Production of tambaqui juveniles (Colossoma macropomum) and arugula microgreens (Eruca sativa) in small-scale aquaponic systems: technical and economic viability

Metadata

• Cite key: dasilvaProductionTambaquiJuveniles2025
• Item type: Journal Article
• Authors: B. Da Silva, F. Da Silva, K. Oliveira, R. Filho, W. Severi, M. Medeiros, M. Coimbra
• Affiliation: Applied Genetics Laboratory, Aquaponics Laboratory, and Limnology Laboratory, Department of Fisheries and Aquaculture, Federal Rural University of Pernambuco (UFRPE), Recife, PE, Brazil
• Journal: Aquaculture International 33(1) (2025) Article 41
• Date: 01/2025
• Date added: [not reported]
• DOI: 10.1007/s10499-024-01759-y
• Funding: Graduate Program Support of the Federal Rural University of Pernambuco (PROAP-UFRPE)
• URL: https://doi.org/10.1007/s10499-024-01759-y
• PDF: Da Silva et al. - 2025 - Production of tambaqui juveniles (Colossoma macrop.pdf

Opinion

A clean, honestly-reported small-scale family-aquaponics trial with a genuinely useful head-to-head NFT-vs-FRT comparison and an unusually thorough economic-viability section (rare in this literature). The core growth/yield findings (NFT > FRT for both fish and microgreens) are internally consistent and well-supported by significance testing. However, re-reading the paper fully surfaced two BLOCK-level contradictions between the Results narrative and Table 1 (pH and water temperature trial means disagree on both mean and, for pH, even on which treatment is higher) with no reconciling statement anywhere — worth flagging before citing either water-quality figure. Published online 12 November 2024, ahead of its January 2025 issue date.

Abstract

Aquaponics is an activity that has been growing worldwide as a sustainable alternative for food production and practical implementation throughout society. The objective of this study was to develop micro-scale aquaponics with a native Brazilian species, the tambaqui (Colossoma macropomum), associated with the cultivation of microgreens, an agro-food trend, using arugula (Eruca sativa), with the aim of evaluating the economic viability of family production. For 70 days, the growth performance of 80 tambaqui juveniles (average initial weight=61.88 g±8.31 g) was monitored. Two treatments with two hydroponics structures, Nutrient Film Technique (NFT) and Floating Raft Technology (FRT), each with four replications, were compared. Tambaqui survival was 100% in both treatments. The final weight of juveniles was 136.38 g (±35.67 g) in NFT systems and 116.5 g (±24.63 g) in FRT systems. In the four production cycles carried out, the arugula microgreens performed better in the NFT systems (157.38 g±21.74 g; 6.68 cm±0.17 cm) compared to FRT (41 g±6.18 g; 4.46 cm±0.58 cm), identifying the best hydroponic unit for the cultivation of E. sativa microgreens. Finally, the economic viability study was conducted considering the average wage in Brazil of US$ 285.66 for the implementation of a homemade aquaponic system. The possible scenarios considered a family income of two or more minimum wages for the implementation of at least one aquaponic system.

Summary

Researchers at UFRPE built eight small-scale (200 L fish tank) aquaponic units in a greenhouse, split into two treatments — Nutrient Film Technique (NFT) and Floating Raft Technology (FRT) hydroponic structures, four replicate units each — and stocked each with 10 tambaqui (Colossoma macropomum) juveniles for a 70-day grow-out, while running four consecutive 10-day cycles of arugula (Eruca sativa) microgreens fed solely by the fish-system water (no nutrient solution added to either treatment). Fish survival was 100% in both treatments; NFT produced significantly larger final fish (136.4 g vs 116.5 g) and significantly better microgreens (roughly 4x the fresh weight and double the germination rate of FRT), attributed to greater water renewal at the seed pots and less biofilm/phytoplankton competition for nutrients in the tubular NFT channels versus the standing-water FRT boxes. FCR, specific growth rate, and daily biomass gain did not differ significantly between treatments. The paper’s second half is an unusually detailed economic-viability analysis (materials, solar vs. grid electricity, water footprint, and 5-year payback projections for one vs. eight production units), concluding that microgreen sales, not fish sales, drive any plausible return on investment, and that building eight units reaches a positive first-year balance while a single unit does not. The paper’s own two Results-vs-Table 1 numerical mismatches on pH and water temperature (detailed below) mean those two water-quality parameters specifically should be re-verified against the original data before being cited.

---

Experiment data

• Location: Federal Rural University of Pernambuco (UFRPE), Recife, PE, Brazil — greenhouse, 66 m²
• Design: 2 treatments (NFT vs FRT hydroponic structure), 4 replicate production units each, 8 units total; both treatments aquaponic (fish-water-fed only, no nutrient solution) — no separate hydroponic-only control
• Replicates / n: 4 per treatment
• Duration: Fish: 70 days grow-out. Plant: 4 complete 10-day microgreen cycles (irregular intervals between cycles for hydroponic-unit maintenance)
• Organisms: Tambaqui (Colossoma macropomum) / Arugula (Eruca sativa), microgreens
• Statistics: Shapiro-Wilk normality + Bartlett homoscedasticity; t-test (parametric: nitrate, orthophosphate, FCR, biomass gain, SGR, standard length, fresh weight, germination index); Kruskal-Wallis (non-parametric: DO, pH, temperature, nitrite, ammonia, total weight, length); R software (R Core Team 2020)
• Feed Conversion Rate (FCR): NFT 2.18 ± 0.63 vs FRT 2.55 ± 0.61 (ns)
• SGR: NFT 1.06 ± 0.44 %/day vs FRT 0.89 ± 0.21 %/day (ns)
• Fish final weight: NFT 136.38 ± 35.67 g vs FRT 116.50 ± 24.63 g (p<0.05)
• Arugula microgreen fresh weight (per replicate): NFT 157.38 ± 21.74 g vs FRT 41 ± 6.18 g (p<0.05)
• Arugula standard length: NFT 6.68 ± 0.17 cm vs FRT 4.46 ± 0.58 cm (p<0.05)
• Germination index: NFT 65.43 ± 9.91% vs FRT 34.02 ± 13.57% (p<0.05; cycles 3-4 only)

---

Fish growth performance

This paper: Tambaqui grew from ~62 g to 136.4 g (NFT) or 116.5 g (FRT) over 70 days at 100% survival in both treatments (Table 2, p.8). FCR (NFT 2.18±0.63, FRT 2.55±0.61) did not differ significantly, though the Discussion (p.13) considers both values “high in a productive context,” attributing this partly to a cold spell (<26°C) during the first 4 weeks.

Compared with:

• todo Oliveira et al. 2022 — tambaqui in aquaponics reaching ~150 g final biomass at 90 days (p.14)
• todo Nascimento et al. 2023 — FCR 1.8 over 30 days, ~1.0 kg fish, 3 fish/m³ (p.14)
• todo Silva et al. 2020 — FCR 1.8, tambaqui in aquaponics, 67.78 g initial weight, 50 fish/m³, 70 days (p.14)
• todo Cunha and Santos Junior 2011 — FCR 2.92, 50 fish/m³, 120 days (p.14)
• todo Sousa et al. 2016 — high stocking density raises tambaqui ammonia excretion to 93.7% / 0.35 mg/L/day (p.14)

Arugula microgreen performance and hydroponic structure comparison

This paper: NFT outperformed FRT on every measured microgreen metric across 4 cycles: standard length (6.68±0.17 vs 4.46±0.58 cm), fresh weight per replicate (157.38±21.74 vs 41±6.18 g), and germination index (65.43±9.91% vs 34.02±13.57%, cycles 3-4 only) (Table 3, p.9). The authors attribute this to greater water renewal at the seed pots in the flowing NFT channels versus the standing water in FRT boxes, plus a biofilm/phytoplankton layer that formed on the FRT box surface and may have sequestered nutrients (p.14-15).

Compared with:

• todo Kizak and Kapaligoz 2019 — arugula (E. vesicaria) microgreens in aquaponics, no difficulty with high N compounds but stressed biofilter surface area/light protection (p.13)

Water quality

This paper: DO and non-ionized ammonia differed significantly between treatments (DO: NFT 5.28±1.12 vs FRT 5.98±1.12 mg/L; NH3: NFT 0.849±0.187 vs FRT 0.677±0.181 mg/L, Table 1, p.6) — both exceeding the paper’s cited tambaqui ammonia tolerance limit of 0.46 mg/L (p.14). Nitrite and nitrate stayed low throughout (Table 1). pH trended acidic after week 4, attributed to low system alkalinity unable to buffer against the region’s reservoir water. See Extraction notes below for two unresolved Results-text-vs-Table-1 numerical conflicts on pH and water temperature.

Compared with:

• todo Boyd et al. 2016 — ideal recirculating-system alkalinity 40-100 mg/L CaCO3 (p.13)
• todo Suárez-Cáceres et al. 2021 — pH<6 compromises bacterial nitrification (p.3, p.13)

Economic viability

This paper: One production unit costs US$231.52(conventionalelectricity)tobuild;eightunitscostUS$ 991.71. Annual revenue is dominated by microgreen sales (US$327.35/unitvsUS$ 4.91/unit for tambaqui juveniles) (Table 5, p.12). Payback period: ~1.2 years (one unit, grid electricity) to ~3.7 years (one unit, solar); ~7 months to ~1 year for eight units. The authors conclude juvenile tambaqui sales are not commercially advantageous at this scale/duration, and recommend targeting the “baby tambaqui” (~300 g) niche market instead by extending the grow-out period at reduced density.

Compared with:

• todo Rizal et al. 2018 — optimistic on short value-chain aquaponic ROI via direct-to-consumer sales
• todo Greenfeld et al. 2019 — categorizes aquaponics as “beneficial to society, but with low profitability,” argues for public subsidy
• todo Silva and Van Passel 2020 — proposes including aquaponics in Brazilian Inovagro/BNDES rural financing criteria

Linked claims

• NFT outperforms floating raft for microgreen production in aquaponics
• Aquaponic plant revenue typically exceeds fish revenue at small scale
• Low system alkalinity drives pH instability in aquaponics

Citations to chase

• todo Oliveira, G.F. et al. (2022) — tambaqui aquaponics, stocking density optimum at 20 animals/m³
• todo Nascimento, E.T.D.S. et al. (2023) — açaí seedlings in tambaqui aquaponics, FCR 1.8
• todo Silva, T.B.F. et al. (2020) — tambaqui + hydroponics recirculation system, FCR 1.8
• todo Cunha, V.V., Santos Júnior, A. (2011) — tambaqui net-cage density trial, FCR 2.92
• todo Sousa, R.G.C. et al. (2016) — tambaqui stocking density and ammonia excretion
• todo Kizak, V., Kapaligoz, S. (2019) — arugula microgreens in aquaponic/recirculating water quality comparison
• todo Rizal, A. et al. (2018) — economic/social benefits of integrated fish-plant aquaponics
• todo Greenfeld, A. et al. (2019) — economically viable aquaponics, cost-benefit categorization
• todo Silva, M.F., Van Passel, S. (2020) — climate-smart aquaponics assessment, Northeast Brazil

Extraction notes

Paper type: Confirmed experiment — randomized fish selection, two defined treatments (NFT/FRT), 4 replicates each, formal significance testing (t-test/Kruskal-Wallis) throughout. Not a review or exploratory pilot.

Trial structure: 2 rows: T1 = NFT, T2 = FRT. Both are aquaponic treatments (differing only in hydroponic structure); there is no hydroponic-only (nutrient-solution) control anywhere in this study, so all HYD-specific columns (FUE HYD, Tissue nitrate HYD, HYD) are NA for both rows — a structural absence, not the paper’s silence.

Contradictions found (severity-tagged):

• ⚠️BLOCK Aq pH. Results text (p.6): “The pH remained below neutrality for most of the cultivation period (NFT=6.15±0.53; FRT=6.28±0.53).” Table 1 (p.6): pH NFT=6.38±0.5ᵃ; FRT=6.31±0.48ᵃ. Both mean and SD differ for each treatment, and the relative ranking flips: the text has FRT>NFT, the table has NFT>FRT. No reconciling statement anywhere in the paper. Recorded UNCLEAR in the Aq pH cell for both T1 and T2; both value sets preserved in Experimental Remarks. Affects: the Discussion’s “pH<6.0” acidification narrative (p.13), which never specifies which treatment.
• ⚠️BLOCK Water temperature. Results text (p.6): “water temperature remained within the animals’ optimal range (NFT=28.1°C±1.72; FRT=28.3°C±1.74).” Table 1 (p.6): Temperature NFT=27.9±1.23ᵃ; FRT=28.2±1.47ᵃ. Both mean and SD differ for each treatment (ranking direction agrees — FRT>NFT in both — but magnitudes do not, and no basis exists to prefer either source). Recorded UNCLEAR for both T1 and T2’s Water temperature cell; both value sets preserved in remarks.
• ⚠️MATERIAL NO3-N SD (NFT only). Results text (p.6-7): “nitrate (N-NO3) also resulted in small accumulations (NFT=0.293 mg.L⁻¹±0.0041; FRT=0.291 mg.L⁻¹±0.028).” Table 1: Nitrate NFT=0.294±0.0412ᵃ; FRT=0.291±0.028ᵃ. FRT matches exactly between sources. NFT mean agrees within rounding (0.293≈0.294); NFT SD differs by a factor of 10 (0.0041 vs 0.0412). Table 1’s SD is order-of-magnitude consistent with the paired FRT SD (0.028) and with NFT’s own nitrite/ammonia SDs in the same table; 0.0041 would be anomalously tight for this measurement and is most plausibly a dropped-digit typo. Table 1’s value (0.294±0.0412) recorded for T1’s NO3-N; text’s 0.293±0.0041 noted as the conflicting alternative.
• ⚠️CHECK Plant fresh weight unit. Methods (p.5) defines fresh weight as “weight in grams of the microgreens from each replicate” — a per-replicate (per-box) aggregate, not a per-plant weight, but the trials.csv schema column is defined as g/plant. No per-plant figure is derivable without the exact number of germinated plants per replicate (only a % germination index against 60 sown seeds is given, and only for cycles 3-4). Table 3’s per-replicate means (NFT 157.38±21.74 g; FRT 41±6.18 g) recorded in the Plant fresh weight cell, explicitly labelled per-replicate rather than per-plant.
• ⚠️MINOR Alkalinity SD. Results text (p.6): NFT=11.3±4.44, FRT=10.8±4.44 mg/L CaCO3. Table 1: NFT=11.31±4.68ᵃ, FRT=10.8±4.68ᵃ. Means agree closely (rounding); both treatments’ SDs are off from the table by the same ~0.24 mg/L. No cell affected — alkalinity has no dedicated trials.csv column.

Severity tally: 2 BLOCK, 1 MATERIAL, 1 CHECK, 1 MINOR. Per SCHEMA.md scoring (2+ BLOCK → suspect), quality: suspect — driven entirely by the pH/temperature Results-vs-Table mismatches, not by the paper’s primary yield/growth findings (fish and microgreen growth results are internally consistent and not part of any BLOCK). Statistically analysed left Y for both rows since the BLOCKed parameters (pH, water temperature) are secondary environmental monitoring variables, not the study’s primary outcomes (fish and plant growth/yield), which carry no BLOCK.

Not a contradiction (different time points, both explicitly stated): fish arrived from the aquaculture station at 53.6±6.59 g (p.2, pre-quarantine), grew during a 20-day quarantine to a pooled 61.88±8.31 g at stocking (p.3, matches abstract), which then splits into Table 2’s per-treatment initial weights (NFT 62.6±6.46; FRT 61.15±9.94) used in trials.csv.

[not reported] / [unclear] fields grouped by field:

• Fish: Initial Stock density (paper states “10 juveniles/0.2 m³” in Discussion p.14 but never converts to kg/m³ — not derived); N, P, K of feed (only 36% crude protein given); Total Feed (kg) (FCR formula given, no absolute feed weight stated); Fish biomass created (kg) (only a daily rate — biomass gain g/day — is given, not a stated final-minus-initial total); Fish weight gain (g) (same reasoning); Fish Category; FUE AP; WUE.
• Water: Water classification; EC (never measured); pHOptimal (a <6 correction trigger is stated, not a formal target/optimum); Daily Water exchange rate (only weekly evaporative top-up, ~60 L/tank/week, is described — replenishment, not a fractional exchange rate); Average room Temperature (only water temperature reported).
• Plant: SPAD; Leaf count; Plant dry matter; Plants/m²; Days Plant after transplant (this is a direct-seeded microgreen crop with no transplant step — each 10-day cycle runs from sowing to harvest, not from a transplant date).
• Location: Lat/Long (paper gives only the institution’s name and city, no coordinates — not filled from background knowledge per the prime directive).

NO COLUMN items (in Experimental Remarks, both trial rows):

• Alkalinity (mg/L CaCO3) — see MINOR flag above.
• Orthophosphate (mg/L PO4³⁻) — NFT 0.009±0.0063, FRT 0.011±0.0068 (Table 1 and Results text agree exactly); water chemistry, so does not route to plant.csv either.
• Germination index (%) — NFT 65.43±9.91, FRT 34.02±13.57 (Table 3, cycles 3-4 only; cycles 1-2 not reported, no explanation given). Fits none of plant.csv’s four analyte categories (biochemistry/mineral/microbiology/proximate) and no trials.csv column.
• Daily biomass gain (g/day) — NFT 1.065±0.27, FRT 0.822±0.15 (Table 2); a rate, not a total gain, so does not fit the “Fish weight gain” column as defined.
• Full economic-viability dataset (Tables 4-5: build costs, solar vs. grid electricity, water/energy bills, payback periods) — entirely out of schema scope, not extracted.

Water volume in the system: recorded as 200 L (fish tank only, explicitly stated, p.3) for both rows. Total system volume (+25 L biological filter; +75 L FRT boxes for FRT; NFT tubular-channel volume never stated) is not summed anywhere in the paper and was not computed here (would require an unstated NFT channel volume).

TAN / NH4-N column caveat: the paper measures and reports non-ionized ammonia (NH3) specifically, not total ammonia nitrogen (TAN) or NH4-N (a broader/different measure). No TAN or NH4-N figure is separately given anywhere. The NH3 value is recorded in this column as the closest available proxy — treat as NH3, not true TAN.

New tags introduced: Meta/Fish/Tambaqui (no existing tambaqui/Colossoma facet in the vault). Reused Meta/Plant/Arugula (already exists) even though this paper studies the microgreen growth stage specifically, not mature arugula — noted here as a judgment call, not a new facet.

---

Source: Da Silva et al. - 2025 - Production of tambaqui juveniles (Colossoma macrop.pdf

---

Data Tables

Structured data extracted from this paper into the vault's trials.csv / plant_measurements.csv datasets. Fields the paper didn't report are omitted. Download the full datasets (measurements).

Trial Parameters

dasilvaProductionTambaquiJuveniles2025-T1

Fish

Field	Value
Fish	Colossoma macropomum (tambaqui), juveniles
FCR	2.18 +/- 0.63
SGR	1.06 +/- 0.44
Protein	36
% of body weight	2.5
Fish size initial	62.6 +/- 6.46
Fish size final	136.38 +/- 35.67
Feed routine	Twice daily (9:00 am and 3:00 pm)
Feed regime	Commercial feed 36% crude protein (Nutripiscis TR 36%) at 2.5% of biomass; feeding rate readjusted every 15 days after biometry based on updated biomass
Fish survival rate	100
Fish trial duration (days)	70

Water

Field	Value
Water recycle	3 L/min
Water volume in the system	200 L (fish tank only, see remarks)
Water type	Reservoir water (regional supply) (p.13)
Aq pH	UNCLEAR (see remarks - WARN-BLOCK)
Dissolved Oxigen	5.28 +/- 1.12
Water temperature	UNCLEAR (see remarks - WARN-BLOCK)
TAN / NH4-N	0.849 +/- 0.187
NO2-N	0.195 +/- 0.055
NO3-N	0.294 +/- 0.0412

Plant

Field	Value
Plant	Arugula (Eruca sativa), microgreens
Details	60 pesticide-free seeds (+/-5)/pot; no nutrient solution added to either treatment; 4 complete 10-day cycles harvested; standard length measured stem-to-highest-leaf; fresh weight measured per replicate (box), not per plant
Plant Category	Microgreens / baby leaf (p.1-2, Keywords, Methods)
Plant height	6.68 +/- 0.17
Plant fresh weight	157.38 +/- 21.74

System & Setup

Field	Value
System type	Nutrient Film Technique (NFT)
Media Details	Tubular PVC channels, 1.40 m length, 11 holes each, 3 channels/replicate; circular pots with fragmented expanded clay pebbles as germination substrate
Biological system already in use	Y (Biofilters matured 20 days without fish or plants prior to stocking, to establish nitrifying bacteria (p.3))
pH Buffers	Y (Sodium bicarbonate (NaHCO3) added whenever pH fell below 6, to avoid compromising bacterial nitrification (p.3))
Nutrient supplemented	N (No nutrient solution added to any treatment; vegetables received only nutrients from fish tank water (p.5))
Equipment	Circular fish tank (200 L); submersible pump (800 L/h); mechanical filter with perlon mesh; biological filter with expanded clay pebbles (25 L); Instrutherm MO-920 oximeter; digital pH meter; photometer for N-NO2/N-NO3/NH3/alkalinity/orthophosphate; 400 W monophase photovoltaic panel; 150 A.h stationary battery; 30 A charge controller; 12-V/20 W submersible pumps
Control Parameters	pH corrected with NaHCO3 whenever below 6; feeding rate readjusted every 15 days per updated biomass from biometry; water topped up weekly to replace evaporative losses
Combination	Tambaqui (Colossoma macropomum) and arugula (Eruca sativa) microgreens; NFT vs FRT hydroponic units, both aquaponic (fed only by fish-system water, no external nutrient solution added to either); no true hydroponic-only control in this study

Site

Field	Value
Region	South America
Country	Brazil

Results & Statistics

Field	Value
Measured Unit	cm (standard length); g/replicate (fresh weight); % (germination index)
Statistic Details	Shapiro-Wilk normality test; Bartlett homoscedasticity test; t-test (parametric: nitrate, orthophosphate, FCR, biomass gain, SGR, standard length, fresh weight, germination index); Kruskal-Wallis test (non-parametric: dissolved oxygen, pH, temperature, nitrite, ammonia, total weight, length); R statistical software (R Core Team 2020)
Statistically analysed	Y
Replicates (n)	4
AP	157.38

Experimental Remarks: TRIAL DEFINITION: T1 = Nutrient Film Technique (NFT) aquaponic treatment (10 tambaqui juveniles/tank, 4 tank replicates, 3 tubular PVC channels/replicate for arugula microgreens). Paired comparison = T2 (FRT), same paper, same fish husbandry; both treatments are aquaponic (fed solely by fish-tank water, no nutrient solution added to either) - there is no separate hydroponic-only (nutrient-solution-fed) control in this study, so HYD-specific columns (FUE HYD, Tissue nitrate HYD, HYD) = NA throughout, per schema (structural absence, not silence). | WARN-BLOCK Aq pH: Results text (p.6): ‘The pH remained below neutrality for most of the cultivation period (NFT=6.15+/-0.53; FRT=6.28+/-0.53)’. Table 1 (p.6): pH NFT=6.38+/-0.5a; FRT=6.31+/-0.48a. Both mean and SD differ for each treatment, and the ranking reverses (text: FRT>NFT; table: NFT>FRT), with no reconciling statement anywhere in the paper. Recorded UNCLEAR for both trials’ Aq pH cell; both value sets preserved here: text NFT 6.15+/-0.53 / FRT 6.28+/-0.53; Table 1 NFT 6.38+/-0.5 / FRT 6.31+/-0.48. Affects: Discussion’s pH-acidification narrative (p.13), which never specifies which treatment. | WARN-BLOCK Water temperature: Results text (p.6): ‘water temperature remained within the animals’ optimal range (NFT=28.1 degC+/-1.72; FRT=28.3 degC+/-1.74)’. Table 1 (p.6): Temperature NFT=27.9+/-1.23a; FRT=28.2+/-1.47a. Both mean and SD differ for each treatment (ranking direction agrees, FRT>NFT in both, but magnitudes do not); no basis to prefer either source. Recorded UNCLEAR for both trials’ Water temperature cell; both value sets preserved here: text NFT 28.1+/-1.72 / FRT 28.3+/-1.74; Table 1 NFT 27.9+/-1.23 / FRT 28.2+/-1.47. | WARN-MATERIAL NO3-N (NFT only): Results text (p.6-7): ‘nitrate (N-NO3) also resulted in small accumulations (NFT=0.293 mg.L-1+/-0.0041; FRT=0.291 mg.L-1+/-0.028)’. Table 1 (p.6): Nitrate NFT=0.294+/-0.0412a; FRT=0.291+/-0.028a. FRT matches exactly between sources; NFT mean agrees within rounding (0.293 approx 0.294) but NFT SD differs by a factor of 10 (0.0041 text vs 0.0412 table). Table 1’s SD is order-of-magnitude consistent with the paired FRT SD (0.028) and with NFT’s own nitrite/ammonia SDs in the same table; 0.0041 would be anomalously tight and is most plausibly a dropped-digit typo. Table 1 value (0.294+/-0.0412) recorded for NFT’s NO3-N; text’s 0.293+/-0.0041 noted as the conflicting alternative (affects T1/NFT row only, FRT/T2 unaffected). | WARN-CHECK Plant fresh weight unit: Methods (p.5) defines fresh weight as ‘weight in grams of the microgreens from each replicate’ - a per-replicate (per-box) aggregate, not a per-plant weight, but the schema column is defined as g/plant. No per-plant figure is derivable without the exact number of germinated plants per replicate (only a % germination index against 60 sown seeds is given, and only for cycles 3-4); computing one would be derivation, not performed. Table 3’s per-replicate means recorded in Plant fresh weight, explicitly labelled per-replicate not per-plant. | WARN-MINOR Alkalinity SD: Results text (p.6) NFT=11.3+/-4.44, FRT=10.8+/-4.44 mg/L CaCO3; Table 1 NFT=11.31+/-4.68a, FRT=10.8+/-4.68a. Means agree closely (rounding); both treatments’ SDs are off from the table by the same ~0.24 mg/L. No cell affected - alkalinity has no dedicated trials.csv column. | NOT DERIVED, left NR: Initial Stock density (Discussion p.14 states ‘10 juveniles/0.2 m3’ but never converts to kg/m3); N, P, K of feed (only 36% crude protein given); Total Feed (kg) (FCR=Feed(kg)/BG formula given, no absolute feed weight stated); Fish biomass created (kg) (only a daily rate, biomass gain g/day, is given - NFT 1.065+/-0.27, FRT 0.822+/-0.15 g/day, Table 2 - not a stated final-minus-initial total); Fish weight gain (g) (same reasoning, only the daily rate is given); Fish Category; Water classification; EC (never measured); pHOptimal (a <6 correction trigger is stated, p.3, not a formal optimal target); Daily Water exchange rate (only weekly evaporative top-up ~60 L/tank/week / ~240 L/month/unit is described, p.7 - replenishment, not a fractional exchange rate); Average room Temperature (only water temperature is reported); SPAD; Leaf count; Plant dry matter; Plants/m2; Days Plant after transplant (direct-seeded microgreen crop, no transplant step - each 10-day cycle runs sowing-to-harvest); Lat/Long (paper gives only institution name/city, no coordinates - not filled from background knowledge per prime directive); FUE AP; FUE HYD=NA; WUE (not computed under these names, though water consumption ~240 L/month/unit is discussed narratively). | NOT A CONTRADICTION (different time points, both explicitly stated): fish arrived from the aquaculture station at 53.6+/-6.59 g (p.2, pre-quarantine), grew during a 20-day quarantine to a pooled 61.88+/-8.31 g at stocking (p.3, matches abstract), which splits into Table 2’s per-treatment initial weights (NFT 62.6+/-6.46; FRT 61.15+/-9.94) used here. | NO COLUMN: Alkalinity (mg/L CaCO3, see MINOR flag above); Orthophosphate (mg/L PO4 3-) - NFT 0.009+/-0.0063, FRT 0.011+/-0.0068 (Table 1 and Results text agree exactly, p.6-7) - water chemistry, does not route to plant.csv either; Germination index (%) - NFT 65.43+/-9.91, FRT 34.02+/-13.57 (Table 3, cycles 3-4 ONLY - cycles 1-2 not reported, no explanation given) - fits none of plant.csv’s four analyte categories nor any trials.csv column; Daily biomass gain (g/day) - NFT 1.065+/-0.27, FRT 0.822+/-0.15 (Table 2) - a rate, not a total gain, does not fit ‘Fish weight gain’; full economic-viability dataset (Tables 4-5: build costs, solar vs grid electricity, water/energy bills, payback periods) - entirely out of schema scope, not extracted. | Water volume in the system recorded as 200 L (fish tank only, explicitly stated p.3); total system volume (+25 L biological filter; +75 L FRT boxes for FRT; NFT tubular-channel volume never stated) is not summed anywhere in the paper and not computed here (would require an unstated NFT channel volume). | TAN/NH4-N column caveat: paper measures and reports non-ionized ammonia (NH3) specifically, not total ammonia nitrogen (TAN) or NH4-N. No TAN/NH4-N figure is separately given. NH3 value recorded here as the closest available proxy - treat as NH3, not true TAN. | Statistically analysed left Y for both rows since the BLOCKed parameters (pH, water temperature) are secondary environmental monitoring variables, not the study’s primary outcomes (fish and plant growth/yield), none of which carry a BLOCK.

dasilvaProductionTambaquiJuveniles2025-T2

Fish

Field	Value
Fish	Colossoma macropomum (tambaqui), juveniles
FCR	2.55 +/- 0.61
SGR	0.89 +/- 0.21
Protein	36
% of body weight	2.5
Fish size initial	61.15 +/- 9.94
Fish size final	116.50 +/- 24.63
Feed routine	Twice daily (9:00 am and 3:00 pm)
Feed regime	Commercial feed 36% crude protein (Nutripiscis TR 36%) at 2.5% of biomass; feeding rate readjusted every 15 days after biometry based on updated biomass
Fish survival rate	100
Fish trial duration (days)	70

Water

Field	Value
Water recycle	3 L/min
Water volume in the system	200 L (fish tank only, see remarks)
Water type	Reservoir water (regional supply) (p.13)
Aq pH	UNCLEAR (see remarks - WARN-BLOCK)
Dissolved Oxigen	5.98 +/- 1.12
Water temperature	UNCLEAR (see remarks - WARN-BLOCK)
TAN / NH4-N	0.677 +/- 0.181
NO2-N	0.185 +/- 0.07
NO3-N	0.291 +/- 0.028

Plant

Field	Value
Plant	Arugula (Eruca sativa), microgreens
Details	60 pesticide-free seeds (+/-5)/pot; no nutrient solution added to either treatment; 4 complete 10-day cycles harvested; standard length measured stem-to-highest-leaf; fresh weight measured per replicate (box), not per plant
Plant Category	Microgreens / baby leaf (p.1-2, Keywords, Methods)
Plant height	4.46 +/- 0.58
Plant fresh weight	41 +/- 6.18

System & Setup

Field	Value
System type	Floating Raft Technology (FRT)
Media Details	Rectangular floating boxes, 25 L volume, 3/replicate; polystyrene floating plates; 11 circular pots/plate with fragmented expanded clay pebbles as germination substrate; pots arranged on water surface to avoid soaking substrate, dispensing with artificial root aeration (p.3)
Biological system already in use	Y (Biofilters matured 20 days without fish or plants prior to stocking, to establish nitrifying bacteria (p.3))
Air supplement	N (FRT design explicitly dispenses with artificial oxygenation for plant roots - pots kept on water surface, roots not submerged (p.3))
pH Buffers	Y (Sodium bicarbonate (NaHCO3) added whenever pH fell below 6, to avoid compromising bacterial nitrification (p.3))
Nutrient supplemented	N (No nutrient solution added to any treatment; vegetables received only nutrients from fish tank water (p.5))
Equipment	Circular fish tank (200 L); submersible pump (800 L/h); mechanical filter with perlon mesh; biological filter with expanded clay pebbles (25 L); Instrutherm MO-920 oximeter; digital pH meter; photometer for N-NO2/N-NO3/NH3/alkalinity/orthophosphate; 400 W monophase photovoltaic panel; 150 A.h stationary battery; 30 A charge controller; 12-V/20 W submersible pumps
Control Parameters	pH corrected with NaHCO3 whenever below 6; feeding rate readjusted every 15 days per updated biomass from biometry; water topped up weekly to replace evaporative losses
Combination	Tambaqui (Colossoma macropomum) and arugula (Eruca sativa) microgreens; NFT vs FRT hydroponic units, both aquaponic (fed only by fish-system water, no external nutrient solution added to either); no true hydroponic-only control in this study

Site

Field	Value
Region	South America
Country	Brazil

Results & Statistics

Field	Value
Measured Unit	cm (standard length); g/replicate (fresh weight); % (germination index)
Statistic Details	Shapiro-Wilk normality test; Bartlett homoscedasticity test; t-test (parametric: nitrate, orthophosphate, FCR, biomass gain, SGR, standard length, fresh weight, germination index); Kruskal-Wallis test (non-parametric: dissolved oxygen, pH, temperature, nitrite, ammonia, total weight, length); R statistical software (R Core Team 2020)
Statistically analysed	Y
Replicates (n)	4
AP	41

Experimental Remarks: TRIAL DEFINITION: T2 = Floating Raft Technology (FRT) aquaponic treatment (10 tambaqui juveniles/tank, 4 tank replicates, 3 rectangular 25-L floating boxes/replicate for arugula microgreens). Paired comparison = T1 (NFT), same paper, same fish husbandry; both treatments are aquaponic (fed solely by fish-tank water, no nutrient solution added to either) - there is no separate hydroponic-only (nutrient-solution-fed) control in this study, so HYD-specific columns (FUE HYD, Tissue nitrate HYD, HYD) = NA throughout, per schema (structural absence, not silence). | WARN-BLOCK Aq pH: Results text (p.6): ‘The pH remained below neutrality for most of the cultivation period (NFT=6.15+/-0.53; FRT=6.28+/-0.53)’. Table 1 (p.6): pH NFT=6.38+/-0.5a; FRT=6.31+/-0.48a. Both mean and SD differ for each treatment, and the ranking reverses (text: FRT>NFT; table: NFT>FRT), with no reconciling statement anywhere in the paper. Recorded UNCLEAR for both trials’ Aq pH cell; both value sets preserved here: text NFT 6.15+/-0.53 / FRT 6.28+/-0.53; Table 1 NFT 6.38+/-0.5 / FRT 6.31+/-0.48. Affects: Discussion’s pH-acidification narrative (p.13), which never specifies which treatment. | WARN-BLOCK Water temperature: Results text (p.6): ‘water temperature remained within the animals’ optimal range (NFT=28.1 degC+/-1.72; FRT=28.3 degC+/-1.74)’. Table 1 (p.6): Temperature NFT=27.9+/-1.23a; FRT=28.2+/-1.47a. Both mean and SD differ for each treatment (ranking direction agrees, FRT>NFT in both, but magnitudes do not); no basis to prefer either source. Recorded UNCLEAR for both trials’ Water temperature cell; both value sets preserved here: text NFT 28.1+/-1.72 / FRT 28.3+/-1.74; Table 1 NFT 27.9+/-1.23 / FRT 28.2+/-1.47. | WARN-MATERIAL NO3-N (NFT only): Results text (p.6-7): ‘nitrate (N-NO3) also resulted in small accumulations (NFT=0.293 mg.L-1+/-0.0041; FRT=0.291 mg.L-1+/-0.028)’. Table 1 (p.6): Nitrate NFT=0.294+/-0.0412a; FRT=0.291+/-0.028a. FRT matches exactly between sources; NFT mean agrees within rounding (0.293 approx 0.294) but NFT SD differs by a factor of 10 (0.0041 text vs 0.0412 table). Table 1’s SD is order-of-magnitude consistent with the paired FRT SD (0.028) and with NFT’s own nitrite/ammonia SDs in the same table; 0.0041 would be anomalously tight and is most plausibly a dropped-digit typo. Table 1 value (0.294+/-0.0412) recorded for NFT’s NO3-N; text’s 0.293+/-0.0041 noted as the conflicting alternative (affects T1/NFT row only, FRT/T2 unaffected). | WARN-CHECK Plant fresh weight unit: Methods (p.5) defines fresh weight as ‘weight in grams of the microgreens from each replicate’ - a per-replicate (per-box) aggregate, not a per-plant weight, but the schema column is defined as g/plant. No per-plant figure is derivable without the exact number of germinated plants per replicate (only a % germination index against 60 sown seeds is given, and only for cycles 3-4); computing one would be derivation, not performed. Table 3’s per-replicate means recorded in Plant fresh weight, explicitly labelled per-replicate not per-plant. | WARN-MINOR Alkalinity SD: Results text (p.6) NFT=11.3+/-4.44, FRT=10.8+/-4.44 mg/L CaCO3; Table 1 NFT=11.31+/-4.68a, FRT=10.8+/-4.68a. Means agree closely (rounding); both treatments’ SDs are off from the table by the same ~0.24 mg/L. No cell affected - alkalinity has no dedicated trials.csv column. | NOT DERIVED, left NR: Initial Stock density (Discussion p.14 states ‘10 juveniles/0.2 m3’ but never converts to kg/m3); N, P, K of feed (only 36% crude protein given); Total Feed (kg) (FCR=Feed(kg)/BG formula given, no absolute feed weight stated); Fish biomass created (kg) (only a daily rate, biomass gain g/day, is given - NFT 1.065+/-0.27, FRT 0.822+/-0.15 g/day, Table 2 - not a stated final-minus-initial total); Fish weight gain (g) (same reasoning, only the daily rate is given); Fish Category; Water classification; EC (never measured); pHOptimal (a <6 correction trigger is stated, p.3, not a formal optimal target); Daily Water exchange rate (only weekly evaporative top-up ~60 L/tank/week / ~240 L/month/unit is described, p.7 - replenishment, not a fractional exchange rate); Average room Temperature (only water temperature is reported); SPAD; Leaf count; Plant dry matter; Plants/m2; Days Plant after transplant (direct-seeded microgreen crop, no transplant step - each 10-day cycle runs sowing-to-harvest); Lat/Long (paper gives only institution name/city, no coordinates - not filled from background knowledge per prime directive); FUE AP; FUE HYD=NA; WUE (not computed under these names, though water consumption ~240 L/month/unit is discussed narratively). | NOT A CONTRADICTION (different time points, both explicitly stated): fish arrived from the aquaculture station at 53.6+/-6.59 g (p.2, pre-quarantine), grew during a 20-day quarantine to a pooled 61.88+/-8.31 g at stocking (p.3, matches abstract), which splits into Table 2’s per-treatment initial weights (NFT 62.6+/-6.46; FRT 61.15+/-9.94) used here. | NO COLUMN: Alkalinity (mg/L CaCO3, see MINOR flag above); Orthophosphate (mg/L PO4 3-) - NFT 0.009+/-0.0063, FRT 0.011+/-0.0068 (Table 1 and Results text agree exactly, p.6-7) - water chemistry, does not route to plant.csv either; Germination index (%) - NFT 65.43+/-9.91, FRT 34.02+/-13.57 (Table 3, cycles 3-4 ONLY - cycles 1-2 not reported, no explanation given) - fits none of plant.csv’s four analyte categories nor any trials.csv column; Daily biomass gain (g/day) - NFT 1.065+/-0.27, FRT 0.822+/-0.15 (Table 2) - a rate, not a total gain, does not fit ‘Fish weight gain’; full economic-viability dataset (Tables 4-5: build costs, solar vs grid electricity, water/energy bills, payback periods) - entirely out of schema scope, not extracted. | Water volume in the system recorded as 200 L (fish tank only, explicitly stated p.3); total system volume (+25 L biological filter; +75 L FRT boxes for FRT; NFT tubular-channel volume never stated) is not summed anywhere in the paper and not computed here (would require an unstated NFT channel volume). | TAN/NH4-N column caveat: paper measures and reports non-ionized ammonia (NH3) specifically, not total ammonia nitrogen (TAN) or NH4-N. No TAN/NH4-N figure is separately given. NH3 value recorded here as the closest available proxy - treat as NH3, not true TAN. | Statistically analysed left Y for both rows since the BLOCKed parameters (pH, water temperature) are secondary environmental monitoring variables, not the study’s primary outcomes (fish and plant growth/yield), none of which carry a BLOCK.