feeding drop stays above target.
6. As temperature rises, split the ration into more, smaller meals, as FAO recommends for trout, to spread the oxygen demand of digestion across the day.
7. Treat a sudden loss of appetite as a prompt to check oxygen and fish health, and record uneaten feed at every meal.

To see how your own temperature and oxygen readings change expected feed performance for your species, run an optimization with FeedOptima's free analysis.