Effect of wastewater from a pikeperch (Sander lucioperca L.) recirculated aquaculture system on hydroponic tomato production and quality

Metadata

• Cite key: delaideEffectWastewaterPikeperch2019
• Item type: Journal Article
• Authors: B. Delaide, S. Teerlinck, A. Decombel, P. Bleyaert
• Affiliation: INAGRO, Ieperseweg 87, Rumbeke-Beitem, B-8800, Belgium
• Journal: Agricultural Water Management 226 (2019) 105814
• Date: 12/2019
• Date added: 2021-02-04
• DOI: 10.1016/j.agwat.2019.105814
• Funding: INAPRO project, European Union’s Seventh Framework Programme for research, technological development and demonstration, grant agreement number 619137; partly funded by the European Regional Development Fund via Aquavlan2 (Interreg V program Flanders - The Netherlands)
• URL: https://doi.org/10.1016/j.agwat.2019.105814
• PDF: Delaide et al. - 2019 - Effect of wastewater from a pikeperch (Sander luci.pdf

Opinion

A methodologically careful, semi-practice-scale comparison — genuine randomized plots, 3 years of replication, and an honest statistical treatment of the low degrees of freedom available for the treatment effect (years x treatment interaction). The authors are transparent about a labelling inconsistency in their own Table 2 (EC values that cannot be dS/m as headed) and about several outcomes they didn’t manage to measure (Mn tissue content, root mass, Brix). Good candidate to cite for DAPS tomato yield parity and the BER/Mn finding, though nearly all fish-side production data (FCR, survival, stocking, feed totals) are simply absent because the RAS was an ongoing commercial farm rather than a bounded fish trial.

Abstract

Decoupled aquaponic systems (DAPS) use the wastewater of recirculated aquaculture systems (RAS) as water source for plant production in recirculated hydroponic systems. RAS wastewater is complemented with macro- and micronutrients to obtain equivalent concentrations and pH as in standard hydroponic nutrient solutions (NS). Unlike in single recirculating aquaponic systems, optimal growth conditions can be established in each production part of a DAPS (i.e. fish and plant parts) avoiding compromises. DAPS design seems more adapted for commercial farming operations but feasibility studies on large-scale systems are lacking. Therefore, the production of tomatoes (Solanum lycopersicum L., cv. Foundation) grown in a NS based on complemented pikeperch RAS wastewater (i.e. AP treatment) has been compared to that of tomatoes grown in conventional hydroponic NS (i.e. HP treatment), in semi-practice conditions. During 3 consecutive years, tomatoes were grown on rockwool slabs, in a large-scale Venlo-type climate-controlled greenhouse, using a recirculated drip irrigation system identical to the ones used by the professionals of the hydroponic tomato sector.

While the electroconductivity was significantly higher in the AP treatment due to the presence of NaCl in the RAS wastewater, no significant differences for the total and marketable fruit yields, fruit number, and size were found between the AP and HP treatments. However, while the level of blossom-end rot (BER) varied substantially (0.9–18.6 %) in the HP treatment, it was remarkably constant and low (0.2-0.4%) over the years in the AP treatment, suggesting a beneficial effect of RAS wastewater. Our results clearly indicate the suitability of complemented pikeperch RAS wastewater as feeding water for professional HP tomato production using drip irrigation for DAPS. As RAS water contains a diversity of microorganisms and dissolved organic matter, it is assumed that some of these acted as plant biostimulants and mitigated the salinity stress and the BER symptoms.

Summary

Over three consecutive growing seasons (2015-2017) at a Belgian research greenhouse, the authors compared tomatoes (cv. Foundation, grafted, grown on rockwool with drip irrigation) fed either complemented pikeperch RAS wastewater (decoupled aquaponics, AP) or a conventional hydroponic nutrient solution (HP), each with 4 randomized plots of 10 stems. Macro- and micronutrient concentrations were successfully matched between the two feeding solutions except for sodium and chloride, which were markedly higher in the AP feed because NaCl was periodically added to the fish system for pikeperch health management; this produced a significantly higher electrical conductivity (EC) in the AP nutrient solution and slabs. Despite this added salinity, total and marketable tomato yield, fruit number and fruit diameter did not differ significantly between treatments across the 3 years, and blossom-end rot (BER) was consistently lower and far less variable in the AP treatment (0.2-0.4%) than in HP (0.9-18.6%, significantly lower in 2016 and 2017). The only significant nutrient-concentration difference found in the root-zone slabs (beyond Na/Cl) was manganese, which was consistently less concentrated in the AP slabs, implying greater plant Mn uptake — a factor previously linked to reduced BER — though fruit/leaf Mn content itself was not measured. The authors attribute the apparent salinity tolerance and reduced BER to microorganisms and dissolved organic matter present in the RAS wastewater acting as plant biostimulants, though this remains speculative since no microbial or DOM characterisation was performed in this study.

---

Experiment data

• Location: Inagro research institute, Rumbeke-Beitem, Belgium (50°54’06.2”N, 3°07’28.0”E)
• Design: Two feeding-water treatments (AP: complemented pikeperch RAS wastewater; HP: conventional hydroponic NS) compared under otherwise identical rockwool drip-irrigation culture, in a semi-practice climate-controlled Venlo greenhouse; combined ANOVA over 3 years with year as random variable, treatment tested against the years x treatment interaction; Duncan test for multiple comparisons
• Replicates / n: 4 randomized plots per treatment per year, 10 tomato stems per plot
• Duration: 3 consecutive growing seasons, 06.01.2015 to 07.11.2017 (sown November, transplanted January, harvested March-November each year)
• Organisms: Pikeperch (Sander lucioperca) / Tomato (Solanum lycopersicum), cv. Foundation (2015-2017); cvs. Axxy and Merlice also tested in 2017 only
• Statistics: ANOVA combined over years (year = random effect) + Duncan test for multiple comparisons; AGROVA statistical package (Inagro)
• Marketable tomato yield: 47.3 +/- 2.4 kg/m2 (AP) vs 46.3 +/- 1.8 kg/m2 (HP) — ns, p > 0.05
• Blossom-end rot: 0.2-0.4% (AP, range over 3 years) vs 0.9-18.6% (HP, range over 3 years) — significantly lower in AP in 2016 and 2017 (per-year Duncan test); combined ANOVA found no significant NS-type effect due to a large years x treatment interaction
• Electrical conductivity (EC): feeding NS trial mean 3.40 +/- 0.70 dS/m (AP) vs 3.15 +/- 0.53 dS/m (HP) — significantly higher AP slab EC vs HP slab EC in 2016 (+1.2 dS/m) and 2017 (+1.3 dS/m)
• Manganese (Mn): slab Mn 1.9x less concentrated in AP than HP overall; significant in 2016 and 2017 slabs (Table 2)

---

Water quality (feeding nutrient solution, trial mean, Table 1, “average over the overall experiment”)

This paper: AP feeding NS (n=24): EC 3.40 +/- 0.70 dS/m; pH 6.34 +/- 0.33; TAN 1.16 +/- 0.74 mmol/L; NO3 16.97 +/- 4.59 mmol/L; PO4 1.54 +/- 0.42 mmol/L; K 8.60 +/- 2.87 mmol/L; Ca 6.38 +/- 1.60 mmol/L; Mg 2.61 +/- 0.80 mmol/L; Na 4.32 +/- 3.04 mmol/L; Cl 3.58 +/- 3.14 mmol/L. HP feeding NS (n=24): EC 3.15 +/- 0.53 dS/m; pH 6.24 +/- 0.32; TAN 1.10 +/- 0.66; NO3 17.93 +/- 3.92; PO4 1.47 +/- 0.38; K 8.84 +/- 3.96; Ca 6.60 +/- 1.37; Mg 2.65 +/- 0.72; Na 1.49 +/- 0.35; Cl 0.60 +/- 0.12 mmol/L. Only Na and Cl (and consequently EC) differed meaningfully between treatments; all other macro/micronutrients were not significantly different (p.5-6, Table 1).

Compared with:

• todo Suhl et al. 2016 — NFT tomato in complemented RAS water vs hydroponics, similar yields, lower BER in complemented-RAS treatment. (p.7)
• todo Crappé et al. 2017 — complemented RAS water vs hydroponics with drip irrigation, similar yields but higher BER in RAS-water treatment (attributed to K/Ca imbalance, not RAS water itself). (p.7)
• todo Delaide et al. 2016a/2016b — lettuce in complemented aquaponic solution outperformed hydroponics; consistent with increased root development also observed here in AP tomato slabs (not weighed/scored). (p.7)
• todo Saha et al. 2016 — basil in RAS-derived nutrient solution, higher vegetative yield vs hydroponic control. (p.7)
• todo Goddek and Vermeulen 2018 — lettuce, RAS-based hydroponic vs conventional. (p.7)
• todo Silber et al. 2009; Aktas et al. 2005 — higher fruit Mn content correlated with reduced BER in pepper. (p.7-8)

Manganese and blossom-end rot

This paper: Mn was the only micronutrient significantly different between AP and HP slabs (2016, 2017; Table 2), consistently lower in AP (1.9x less than HP on average across years; 1.7x less than its own feeding NS concentration, whereas HP slab Mn was slightly more concentrated than its feeding NS). No significant AP/HP difference in feeding-NS Mn or in slab pH, ruling out precipitation as the cause. The authors interpret lower AP slab Mn as evidence of greater plant Mn uptake, hypothesized (via cited pepper literature) to be linked to the AP treatment’s markedly lower and more stable BER incidence, but tomato leaf/fruit Mn content itself was not measured in this study — this causal chain is the authors’ own explicit speculation, not a measured link.

Extraction notes

Fish-side data are almost entirely absent. The pikeperch RAS was an ongoing 700 m2 / 160 m3 commercial farm (average fish load 15 kg/m3, feed 56% protein / 16% fat meal, able to produce 2000-4000 kg pikeperch/year, average daily water exchange 15% to meet a NO3 discharge limit of <2.42 mmol/L) rather than a fish trial bounded to this tomato experiment. FCR, SGR, fish size initial/final, feed totals, biomass created, survival, weight gain, and fish trial duration are [not reported] — the paper is silent on all of these for a discrete cohort, not merely non-applicable.

⚠️WARN-MATERIAL Table 2 EC unit. Table 2 is headed “EC (dS/m)” but its slab values (e.g. AP 2015 = 4362 +/- 383, AP 2016 = 6806 +/- 1590, AP 2017 = 6963 +/- 1957) are physically impossible as dS/m (would be extreme brine, ~1000x normal nutrient-solution EC) and are inconsistent in scale with Table 1’s correctly-labelled dS/m values (AP feeding NS = 3.40 +/- 0.70). Fig. 4’s right-hand axis, plotting the same slab EC data, is explicitly labelled µS/cm with a 0-12000 range matching Table 2’s numbers. Cross-checking against the Discussion’s own stated deltas confirms the µS/cm reading: Table 2’s 2016 AP-HP difference (6806-5539=1267) / 1000 = 1.27 dS/m, matching the text’s “+1.2 dS/m in 2016”; the 2017 difference (6963-5593=1370) / 1000 = 1.37 dS/m, matching “+1.3 dS/m in 2017” (p.5). Recoverable with high confidence via this independent cross-check (analogous to the DMS>59 coordinate-recovery case) — Table 2’s header unit is treated as a typo for µS/cm, not used as-is. Not used in any trials.csv cell (this paper’s EC/pH cells are sourced from Table 1’s feeding-NS trial mean instead, see below); recorded here for transparency since Table 2 is otherwise unusable as printed.

⚠️WARN-CHECK Water quality compartment (pH, EC, TAN, NO2/NO3, macro/micronutrients). Two compartments are reported: the feeding nutrient solution (Table 1, “average over the overall experiment,” n=24, a genuine 3-year trial mean) and the root-zone slabs (Table 2, means given separately per year only — 2015 n=6, 2016 n=11, 2017 n=14 — with no combined 3-year slab mean stated anywhere in the paper, only Na and Cl slab concentrations get an explicit combined-average statement in running text). SCHEMA.md’s general preference is the plant-bed/hydroponic-unit value, but no single “trial mean” exists for slab pH/EC/nutrients without averaging the three yearly Table 2 rows myself, which would be derivation. Recorded the feeding-NS (Table 1) trial means instead, since that is the only value matching the required “trial mean +/- SD” format; per-year slab values noted here as the alternative, unused candidate. This paper is added to REVIEW.md for this choice.

⚠️WARN-CHECK NO3/TAN units. Table 1 and Table 2 report nitrogen species in mmol/L (e.g. AP feeding NO3 16.97 +/- 4.59 mmol/L, TAN 1.16 +/- 0.74 mmol/L), not mg/L NO3-N or mg/L TAN as the trials.csv schema columns require. Converting mmol/L to mg/L N requires applying molar mass (a stoichiometric constant not stated in the paper) and was judged out of scope for permitted “unit conversion only” (kg<->g, weeks<->days, DMS<->decimal are the sanctioned examples; this is closer to the explicitly-flagged NO3-vs-NO3-N ambiguity). TAN / NH4-N, NO2-N and NO3-N columns therefore recorded NR; the paper’s own mmol/L values are given above under Water quality and in the trials.csv Experimental Remarks.

⚠️WARN-CHECK Plants/m2 vs stems/m2. The paper states a plant density of “2.27 stems per square meter” derived explicitly from its own stated inter-plant spacing (0.55 m x 0.80 m gutter spacing) with 2 stems/plant (p.2) — i.e., 1.135 plants/m2 if halved, but the paper never states the halved plants/m2 figure itself, only stems/m2. Separately, for 2015-2016, the paper states 498 HP stems on 180 m2 (2.77 stems/m2) and 411 AP stems on 149 m2 (2.76 stems/m2) — both different from the explicitly stated 2.27 stems/m2 design density. No basis to prefer one area/stem-count figure over the other; Plants/m2 recorded NR rather than deriving or choosing among these. This paper is added to REVIEW.md for this discrepancy.

⚠️WARN-MINOR “Na concentration in AP feeding solution (?)” — p.5 body text includes a literal ”(?)” after “In the HP feeding solution Na concentration” with no footnote or explanation; likely an editing artifact left in the published text. The stated average (1.49 mmol/L) matches Table 1’s HP feeding Na value exactly, so no cell is affected.

⚠️WARN-MINOR Final stem length reported to inconsistent decimal precision: “868.6 cm” (AP) vs “878.64 cm” (HP), p.5. Not used in any trials.csv cell (see below), no downstream effect.

[not reported]: Fish size initial/final, FCR, SGR, feed routine/schedule, total feed (kg), fish biomass created, fish survival rate, fish weight gain, fish trial duration (days), Fish Category, Water recycle (L/min), Water temperature, Dissolved Oxygen, Days Plant after transplant, Plants/m2 (see CHECK above), SPAD, Plant height (see NO COLUMN below), Leaf count, Plant fresh weight, Plant dry matter, Tissue nitrate AP/HYD, FUE AP/HYD, WUE, pHOptimal, Plant Category, Water classification, Biological system already in use, Air supplement, Remineralization, Artificial Lighting.

[unclear]: none beyond the items flagged above.

NO COLUMN items (routed to Experimental Remarks in trials.csv, not entered as cells):

• Total yield (as distinct from marketable yield): 48.3 +/- 2.4 kg/m2 (AP) vs 48.0 +/- 1.8 kg/m2 (HP), ns — the schema’s singular AP/HYD columns were populated with marketable yield instead (the graphed, more commercially standard metric); this is a different-metric pairing, not a contradiction.
• Final stem length (AP 868.6 cm vs HP 878.64 cm over 3 years, ns) — not entered as Plant height because the paper measures and reports cumulative stem length on a trained/topped indeterminate crop, not a height, and no explicit “height” term or method is used.
• Fruit diameter: overall average 67 +/- 3 mm, constant across years and treatments, no significant AP/HP difference (values not split by treatment in text).
• RAS annual production capacity: 2000-4000 kg pikeperch/year (a facility capacity range, not a measured biomass-created figure for this study’s period).
• Brix measurements: methodology described in Statistical analysis (2.7, n=120, compared via Duncan test) but no Brix results were found anywhere in Results, Discussion, or Conclusion — an apparent reporting gap, not a contradiction.
• Slab Na/Cl concentrations (trial-wide averages given in running text, p.5): Na 11.48 +/- 7.32 mmol/L (AP slabs) vs 4.44 +/- 1.71 mmol/L (HP slabs); Cl deviation from Na by an approximately constant amount, -4.89 mmol/L (AP) vs -4.25 mmol/L (HP) — no dedicated trials.csv column for slab-specific ionic detail beyond the feeding-NS values already recorded.
• Root observations: “slightly more roots were observed on the side of the AP slabs” at end of cultivation, not weighed or scored (p.7) — qualitative only.

Water panel note: No water-quality panel was excluded from plant.csv — this paper has no plant-tissue analyte data at all (no leaf/fruit mineral, biochemistry, microbiology, or proximate analysis was performed; the Mn-uptake conclusion is inferred entirely from nutrient-solution depletion in slabs, not from plant tissue), so plant.csv is empty by design, not because a usable panel was discarded.

Linked claims

• Aquaponic wastewater can match hydroponic yields when nutrients are complemented
• Aquaponic wastewater can reduce blossom-end rot in tomato
• RAS wastewater microorganisms and dissolved organic matter may act as plant biostimulants
• Manganese uptake is linked to reduced blossom-end rot

Citations to chase

• todo Suhl et al. (2016) — NFT tomato, complemented RAS water vs hydroponics, similar yield, lower BER
• todo Crappé et al. (2017) — complemented RAS water vs hydroponics with drip irrigation, similar yields, higher BER (K/Ca imbalance)
• todo Delaide et al. (2016a, 2016b) — lettuce in complemented aquaponic solution outperforms hydroponics
• todo Saha et al. (2016) — basil, RAS-derived nutrient solution vs soilless control
• todo Goddek and Vermeulen (2018) — lettuce, RAS-based vs conventional hydroponic systems
• todo Silber et al. (2009); Aktas et al. (2005) — Mn nutrition and blossom-end rot in pepper

---

Source: Delaide et al. - 2019 - Effect of wastewater from a pikeperch (Sander luci.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

delaideEffectWastewaterPikeperch2019-T1

Fish

Field	Value
Fish	Pikeperch (Sander lucioperca L.)
Initial Stock density	15 (average fish load, not explicitly ‘initial’; see remarks)
Protein	56
Feed regime	Fish protein-based meal, 56% protein and 16% fat (Skretting, Fontaine les Vervins, France)

Water

Field	Value
Water volume in the system	160000 (RAS total facility volume; hydroponic loop total not stated as a single figure, see remarks)
Water type	RAS wastewater (pikeperch), complemented with macro/micronutrients (p.2)
Daily Water exchange rate	15 (RAS/fish system; see remarks)
Aq pH	6.34 +/- 0.33
EC	3.40 +/- 0.70

Plant

Field	Value
Plant	Tomato (Solanum lycopersicum L.), cv. Foundation (also cvs. Axxy and Merlice, 2017 only)
Details	Grafted on rootstock Maxifort (2015) or DR0141TX (2016-2017); sown November, transplanted January (first truss visible), harvested March-November; 2 stems/plant, topped early September; slabs renewed yearly

System & Setup

Field	Value
System type	Rockwool slab, drip irrigation (Venlo-type climate-controlled greenhouse)
Media Details	Grodan Vital rockwool slabs 1 x 0.20 x 0.075 m, 2 plants/slab, one dripper/plant, slabs on drainage gutters (Meteor Systems), 0.55 m between-plant spacing on gutters x 0.80 m between gutters (2.27 stems/m2 design density, 2 stems/plant)
Iron supplemented	Y (DTPA-Fe included in both AP and HP commercial fertiliser mix (Yara); feeding NS Fe trial mean 45.01 +/- 13.59 umol/L (AP) vs 22.45 +/- 7.50 umol/L (HP), Table 1)
Climate control	Y (Venlo-type climate-controlled greenhouse, PRIVA Connect computer software; heating setpoints 20.0C day / 15.5C night targeting 18.0C average; ventilation opened at 22C (19C from mid-May to September))
Nutrient supplemented	Y (RAS wastewater complemented with commercial fertilisers (Yara: Amnitra, Calsal, Magnesul, Baskal, BFK, Sulfakal, SZ-38, DTPA-Fe, Fervent Mn/Zn/B/Cu/Mo) to match HP nutrient solution targets: macronutrients (mmol/L) 16 NO3, 1.2 TAN, 1.4 PO4, 9.5 K, 5.4 Ca, 2.4 Mg, 4.4 SO4; micronutrients (umol/L) 30 B, 15 Fe, 0.75 Cu, 10 Mn, 5 Zn, 0.5 Mo)
Equipment	TAF filter (TAF 750, Amiad); sand filter + UV disinfection unit for drain water reuse; Nutronic dosing device (PRIVA); PRIVA Connect climate/irrigation software; conductivity tester (WTW ProfiLine Cond 3110); pH meter (WTW ProfiLine pH 3110); continuous flow analyser (Skalar SFA 4000) for N species; ICP-OES (PerkinElmer Optima 8300) for cations; ion chromatography (Metrohm 850 Professional IC) for Cl/SO4
Control Parameters	NS pH target 5.6 (both treatments); drain water quantity target 20-40% of inflow; RAS discharge NO3 limit <2.42 mmol/L; EC not actively controlled/matched in AP (2016-2017, higher AP EC tolerated for better drain-water reuse; 2015 constant EC aimed in both)
Combination	Pikeperch (RAS, professional-scale) and tomato (semi-practice greenhouse); decoupled aquaponic (DAPS) wastewater-fed hydroponics vs conventional hydroponic nutrient solution, both on rockwool drip irrigation

Site

Field	Value
Region	Europe
Country	Belgium
Lat	50.9017
Long	3.1244
Average room Temperature	18.0 (targeted average, not measured mean; see remarks)

Results & Statistics

Field	Value
Measured Unit	kg/m2 (yield); dS/m (EC); mmol/L (nutrient concentrations); % (BER incidence)
Statistic Details	AGROVA statistical package (Inagro); combined ANOVA over 3 years (years = random variable), treatment tested against years x treatment interaction; Duncan test for multiple comparisons
Statistically analysed	Y
Replicates (n)	4
AP	47.3 +/- 2.4 (marketable yield, kg/m2)
HYD	46.3 +/- 1.8 (marketable yield, kg/m2)

Experimental Remarks: TRIAL DEFINITION: T1 = aquaponic (AP) treatment (tomato fed complemented pikeperch RAS wastewater), paired control = conventional hydroponic (HP) treatment (80% rainwater + 20% tap water plus commercial fertilisers), recorded in the HYD-labelled columns/cells. Single AP treatment repeated over 3 years (2015-2017) with year treated as a random effect in one combined ANOVA (not three separate arms), so one row. | WARN-MATERIAL Table 2 EC unit: headed ‘(dS/m)’ but values (e.g. AP 2016 6806 +/- 1590) are physically impossible as dS/m and match Fig. 4’s uS/cm axis; confirmed via Discussion deltas (+1.2 dS/m 2016, +1.3 dS/m 2017 = Table 2 differences /1000). Not used for any cell (EC/pH cells here sourced from Table 1 feeding-NS trial mean instead); full detail in the note. | WARN-CHECK Water quality compartment: Aq pH/EC recorded from Table 1 feeding NS overall trial mean (n=24); Table 2 gives only per-year slab means (2015 n=6/2016 n=11/2017 n=14), no combined slab trial mean stated, so slabs not used (would require averaging years myself). Slab AP/HP pH and EC per year, and slab Na (11.48 +/- 7.32 AP vs 4.44 +/- 1.71 HP mmol/L) and Cl (-4.89 AP vs -4.25 HP mmol/L, deviation from Na) given in the note. Added to REVIEW.md. | WARN-CHECK NO3/TAN units: paper reports NO3 and TAN in mmol/L (AP feeding: TAN 1.16 +/- 0.74, NO3 16.97 +/- 4.59 mmol/L; HP feeding: TAN 1.10 +/- 0.66, NO3 17.93 +/- 3.92 mmol/L), not mg/L NO3-N/TAN as required by the schema; molar-mass conversion judged out of scope for permitted unit conversion, so TAN/NH4-N, NO2-N, NO3-N left NR. | WARN-CHECK Plants/m2 vs stems/m2: paper states plant density 2.27 stems/m2 from stated spacing (2 stems/plant, so ~1.135 plants/m2 if halved, not stated directly); separately, 2015-2016 areas/stem-counts compute to ~2.77 stems/m2 (498 HP stems/180 m2; 411 AP stems/149 m2), inconsistent with the stated 2.27. No basis to prefer one; Plants/m2 recorded NR rather than derive/choose. Added to REVIEW.md. | UNIT CONVERSION ONLY: coordinates 50 deg 54’ 06.2” N / 3 deg 07’ 28.0” E -> decimal 50.9017 / 3.1244; RAS water volume 160 m3 -> 160000 L (RAS total facility volume, not hydroponic loop total, which is not stated as a single figure - components given: 0.4 m3 feeding sump tank + 1 m3 drain storage tank per treatment). | NOT DERIVED, left NR: fish stocking density beyond the stated 15 kg/m3 average RAS load (no per-batch initial value given); FCR; SGR; feed N/P/K; % of body weight ration; fish size initial/final; total feed (kg); fish biomass created (RAS states a general 2000-4000 kg/year production capacity, not a measured value for this study period); fish survival rate; fish weight gain; fish trial duration (RAS operated continuously as a commercial farm, not bounded to the tomato trial); Water recycle (L/min, not stated for the hydroponic loop); Water temperature; Dissolved Oxygen; Days Plant after transplant (harvest spans March-November after January transplant, no single figure given); SPAD; Plant height (paper reports final stem length 868.6 cm AP vs 878.64 cm HP, ns, not termed ‘height’ and not used, see note); Leaf count; Plant fresh weight per plant; Plant dry matter; Tissue nitrate AP/HYD (no tissue analysis performed); FUE AP/HYD; WUE; pHOptimal; Fish Category; Plant Category; Water classification. | Initial Stock density recorded as 15 kg/m3, the RAS’s stated ‘average fish load’ (p.2) - not explicitly termed an ‘initial’ stocking value, since the RAS is a continuously operated professional farm rather than a single stocked cohort; no other density figure is given. | Average room Temperature recorded as 18.0C, the paper’s stated greenhouse target/aim, not an explicitly measured mean. | Daily Water exchange rate (15%) is the RAS/fish-side value (to meet a NO3 discharge regulation), not a stated hydroponic-loop exchange rate; the hydroponic drain-water reuse target (20-40% of inflow) is a different, unrelated parameter, given here for context only. | AP/HYD columns use marketable yield (kg/m2); total yield also reported (48.3 +/- 2.4 AP vs 48.0 +/- 1.8 HP kg/m2, ns) as an alternative, unused metric - not a contradiction, both explicitly labelled and reported for different definitions. | NO COLUMN: fruit diameter overall average 67 +/- 3 mm (constant across years/treatments, not split by treatment in text); Brix methodology described (n=120, Duncan test) but no Brix results reported anywhere in the paper; qualitative note of ‘slightly more roots’ observed on AP slabs, not weighed/scored; RAS annual production capacity 2000-4000 kg pikeperch/year (facility capacity, not a measured value for this study).