Print This Page

Diffusion/Leaching of Nutrients from Shrimp Feeds


Dallas Weaver ( The complexity of shrimp feeds is much greater than most feed millers and nutritionists are willing to admit.  Shrimp are slow feeders, and the feed can be in the water for hours before it is consumed.  Shrimp pellets rapidly absorb water when they hit the pond, and small water-soluble components such a taurine, Vitamin C and ionic salts leach out into the surrounding water.


This leaching (really diffusion) of critical components means that “what you feed shrimp is not what they eat.”  A lot of nutrition researchers think a pellet is consumed before anything leaches out into the water.  Not true.  Even when a pellet is held together with a binder, the leaching happens and is not slowed down very much.


For an interesting experiment on leaching kinetics, fill a beaker with distilled, very pure water.  Add some feed, mix it with the water, and test it with and electrical conductivity meter.  The conductivity, a measure of what ionic components went into the water, will increase.  For a one-millimeter trout pellet, 63% of its ionic components will be in the surrounding water after one minute!  Small soluble non-ionic molecules will diffuse at similar rates.


Controlling this diffusion problem in diets with small particles, like those in replacement diets for rotifers and Artemia or with slow, messy eaters like shrimp, is very difficult.  This is why rotifers, Artemia and live algae along with fresh feeds (for example squid and bloodworms) are used for shrimp maturation and larval rearing.  Live feeds and fresh dead feed (but not dried/frozen dead feeds) contain hydrophobic cell walls that stop diffusion of water-soluble components.  I am reasonably sure the reason live feeds out perform formulated prepared diets in larval culture of all kinds (fish and shrimp) is related to this diffusion phenomenon.


With the increase in fishmeal prices and the shift to other feedstuffs, we could have changed how critical components diffuse out of the diets, creating health issues ranging from vitamins and mineral deficiencies to shortages of things like taurine (a small, fast leaching molecule).  To make things worse, the leaching rates depend upon the history of the ingredients and their formulations.  Like good cooking, the order and mixing details of how the compound diets are created is important.  Adding ingredients in a different order when mixing up the feed batch can make a difference in the effective diffusion kinetics by factors of 10 to 100s.


I have been doing some experiments on this diet diffusion problem with the objective of controlling it in shrimp maturation diets.  I have had some success with abalone diets and am now looking at a synthetic worm for shrimp maturation diets.


Daniel Gruenberg ( Hi Dallas, this is the reason we have moved to more frequent feeding with automatic feeders.  The shrimp grab the pellets more quickly.  With our system almost no feed makes it to the bottom of the pond!


Billy Setio ( Hi Dallas, compared to manually throwing the feed, do automatic feeders minimize the leaching of nutrients from the feed?


I am experimenting with automatic feeders, assuming that once the shrimp get used to their locations, they will find and consume the feed quicker that they do with manual feeding.  Assuming that 30% of the feed goes uneaten with manual feeding, would it be accurate to assume that only 20% of the feed goes uneaten with automatic feeders?


Durwood Dugger (, Dallas, Daniel and others: While I agree that it’s important for shrimp feed to be consumed quickly, so the shrimp get the nutrients before they are diffused into the water column, you are not addressing the physical feeding limitations that prevent shrimp from ingesting whole pellets.  Shrimp have rather small mouths, making it impossible for them to consume whole pellets.  Consequently, pellets are torn apart by the shrimp in both the “tasting/evaluation” and mandibular mastication processes—both of which increase the diffusion of nutrients as more and more of the pellet and its ingredient surfaces are exposed to water.


A partial solution to the leaching of nutrients would involve much finer ingredient grinds and greater ingredient homogenization than commercial feed producers currently use.  Another approach would be to look for ways to physically or chemically prevent leaching through the use of better binding agents.  Theoretically—if not economically—finely ground homogenous ingredient binding is the simplest way to prevent at least some leaching.


Additionally, at least with Penaeus vannamei, you are seemingly forgetting the role that bacteria play in shrimp nutrition.  Feeds quickly pass through the shrimp’s digestive system.  Once excreted, they attract a broad array of nutritious bacteria and are consumed again by the shrimp (autocoprophagy), allowing the shrimp to achieve highly efficient growth rates, even with leaching and poorly constructed and formulated diets.  This is clearly demonstrated in growth and yield achievements in biofloc systems, where feed nutrients are recycled into bacterial flocs that the shrimp consume.  Our goal is to produce more shrimp for less money.  A diet that is nutritionally complete may not reduce our feed costs from the current 50% to 60% of operating costs.


Biofloc systems highlight the role of bacteria and other organisms in P. vannamei feed utilization, which should not be too surprising because vannamei evolved as a highly effective detrital feeder in primarily bacterial-rich environments.


Dallas, what you consider to be “messy feeders” is the shrimp reducing the pellet to a size that it can ingest.  In this process, it also sorts the ingredients (if they are not finely ground, homogenous and well bound) back into their original particulate forms.  Then it selects the most attractive ingredients like fishmeal and the high-lipid components to eat first.  A few hours later, the shrimp will come back to the uneaten ingredients, which will be coated with yummy bacteria that are breaking those ingredients down into their simpler molecules that the shrimp can absorb directly, and consume them.


We often assume that shrimp are more like fish—ingesting whole feed pellets and taking a long time to digest them.  This fish paradigm isn’t the reality of shrimp’s digestive processes.  Feeds pass through a shrimp’s digestive tract very quickly, and digestion is not as complete as it with fish.


Perhaps the questions for vannamei biofloc farmers and shrimp feed companies should be: How simple and inexpensive can the feed be in order to produce good growth rates, and how can pond bacteria be selected to convert the most basic feed ingredients into a nutritionally complete diets?  These seem to be questions that shrimp feed companies have long ignored, or rather not chosen to consider, but they are questions that need to be answered.


Daniel Gruenberg ( Hi Durwood, you bring up some important points.  The vast majority of shrimp feeds are steam pelleted and too big for shrimp to handle efficiently.  When you observe shrimp eating pellets, there is a large amount of waste that just becomes expensive pond fertilizer.


First, I have moved away from steam-pelleted feeds and moved to extruded feeds that behave much differently in the water and the shrimp’s mouth.  Second, I have adjusted the pellet geometry so that the shrimp can ingest the whole thing without masticating it with their mandibles.  And third, I optimized the spatial and temporal feeding pattern so that the shrimp get nearly 100% of the pellets before they reach the pond bottom.


I do not agree with you completely on the benefit of bioflocs.  We have compared many pond cycles with bioflocs versus optimized feeding with NPM (natural pond management with diatoms and copepods).  With NPM, we got farm-wide average FCRs of around 0.8; this was before EMS came to Thailand.


EMS threw a wrench in our plans, but this year we are targeting FCRs even lower than 0.8!  I’ve seen a lot of biofloc systems, but never any with FCRs as low as 0.8, even when pelleted feed is supplemented with flocs.  There is something wrong with the theory that the flocs process waste back into feed for the shrimp.  In a meta study done of many floc systems versus NPM ponds, there was a clear benefit for extruded pellets over steam-pelleted feeds that included flocs.


I am not attempting to say that flocs are bad.  They have many benefits, but something else is obviously going on here other than what the theory says, and I think Durwood may have touched upon it in your last message.  Perhaps steam-pelleted feeds that fall to the bottom of the pond do attract a nutritious flora that replaces the leached nutrients.


You have a negative factor of soluble nutrient leaching that’s balanced out by micro flora that establishes itself on the pellets and assists with the nutrients that the shrimp do finally consume.  In any case, I think it’s quite clear that the feed that comes from the mill and the feed that the shrimp eats are quite different.


At the end of the day, since we all know there are many ways to skin a cat in this business, I don’t want to arrogantly say one way is better or worse.  I think Durwood makes some excellent points.  There are various approaches that we can take to mitigate the problem of leaching.


I think a FCR of 0.5 is going to be possible within the next 3-5 years and that we can lower feed costs, improve the environment and produce more sustainable, high-quality shrimp!


Durwood Dugger (, Hi Daniel, the problem in drawing conclusions from biofloc comparisons is that bioflocs tend to be highly, if not infinitely, variable—not at the beginning of the growout cycle, but over time, regardless of their initial speciation.


Additionally, in my experience, extrusion does not change feed ingredient size; it only provides better ingredient mixing, greater starch plasticity and bonding.


That said, some larval diets tend to be extruded and reground until they are reasonably homogenous nutritionally, but of course, there’s a lot of heat degradation and loss of critical volatiles in the process.  All of our past tests failed to produce performance differences between similar formulations of pelleted and extruded diets in pond growout and/or RAS environments.


I am curious how you can get a one-gram shrimp to ingest a whole pellet given the oral opening is sub-millimeter in diameter (“adjusted geometry” or not)?  Are you producing sub-millimeter pellets?  Or, are you saying that your diet is completely homogenous at a microscopic level and no matter whether a bite or whole pellet, the same nutrition is ingested?


Dallas Weaver ( Durwood and Daniel, I am not sure that “biofloc” is anywhere close to the same stuff with the same “feed value” when the operating details are different.


Different effective sludge ages should have different nutritional values with very young fast growing sludges (aka biofloc) having higher nutrition values and more “mature sludges” having built up higher concentrations of refractor organic waste products from the microbiological ecology of the system.  Fully “stabilized” organic sludges are like mature compost that won’t even grow mushrooms.


Pure bacterial systems do not synthesize vitamin C which makes adding some algae/diatoms to the sludge a requirement to improve the nutritional value.  Steve Serfling, when he was doing his ODAS (organic detrital algal soup) culture systems for tilapia back in the late 1970s and early 1980s, did take some of the high nitrate/phosphate system water and settle/recycle the biofloc and then use that high nutrient water in algae raceways with the algae water being added back to the biofloc.  He was playing these biofloc games in 1,500-tons-a-year systems, long before the thought of biofloc systems or C/N ratios even appeared in the scientific literature.


If we want to recycle nutrients via bioflocs with any efficiency, we probably can’t use chemical energy only (adding carbohydrates) and will probably have to use multiple separate interconnected systems to achieve the nutrient recycling.  More like the partitioned aquaculture systems (PAS) type approaches where we can create and have more control over the various steps.


Once the indigestible refractory organic content is high enough in the biofloc it becomes useless as a feedstuff and the option becomes removing these refractor organics from the system.  To get nutrients to recycle, you need to get the SRT (sludge retention time) low while maintaining the total biofloc concentration fairly high, which can probably only be accomplished at a fairly narrow range of feed loadings in a single system.  In a single pond with small shrimp at low feed rates, you can’t grow enough young sludge, and at very high biomass and feed loadings per area your sludge-wasting rate is too high to maintain slow growing bacteria like nitrifying bacteria (nitrite buildup) and the amount of oxygen and carbohydrate demand becomes excessive.


Eric Muylder (, Hello everybody, there are a lot of misunderstandings about bioflocs.  They are the best way to recycle nutrients.  I agree with Daniel that algae is a better system, since shrimp will utilize the algae as well as the plankton, which has a higher nutritional value when grown on algae.  However, once you go more intensive, algae cannot uptake the surplus nutrients in the water anymore, and the only option is bacterial bioflocs.


It is true that “bioflocs” can reduce the FCR by 30-40 %, but this is compared to clean water (like a RAS system).  Algae is at least as good for reducing your FCRs.  Also, part of the benefit of bioflocs (no changing of water) can probably be attributed to the immediate consumption of the partly digested feces, maybe further digested by some bacteria, or not.


I don’t agree with the idea that we need carbon in biofloc systems.  Nitrifying bacteria can control the ammonia buildup, once you have mature water.  Adding carbon will only increase your electricity cost and buildup of bioflocs, which you will have to remove from the water.


Ramon Macaraig ( Durwood, Dallas and Daniel, our biofloc ponds run on a stocking rate of 150 postlarvae per square meter in 2,000 sqm open ponds.  This seems to point to the fact that a floc system has more effect on stabilizing the pond environment by keeping the dissolved oxygen at 3.8-5.2 parts per million than in providing recycled nutrition via the biomass.  We have harvested the equivalent of 35 metric tons per hectare of 25-gram shrimp and averaged 20 metric tons per hectare of 16-18-gram shrimp!  Good survival rates (70% of what we seed) have made the business quite viable, even at FCRs of 1.3-1.5.


Adding carbon, to which Daniel and Tzachi Samocha have strong opinions against, must have stimulated the bacterial biochemistry to scrub ammonia-nitrogen and stabilized conditions for the shrimp to grow stress-minimized, even when our unlined ponds had whitespot previously.  Once the stress conditions were minimized by the “sludge technology,” the shrimp seem to look for better nutrition.  With the ammonia scrubbed, there were no dissolved oxygen drops in the early morning, and the pH daily fluctuations at alkalinity 140-160 were less than 0.3 unit.  The organic matter is suspended in the floc and does not make underwater mounds in these small ponds.


We did grow copepods and rotifers for the PLs, and the shrimp grew fast, hitting 5 grams in 45 days and 10 grams in 70 days.  We made the mistake of basing the survival on formulated feed consumption and paid dearly for the slower growth thereafter, even if the computed FCRs were only at about 0.8 at that stage.  After the natural foods were consumed or just didn’t meet shrimp’s requirements, the demand for the formulated feeds accelerated.  We had difficulty in developing the floc in this run.


In a previous run, which developed floc early, we were able to get growth rates of 2.0 to 2.5 grams per week when the averaged body weight reached 15 grams.  When the floc developed, the algal composition became immaterial.  The dried floc we analyzed had a proximate analysis of Cprotein (12.07%), Cfat (2.22%), moisture (3.06%), ash (13.8%), fiber (12.43%) and starch (13.6%).  From the feeds that were at 37% Cprotein and 4.7% Cfat, we thought that the shrimp got a good enough share of nutrients from the recycling, described by Durwood, considering that the TAN (total ammonia nitrogen) results were never above 0.5 ppm in the daily samples.


In the next biofloc runs, we will still use natural foods based on algae nutrition, then accelerate growth with formulated feeds at day-45 of growout, targeting development of floc at a density of 8-10 ml at 70 days of culture.  Once the growing fundamentals are there, and only then, we will look at feed compositions intelligently, praying that the stable growing conditions can still shield the shrimp from diseases.  The use of automatic feeders mentioned by Billy is very interesting.


Dallas Weaver ( Billy, frequent feedings from automatic feeders can help a lot in minimizing leaching.


Durwood Dugger (, Dallas, if you have the right bacteria, they will produce vitamin C, and other vitamins are also produced by bacteria.


Unless the biofloc system is operated in absolute darkness, there will be both bacteria and algae in the system, along with their respective capabilities at nutrient production.  As yet, I have not examined any biofloc from a shrimp pond that did not have some portion of the biofloc as algae.  As you point, out biofloc matures, and the stages of maturity have different nutritional processing capabilities, which is why harvesting biofloc from the system is necessary.


As I was saying to Daniel, biofloc is an almost infinite in its species possibilities, as are its capabilities and the opportunities to manipulate its speciation to produce many nutritionally beneficial bi-products.  The random species that develop in a biofloc system represent a very limited perspective of how bioflocs can be used.  It is up to us to manage the various biofloc species in our systems.  Unlike outside environments, closed system/closed environment RAS offers the potential to control, select and manage the systems microbiome.


Dallas Weaver ( Durwood, we can get bacteria to produce almost everything, including vitamins, but when they don’t need it for their own purposes and metabolism, the ecology in an open system usually won’t make it.


With very highly loaded systems, I have seen floc conditions where I can’t detect any algae with my scope (but it is probably there at such low concentrations as to be nutritionally irrelevant).  That is probably why Steve Serfling added a separate algae system to up the concentration.


Victor Suresh( Water-soluble, small molecular compounds like free amino acids, short-chain peptides, nucleotides, biogenic amines and other nitrogenous compounds in feed signal the chemosensory organs of shrimp to identify, approach and consume the feed.  If a feed does not leach out, it will not be recognized as food.  When you reduce the leaching rate, you increase the time for the feed to be found, so the amount of nutrients leached will perhaps remain the same.  One solution: you could tightly bind the feed to reduce leaching, but apply attractants on the surface to provide attractability.


Feeding the shrimp with bite-sized particles is an interesting idea.


Automatic feeders are another solution, but I don’t recall any drastic reduction in FCR from using them.


Dan Fegan (owner and creator of The Shrimp List, Eric, back in the early days of Frippak, we measured leaching in larval and some growout feeds.  Around 10% of soluble protein was lost in the first minute after putting in water.  Of course technology is probably better nowadays, but it really brought home that formulated levels of nutrients may bear little resemblance to the composition of the feed as eaten by the shrimp.


Everyone, the nutritional value, final product condition and water stability of shrimp (or any aquaculture feed) is, in the end, a product of a number of considerations and compromises.


Nutrient levels in ingredients, digestibility, ingredient costs, ingredient quality and ingredient availability determine the raw material set that is available to the formulator to provide the nutrients required.  I have never understood the obsession with crude protein, which in the end is actually just a measure of nitrogen (remember melamine anyone).  It says nothing about feed quality, protein content, protein (and amino acid) digestibility or suitability for shrimp.  However, that hasn’t stopped people and governments putting minimum levels on crude protein content under the fiction that it somehow guarantees quality or protects farmers from sub-standard feeds.


The characteristics of these ingredients for grinding and feed production, their consistency, the equipment in the mill, quality assurance processes and, most underestimated, the skill of the pellet mill (or extruder) operator will all have an impact on the physical properties of the pellets produced.


All of this then has to be set against the price that farmers are willing to pay to buy the feed.  It is an interesting exercise to work back from the feed price and subtract, for example, commissions, profit margins, production and other costs to come to a figure for the budget for ingredients.  This can be quite a surprise and, since ingredient prices rise more quickly and frequently than feed prices, you get some appreciation for the challenges faced by the poor nutritionists and formulators.  It is neither an easy nor a thankful task trying to balance all of the different expectations of the business, the sales team and the customers in a commercial feed business.  Move too far in any one direction and somebody else gets upset.


This is even without taking into consideration the large impact of on-farm variables that affect feeding efficiency, most of which tend to be ignored.  Even on a well managed, successful farm, there are many factors outside of the control of the farmer that will affect feeding efficiency.  However, in my experience, these tend to be ignored when discussing feed management and efficiency.  I’ve even done it myself when managing a farm as putting a bit of pressure on the feed salesman often results in a better deal even when I know the problem wasn’t really the feed.


Sources: 1. The Shrimp List (a mailing list for shrimp farmers).  Subject: Daniel; Feeds and Diffusion.  June 18 to 21, 2017.  2. Bob Rosenberry, Shrimp News International, June 21, 2017.

Print This Page