What and how much should Neocaridina eat?

Evidence review: August 12, 2026

There is no controlled Neocaridina davidi study that turns colony count into a universal number of pellets, feedings per week or hours before removal. Tank surface area, biofilm production, animal size, reproduction, temperature, food water stability and other grazers all change the demand. A schedule copied without those conditions is false precision.

Biofilm can be a diet, but it is not one ingredient

In a purpose-built zero-water-exchange culture system, biofilm made mainly of microalgae, diatoms, cyanobacteria and ciliates was the sole diet. Shrimp completed the life cycle, while juvenile survival, biomass and biochemical reserves were higher with two of the tested substrate treatments.¹ This supports extensive conditioned surfaces. It does not show that every home aquarium film has the same organisms, quantity or nutritional value. Use the habitat guide to separate conditioned surface, shelter and substrate chemistry.

Seeing more worms or microcrustaceans after feeding establishes an observation sequence, not a species identity or an overfeeding diagnosis. Record the organism and a comparable count trend before changing food or treating the tank.

Biofilm and prepared food are not competing religions. A home colony can graze surfaces continuously while receiving a measured formulated diet. First-stage juveniles are not proven to eat only biofilm.

Surviving without food is not thriving

In one laboratory study, estimated point-of-no-return values were 16.15 days for first-stage juveniles and 9.44 days for third-stage juveniles. Longer restriction beginning at the first stage reduced growth and survival.² These point-of-no-return estimates are study endpoints, not feeding intervals.

Adults also survived long laboratory fasts, but their midgut cells changed during starvation and recovery took longer after the longer restriction. The reported adult PNR50 was 24.72 days.³ These are nutritional-vulnerability experiments, not husbandry recommendations. They refute both "missing one feeding is fatal" and "underfeeding is harmless."

A separate adult experiment suppressed algae, placed shrimp individually in 250 mL containers and withheld feed for 14 days. The reported mean share of ROS-positive cells rose from 2.8 to 13.2 percent in the hepatopancreas and from 1.3 to 12.7 percent in the intestine. After 7 to 14 days of re-feeding, both measures declined toward the control values.²⁰

That is evidence of a cell-level oxidative-stress response under forced deprivation, not evidence that a routine fast improves health. The study tested one 14-day restriction, not a weekly fast, and did not measure a survival, growth, molting, behaviour, reproduction or longevity benefit. Its isolated, algae-suppressed containers do not reproduce a conditioned community tank, the assay sample mapping is incomplete and the MnSOD treatment comparison was qualitative. Missing one normal feeding is not the experiment, and the paper supplies no fasting or re-feeding recipe.

A linked 2019 midgut study used the same isolated, algae-suppressed deprivation procedure with 7, 14 and 21 day groups. After 14 days, the authors described reduced reserve material and organelles in digestive cells; after re-feeding, tissue structure and selected cellular markers moved toward the earlier non-starved reference.²²

That paper is not an independent confirmation or a recovery schedule. It imported its non-starved controls from a 2016 paper, did not report the re-feeding amount or frequency, and said most 21 day animals died without supplying mortality counts or a survival analysis. Its animal denominators conflict across tables, its statistical comparisons are underspecified, and the apoptosis table contains impossible zero means with positive standard deviations. Use the paper to understand prolonged deprivation as a cellular stress model, not to prescribe fasting or compensatory feeding.

A daily RNA pattern is not a best feeding hour

A 2024 laboratory series sampled hepatopancreas tissue at 06:00, 12:00, 18:00 and 24:00 after food had been withheld for at least 48 hours. The authors found time-associated RNA differences and annotated several putative clock-related transcripts.²⁴ The sequencing comparison used only ten pooled libraries, distributed 2, 3, 3 and 2 across the four times. Two additional intended libraries were not explained. Only Clk differed significantly among the four qRT-PCR targets.

Time was not the only changing condition. Sampling also followed the light and dark cycle, a reported daily temperature swing from about 10 to 18 C, and a progressively longer fast. There was no feeding-time treatment, constant- temperature comparison, shifted light cycle, constant darkness or second daily cycle. The pathway labels did not measure food intake, digestion, growth, molting, reproduction or health. This study cannot identify a best feeding hour, prove that night feeding is better, require feeding at lights-on or lights-off, or justify withholding food before a normal meal.

A digestive enzyme is not a feeding schedule

A 2026 molecular study found the highest NdTryp expression in the hepatopancreas and detected the transcript from late embryonic development into the larval stage.¹⁰ That supports a role for trypsin in the studied digestive biology. It does not establish a dietary protein percentage, pellet count, hatchling feeding time or home-tank digestive pH and temperature. The separate recombinant-protein assays do not show that trypsin is a safe or useful feed additive because no feeding, digestibility, growth or dose trial tested that proposal.

A separate 2025 study measured amylase, lipase and protease activity from ten-animal whole-body homogenates at four juvenile ages.¹⁶ Three pooled samples represented each checkpoint, but all animals came from one rearing tank and whole-body activity does not isolate the digestive tract. Reported protease activity increased from day 15 through day 60 under one twice-daily formulated-food schedule.

Age changed at the same time as body size, tank history and sampling date. The study assigned no diet, protein level or feeding-frequency treatment, and it did not report the food amount or intake. It therefore cannot show that juveniles need a high-protein food, prescribe a food for an age, set a feeding schedule or prove that one enzyme pattern caused the observed growth.

Experimental hormones are not feed additives

A 2025 study assigned 120 one-week-old N. denticulata juveniles to a basal diet or feeds formulated with 0.4, 4 or 40 micrograms of methyl farnesoate per kilogram for 40 days. The paper reports concentration- and time-associated differences in length growth plus digestive-enzyme, oxidative-stress and gene-expression outcomes.¹³ It describes four diet groups but does not report independent replicate tanks. Therefore the diet treatment may be confounded with its group or culture system, even though many individual shrimp were present.

The hormone stock was prepared in ethanol, but the control is not explicitly described as receiving matched ethanol. Finished-feed concentration, stability, leaching, consumption and internal dose were not verified. The study also did not establish survival, reproduction, withdrawal effects or long-term safety. Its low feed concentration is not a home recipe, supplement recommendation, water dose or growth-product claim.

A separate 2025 experiment exposed excised hepatopancreas tissue from mature female N. davidi to farnesoic acid or methyl farnesoate for three hours and found different gene-expression responses.¹⁴ Each sequencing sample pooled tissue from three females, with three pooled samples per condition. The work did not expose a living shrimp through food or water, and control timing and hormone-vehicle details are not fully clear. Gene-expression and pathway labels do not prove better digestion, immunity, growth, molting or breeding.

A manganese response curve is not a tank dose

A 30-day experiment assigned 900 wild-collected N. heteropoda to six complete diets, with three tanks and 50 shrimp per tank for each diet. The measured manganese concentrations were 2.79, 22.57, 42.55, 62.69, 82.59 and 102.45 micrograms per gram of finished feed.²¹ The 62.69-microgram-per-gram diet had the highest reported length and weight gain percentages, the highest measured SOD, GPX and CAT activity and the lowest superoxide-anion assay value under that protocol.

This does not establish 60 micrograms per gram as a universal requirement. The control feed already contained measured manganese, its mineral premix listed manganese sulfate, and all tanks shared one recirculating-water reservoir. The paper did not report survival, feed conversion, consumed ration, tissue manganese, deficiency signs, molting, reproduction or long-term safety. It also does not explain how the six biochemical samples and pooled hemolymph map back to the three tanks per diet, and it reports conflicting units for water manganese. Use the result as one complete-formulation comparison, not a reason to dose tank water, add powder to finished food or treat enzyme activity as proof of better health.

Eating snail tissue does not make it a complete or safe food

A 56-day undergraduate thesis placed 42 one-week-old shrimp individually in 600 mL beakers and assigned 14 to excess daily pellets, excised Biomphalaria glabrata mantle, or an equal offering of both. The reported two-factor analysis did not detect a feeding-treatment effect on mass gain and did detect greater gain in females than males.²³

That result is not evidence that snail tissue equals formulated feed. The study did not measure intake, leftovers, leaching, feed conversion or snail nutrient composition, and the snails had themselves been raised on the comparator pellet. Failure to find a difference in a small trial is not an equivalence test. The final sex groups were small and uneven, and no diet-by-sex interaction or power calculation was reported.

The internal reporting also requires caution. Forty-two shrimp began the trial, but only 40 appear in the final groups despite a stated 100 percent survival. The abstract swaps the snail-only and mixed sex counts relative to the results, and one female mass-gain mean in the table is impossible relative to that treatment's final mass and the plotted observations. The source provides no raw data to repair those conflicts.

Diet composition changes what you observe

Five tested commercial diets all supported more than 80% survival and satisfactory growth and spermatophore quality over 90 days, but body lipid and carotenoid content differed with diet. The diet with the most carotenoid produced brighter body colour.⁴ In a separate broodstock experiment, three commercial diets produced similar egg size and realized fecundity, yet differed in egg carotenoids, juvenile lipid droplets and 32-day juvenile weight.⁵

A 56-day experiment in juvenile male red shrimp also found that particular astaxanthin treatments changed measured redness, chroma and lightness.⁶ A colour food can change phenotype on the day of grading; it cannot change a shrimp's inherited colour genotype or guarantee permanent grade. Record diet, lighting, background and time since feeding when comparing colour.

A separate 31-day experiment compared gel spheres containing live Haematococcus pluvialis, live Spirulina platensis or no microalgae in three-day-old wild, Fire Red and Diamond Blue juveniles.¹¹ The article reports four baskets of 11 juveniles for the treatment combinations, but the baskets shared recirculating water and its one-way analysis does not state that basket was modeled as the independent unit. End-point uropod colour responses differed by phenotype and supplement. Those labels identify the study stocks, not every seller's same-named line.

All groups had 100 percent survival, so survival cannot distinguish a supplement benefit. The study did not measure growth, health, breeding, persistence after withdrawal or genetic change. It also does not convert three experimental spheres into a spoon dose for a powder, validate home-grown algae, or guarantee a sale grade. Use it as short, formulation-specific colour evidence.

An older 16-week aquarium trial compared four branded feeds in three 50-litre aquaria per product, with 30 shrimp per aquarium and feeding to satiation twice daily. The Spirulina granule group had the highest reported final means for weight, total length and survival, while the pond-stick group had the lowest.⁷ This does not isolate a Spirulina effect. Product formulation and physical form changed together, intake was not a fixed ration and some pairwise differences were not statistically distinct. Use it as evidence that feed choice can change measured outcomes, not as a current brand ranking or a pellet-count prescription.

A separate 11-month trial tested six formulated diets with 0, 1, 3, 5, 8 or 10 percent Arthrospira meal in three aquaria per diet. The diets were kept near the same protein and energy values by replacing part of the fish meal, and shrimp received 3 percent of wet biomass per day. Several growth, survival and reproductive measures were higher in the 8 or 10 percent groups under that exact protocol.⁸

The percentage is a dry feed-formulation percentage, not a powder dose for tank water and not permission to add 8 to 10 percent to a finished food. Control female survival was only 25.7 percent after 11 months, eggs were removed and incubated artificially, and the independent incubation unit is not clear. The reported relative-fecundity advantage also declined after the sixth spawning. This is useful ingredient-replacement evidence, not a universal recipe or current product claim.

Black soldier fly meal is a formula ingredient, not half the bowl

A 60-day study compared five complete diets labelled as replacing 0, 25, 50, 75 or 100 percent of fish meal with black soldier fly larvae meal. Those labels are replacement levels, not the fraction of the complete food. The formulas actually contained 0, 10.8, 21.68, 32.62 and 43.64 percent insect meal.¹⁵ Each diet had five 3-litre aquaria, with one adult male and one adult female per aquarium and Java moss present.

The 50 percent replacement formula, containing 21.68 percent insect meal, had higher reported length gain, weight gain and specific growth rate than the 75 and 100 percent replacement formulas. The table's comparison letters did not distinguish that group from the fish-meal control or the 25 percent replacement group. Survival ranged from 70 to 90 percent without a reported statistical difference. Ovary measurement, reported offspring count and colour intensity also did not differ statistically among diets.

This is not a recipe for making half of a home food from insects. Fish meal, insect meal, fish oil, corn meal and cellulose changed together, measured lipid and ash declined across the formulas, and amino acids, fatty acids and energy were not reported. Every diet also contained the same Spirulina and astaxanthin. Feeding was described as both to satiation and predetermined, without an exact ration, uneaten-feed recovery method or leaching correction. Daily water checks were stated but their values were not reported.

Several reporting problems further limit the molt and reproduction claims. The published molting equation multiplies molts per starting shrimp by 100, while the table reports values from 4.8 to 7.6 percent without reconciling that scale. Ovary size was calculated as length times width but reported in centimetres, one figure caption refers to male ovary development, and the paper does not state whether aquarium or animal observations entered each analysis. Use this as a promising complete-formula comparison, not a whole-insect feeding method, molting rate, colour benefit or proof of reproductive safety.

A 12-week thesis used a two-by-two comparison of commercial shrimp and trout grow-out feeds offered once or three times daily. Each treatment combination had three 40-litre aquaria with ten shrimp. Feeding frequency changed several growth, survival, shed-exoskeleton and reproductive outcomes, while product differences appeared in selected reproductive and colour measures.⁹ The daily ration was not reported, groups consumed different amounts and the products differed in several declared and undeclared properties. The result is a direct frequency experiment, but it does not establish three meals per day for a home colony or a pellet count.

A probiotic feeding trial is not a tank dose

A 30-day experiment assigned a basal feed, four feeds containing 1 to 4 g galacto-oligosaccharide per kg and four corresponding feeds also containing 1 mL of an Enterogermina Bacillus clausii product per kg to three 10-litre tanks each. Some treatment groups had higher reported mass, specific-growth and feed-conversion values than the basal-feed control.¹² The 2 g galacto-oligosaccharide plus product group had the highest table values for several sex-specific growth outcomes under that protocol.

There was no probiotic-only group, so the experiment cannot isolate a B. clausii effect or demonstrate that the two additives acted synergistically. Feeding was to satiation, actual intake was not standardized, finished-feed bacterial viability and additive retention were not verified, and the analysis did not state whether tank was modeled as the experimental unit. The paper's survival language is internally inconsistent, so this guide does not claim a survival benefit. It measured no microbiome, bacterial colonization, immune marker, disease challenge or pathogen outcome. A human probiotic product mixed into an experimental feed is not a dose for tank water, a bacterial starter, a medicine or a universal home-food recipe.

A fermented preparation is not a fresh-vegetable recipe

A 40-day experiment compared a Spirulina-tablet control with mustard greens, carrot, bean sprouts or cabbage prepared by boiling and then fermenting for five days with a named microbial product and molasses. Each feed had three nominal 21.6-litre aquarium compartments, and food was kept available with twice-daily checks.¹⁷

The mustard-green, bean-sprout and cabbage groups were not statistically distinguished from the control for the reported length gain, weight gain or specific growth rate. The carrot group was lower. That result is a comparison of five complete treatment packages. There was no unfermented-vegetable group, microbial-product-only group or molasses-only group, so it does not isolate a vegetable or show that fermentation caused the result.

The feeds were offered ad libitum without an exact ration or measured intake, and the wet preparations were not balanced for dry matter, protein or energy. The paper also does not state clearly whether aquarium or sampled-animal values entered the analysis. Survival means were only 63.33 to 71.66 percent, and the authors attributed many deaths to repeated measurement outside water. This is limited alternative-feed evidence, not permission to feed market scraps, copy a home fermentation recipe, replace formulated food or set a vegetable-removal deadline.

A live-food comparison is not a Tubifex recipe

A 40-day study assigned live Tubifex, Chironomus larvae, Daphnia, or an unidentified Spirulina pellet to three nominal 21.6-litre aquaria per feed. Food remained available ad libitum and was checked twice daily. The Tubifex group had the highest reported survival mean at 85.00 percent and the highest visible-red colour-score mean at 3.50. Length and weight gains were not statistically distinguished among the four feeds.¹⁸

This comparison does not establish a complete food or live-feed dose. The stock was identified only as Neocaridina sp.; age, sex, line, starting size and starting colour balance were not reported. The stated 21.6-litre volume and one shrimp per litre do not resolve an integer starting count. Intake was not measured, live foods and pellets changed physical form as well as composition, and the pellet product was not identified.

The reporting also conflicts. The written method describes four colour levels, while its figure displays five with different descriptions. The table note uses P greater than 0.05 for different comparison letters, while the results text uses P less than 0.05. One discussion passage substitutes insect larvae for Daphnia, and another says Chironomus and Daphnia produced more intense colour than Tubifex even though the table ranks Tubifex highest. Use the experiment as evidence that whole feeding treatments can change reported outcomes, not as a reason to introduce live worms, promise survival or alter inherited grade.

Our measured feeding protocol

Our practice This is a control system for our tanks, not a species-wide prescription. It starts only after a baseline is recorded.

1. Record tank, date, estimated adult and juvenile ranges, reproductive stage, food product, lot and a repeatable portion unit. A small gram scale is better than "a pinch" when the food mass can be resolved.
2. Use a formulated food whose ingredient list and guaranteed analysis are available. Protein percentage alone does not establish nutritional adequacy; lipid, energy, micronutrients, digestibility and leaching also matter.
3. Place the measured portion where it can be inspected. Record first contact, how much remains at the next scheduled observation and whether food is dispersing beyond recovery. Remove material that is visibly decomposing or associated with worsening water readings.
4. Hold the portion and observation interval constant long enough to compare results. If food repeatedly remains, reduce one step. If it disappears and juvenile growth or broodstock condition is poor, increase one step. Do not change food, portion and frequency together.
5. Track ammonia, nitrite and nitrate against the tank's baseline, plus deaths, molts, juvenile growth and reproduction. Feeding response is evidence from a series, not whether shrimp crowd one pellet once. Use the cohort guide when growth is the outcome so size, age, density and reproductive stage remain separate.

Different tanks need different evidence

Vegetables, powders and botanicals

• One direct trial compared four fermented vegetable preparations with a Spirulina-tablet control. It did not test fresh grocery vegetables, an unfermented comparison, a balanced vegetable-only home diet or a universal removal time.
• Fine powders spread beyond a dish and are harder to recover. Measure them and judge them by colony and water outcomes, not the word "biofilm" on the label.
• Leaves, wood and cones add surface and decomposing organic matter. This review found no controlled N. davidi study proving that catappa leaves or alder cones prevent disease, provide a complete diet or require a fixed replacement cadence. Replace or remove them for a recorded husbandry reason, using the habitat and wet-addition guide.
• One 140-day cattail study used eight river-collected shrimp in each four-litre microcosm and found more litter mass loss when shrimp directly contacted the litter.¹⁹ Mass loss was not a food intake measurement, weekly filtered river-water replacement supplied other material, and growth, condition, molting and reproduction were not measured. Reported zero mortality does not make the leaf a complete diet.
• Copper is an essential trace nutrient and may appear in complete foods. A copper ingredient on a label is not equivalent to a toxic waterborne dose. Avoid medications and algaecides without explicit invertebrate directions. Evaluate the exact product, route and test limits.

Whether shrimp graze a surface is a different question from whether they control a visible growth or remove nutrients from the aquarium. Read the grazing and cleanup evidence guide.

Sources

1. Viau et al., 2020. Breeding and life cycle of the ornamental freshwater shrimp Neocaridina davidi in a biofilm-based culture system. Aquaculture Research 51: 3847-3864.
2. Pantaleão et al., 2015. Nutritional vulnerability in early stages of the freshwater ornamental red cherry shrimp Neocaridina davidi. Journal of Crustacean Biology 35(5): 676-681.
3. Sonakowska et al., 2017. The effect of starvation and re-feeding on mitochondrial potential in the midgut of Neocaridina davidi. PLOS ONE 12(3): e0173563.
4. Tomas et al., 2020. Effect of diets on carotenoid content, body coloration, biochemical composition and spermatophore quality. Aquaculture Nutrition 26: 1198-1210.
5. Sganga and López Greco, 2020. Effect of commercial diets on female reproductive performance and offspring quality in the red cherry shrimp Neocaridina davidi. Aquaculture Research 51: 5029-5039.
6. Hou et al., 2024. Enhancing the color and stress tolerance of cherry shrimp using astaxanthin and Bidens pilosa. PLOS ONE 19(12): e0315585.
7. Bingol, Turkmen and Karadal, 2016. Effects of different aquarium feeds on growth performance and survival rate of red cherry shrimp Neocaridina denticulata. Ege Journal of Fisheries and Aquatic Sciences 33(3): 217-222.
8. Namaei Kohal, Esmaeili Fereidouni, Firouzbakhsh and Hayati, 2018. Effects of dietary incorporation of Arthrospira platensis meal on growth, survival, body composition, and reproductive performance of red cherry shrimp Neocaridina davidi over successive spawnings. Journal of Applied Phycology 30: 431-443.
9. Elmas, 2019. Effects of different feed types and feeding frequencies on growth performance, moulting cycle, reproductive efficiency and carapace coloration of red cherry shrimps Neocaridina davidi. Izmir Katip Celebi University MSc thesis.
10. Feng et al., 2026. Physiological and biochemical characterization of trypsin from Neocaridina denticulata sinensis and its roles in ontogenesis and immune response. PLOS ONE 21(2): e0342746.
11. Luna-Vivaldo et al., 2024. The coloration of Neocaridina davidi fed with live microalgae Haematococcus pluvialis and the cyanobacteria Spirulina platensis. Latin American Journal of Aquatic Research 52(2): 298-306.
12. Kaya et al., 2022. Symbiotic effect of Bacillus clausii and Galacto-oligosaccharide on growth and survival rates in red cherry shrimp Neocaridina davidi. Marine and Life Sciences 4(2): 146-151.
13. Chen et al., 2025. Effects of Methyl Farnesoate on the Growth and Antioxidant Capacity of Neocaridina denticulata. Antioxidants 14(6): 635.
14. Luan et al., 2025. Sesquiterpenoid Hormones Farnesoic Acid and Methyl Farnesoate Regulate Different Gene Sets in Shrimp Neocaridina davidi Hepatopancreas. Biomolecules 15(6): 815.
15. Azhar et al., 2025. Growth, reproduction and pigmentation performances of red cherry shrimp Neocaridina davidi fed with black soldier fly larvae Hermetia illucens meal. Planetary Sustainability 3(2): 56-76.
16. Pierre, Kotani and Irabor, 2025. Experimental measurement of enzyme activity during initial crustacean growth using the ornamental shrimp, Neocaridina denticulata, as a model and changes in activity with growth. Crustacean Research 54: 19-33.
17. Wijaya et al., 2023. Effect Of Market Vegetable Waste Fermented Feeding On The Growth Of Red Cherry Ornamental Shrimp Neocaridina davidi. Barakuda 45 5(1): 102-113.
18. Kusuma et al., 2025. The effectiveness of natural feed in enhancing color and survival rate of ornamental shrimp Neocaridina sp.. Juvenil 6(3): 290-300.
19. Kong et al., 2019. Promotive performance of shrimp Neocaridina denticulata on Typha angustifolia leaf litter decomposition. Hydrobiologia 827: 75-87.
20. Włodarczyk et al., 2019. Relationship between ROS production, MnSOD activation and periods of fasting and re-feeding in freshwater shrimp Neocaridina davidi. PeerJ 7: e7399.
21. Wang et al., 2010. Effects of Dietary Manganese Supplementation on Antioxidant Enzyme Activity in the Shrimp Neocaridina heteropoda. Israeli Journal of Aquaculture - Bamidgeh 62(2): 78-84.
22. Włodarczyk, Student and Rost-Roszkowska, 2019. Autophagy and apoptosis in starved and re-fed Neocaridina davidi midgut. Canadian Journal of Zoology 97(4): 294-303.
23. Lima, 2023. Females weigh more: evaluation of Neocaridina davidi growth and use of the alternative food Biomphalaria glabrata snail. UNESP undergraduate thesis, 22 pages.
24. Zhang et al., 2024. Transcriptome profiles of the hepatopancreas of the Chinese swamp shrimp Neocaridina denticulata under different diurnal rhythms. ScienceAsia 50(2): 2024043. Official article and supplement PDF. NCBI BioProject PRJNA832034.