Hydroponic and Aquaponic Floating Raft Systems Elicit Differential Growth and Quality Responses to Consecutive Cuts of Basil Crop
Metadata
- Cite key: modarelliHydroponicAquaponicFloating2023
- Item type: Journal Article
- Authors: G.C. Modarelli, L. Vanacore, Y. Rouphael, A.L. Langellotti, P. Masi, S. De Pascale, C. Cirillo
- Affiliation: Department of Agricultural Sciences, University of Naples Federico II, Portici, Italy; Centre for Innovation and Development in the Food Industry (CAISIAL), University of Naples Federico II, Portici, Italy
- Journal: Plants 12 (2023) 1355
- Date: 03/2023
- Date added: [not reported]
- DOI: 10.3390/plants12061355
- Funding: European Union’s Horizon 2020 research and innovation programme, grant agreement no. 862663 (Food Systems in European Cities, FoodE)
- URL: https://doi.org/10.3390/plants12061355
- PDF:
Modarelli et al. - 2023 - Hydroponic and Aquaponic Floating Raft Systems Elicit Differential Growth and Quality Responses to C.pdf
Opinion
A tightly controlled, well-replicated 2x2 factorial (system x cut) with a genuinely paired hydroponic control run in the same greenhouse — methodologically one of the more rigorous basil aquaponics-vs-hydroponics comparisons in the vault. However, the paper contains a confirmed misattribution (a results sentence labels Total Dry Weight and Total Dry Matter percentage changes as “leaf area” changes) and a near-certain unit error in the nitrate table (g kg⁻¹ f.w. where mg kg⁻¹ f.w. is clearly intended, given the paper’s own regulatory-limit comparison). Neither error invalidates the raw tabulated means, which are internally self-consistent, but they lower confidence in the prose and mean values quoted from this paper elsewhere should be checked against the tables directly rather than trusted from the running text.
Abstract
Basil crops are appreciated for their distinct flavour and appeal to various cuisines globally. Basil production is mainly implemented in controlled environment agriculture (CEA) systems. Soil-less cultivation (e.g., hydroponic) is optimal for producing basil, while aquaponics is another technique suitable for leafy crops such as basil. Shortening the production chain through efficient cultivation techniques reduces basil production’s carbon footprint. While the organoleptic quality of basil demonstrably benefits from successive cuts, no studies have compared the impact of this practice under hydroponic and aquaponic CEA conditions. Hence, the present study evaluated the eco-physiological, nutritional, and productive performance of Genovese basil cv. Sanremo grown in hydroponic and aquaponic systems (combined with tilapia) and harvested consecutively. The two systems showed similar eco-physiological behaviour and photosynthetic capacity, which were on average 2.99 µmol of CO2 m−2s−1, equal numbers of leaves, and fresh yields of on average 41.69 and 38.38 g, respectively. Aquaponics yielded greater dry biomass (+58%) and dry matter content (+37%), while the nutrient profiles varied between the systems. The number of cuts did not influence yield; however, it improved dry matter partitioning and elicited a differential nutrient uptake. Our results bear practical and scientific relevance by providing useful eco-physiological and productive feedback on basil CEA cultivation. Aquaponics is a promising technique that reduces chemical fertiliser input and increases the overall sustainability of basil production.
Summary
The study grew Genovese basil (cv. Sanremo) side by side in a floating-raft hydroponic system and a floating-raft aquaponic system (coupled to a recirculating tilapia RAS) in a shaded Mediterranean greenhouse, harvesting each system twice (35 and 79 days after planting) to test whether cultivation system and cut number interact. Growth (leaf number, fresh weight), gas exchange, chlorophyll fluorescence, photosynthetic pigments, SPAD, and leaf mineral content (NO3, P, K, S, Ca, Mg, NH4, Na, Cl) were measured for each system x cut combination with n=3 replicates. Cut number, not cultivation system, was the dominant driver of most eco-physiological and growth responses — gas exchange and photochemical efficiency fell sharply at the second cut regardless of system, while dry biomass and dry matter rose in both. The cultivation system mainly affected mineral nutrition: aquaponics produced higher leaf Mg, Ca and carotenoids and lower P than hydroponics, and by the second cut aquaponics had accumulated substantially more dry biomass and dry matter than hydroponics despite similar fresh weight. The authors conclude aquaponics can match hydroponic basil yield while reducing mineral fertiliser dependence, though they flag several extraction issues in this note (see Extraction notes) that mean specific headline percentages quoted in the abstract/discussion should not be taken at face value without checking the tables.
Experiment data
- Location: Department of Agricultural Sciences, University of Naples Federico II, Portici, Italy; cold greenhouse, 40.8161°N 14.3504°E, 29 m a.s.l.
- Design: 2 (System: Hydroponic vs Aquaponic floating raft) x 2 (Cut: 35 DAP, 79 DAP) factorial; n = 3 replicates per cell (each replicate = mean of pooled plants)
- Replicates / n: 3 (stated explicitly, Table 1-4 footnotes and section 4.9)
- Duration: 24 June to 10 September 2021 (calendar dates only); plant cuts at 35 and 79 days after planting (DAP)
- Organisms: Oreochromis niloticus (tilapia) / Ocimum basilicum cv. Sanremo (Genovese basil)
- Statistics: Two-way ANOVA (System x Cut), SPSS v.27; Tukey’s HSD post hoc, p<0.05; cluster heatmap via ClustVis (Euclidean distance, complete linkage, log(x+1)-transformed)
- Plant fresh weight: AQ 32.36±1.23 g/plant (Cut I) / 44.38±4.44 g/plant (Cut II); H 38.05±1.67 g/plant (Cut I) / 38.75±2.83 g/plant (Cut II)
- Plant dry matter: AQ 3.64±0.04% (Cut I) / 11.22±0.11% (Cut II); H 2.98±0.06% (Cut I) / 7.91±0.13% (Cut II)
- Tissue nitrate (NO3): AQ 1565.55±156.84 mg/kg fw (Cut I) / 3478.24±321.61 mg/kg fw (Cut II); H 523.29±229.17 mg/kg fw (Cut I) / 3311.48±148.39 mg/kg fw (Cut II) — unit corrected from paper’s stated “g kg⁻¹ f.w.”, see Extraction notes
Growth and yield
This paper: Leaf number and total fresh weight did not differ significantly between hydroponic and aquaponic systems (S main effect ns for both). Total leaf area showed a significant System x Cut interaction: it fell 57% between cuts in hydroponics but rose 48% in aquaponics. Total dry weight and dry matter content rose sharply between cuts in both systems, more so in aquaponics (+58% dry biomass, +37% dry matter content vs hydroponics, per abstract).
Compared with:
- todo Saha Monroe Day 2016 — compared basil in hydroponic vs aquaponic (crayfish) systems, but without a cut-number factor, so a direct comparison with this paper was not possible per the authors (p.6).
- todo Ferrarezi Bailey 2019 — consecutive-cut effects on basil yield under tropical aquaponics; after the fourth cut, production declined (cited p.2, secondary).
Eco-physiology and gas exchange
This paper: Cut number, not cultivation system, was the dominant factor for gas exchange and chlorophyll fluorescence (Pn, gs, E, Fv/Fm, ΦPSII, ETR all fell at the second cut; ΦNO rose 60%). Between systems, hydroponic plants had 14% higher transpiration and 25% higher ETR; no other gas-exchange/fluorescence parameter differed by system (Table 2).
Compared with: none directly cited for gas exchange comparisons in this section of the paper.
Leaf mineral content
This paper: Nitrate rose sharply at the second cut in both systems (see Extraction notes for a flagged discrepancy between the stated percentage increases and the tabulated values). Phosphorus was higher in the first cut and higher in hydroponics than aquaponics. Magnesium and calcium were higher in aquaponics than hydroponics (Mg +108%, Ca higher throughout). Potassium, sulphate, ammonium and chloride showed system- and cut-specific patterns (Table 3).
Compared with:
- todo Ciriello Formisano Corrado Nicoletto (cited collectively p.7 as refs [3,5,6,9,33]) — nitrate content values in hydroponic Genovese basil under similar cut regimes, said to align with this paper’s results.
Linked claims
- Aquaponics can match hydroponic yield while reducing mineral fertiliser input
- Cut number is a stronger driver of basil eco-physiology than cultivation system
- Aquaponic basil accumulates more dry matter than hydroponic basil across successive cuts
Citations to chase
- todo Saha, S.; Monroe, A.; Day, M.R. (2016) — Growth, yield, plant quality and nutrition of basil under soilless agricultural systems (hydroponic vs aquaponic with crayfish); no cut-number factor, direct comparison not possible per this paper’s authors.
- todo Ferrarezi, R.S.; Bailey, D.S. (2019) — Basil performance evaluation in aquaponics (tropical); consecutive-cut/harvest-period effects on yield, production declines after the fourth cut.
- todo Ciriello, M.; Formisano, L.; El-Nakhel, C.; Corrado, G.; Pannico, A.; De Pascale, S.; Rouphael, Y. (2021) — Morpho-physiological responses and secondary metabolites modulation by preharvest factors of three hydroponically grown Genovese basil cultivars (nitrate content comparison, cited p.7).
- todo Khalil, S. (2018) — Growth performance, nutrients and microbial dynamic in aquaponics systems as affected by water temperature (cited p.5 re: nitrate-rich water and bacterial conversion).
Extraction notes
⚠️BLOCK (misattribution): Results section 2.1 (p.2) states “In the second cut, the leaf area increased by 182% and 365% and by 166% and 208%, respectively, under hydroponic and aquaponic conditions” — attributing all four percentages to “the leaf area.” Recomputation from Table 1 shows 182%/365% actually match Total Dry Weight (H: 1.16→3.26 g DW plant⁻¹ = +181.0%; AQ: 1.18→5.5 g DW plant⁻¹ = +366.1%) and 166%/208% match Total Dry Matter (H: 2.98→7.91% = +165.4%; AQ: 3.64→11.22% = +208.2%) — not Total Leaf Area, which the same paragraph had already correctly described three sentences earlier (−57% H / +48% AQ). Total Leaf Area and absolute Total DW have no dedicated trials.csv column, so no cell is overwritten by this error; the Plant dry matter (%) cells used in both trial rows are sourced directly from Table 1 (independently corroborated by this same recomputation), not from the mislabelled sentence. Flagged per the vault’s rule that any confirmed misattribution forces quality: suspect regardless of downstream cell impact.
⚠️MATERIAL: Table 3’s nitrate (NO3) column is headed “g kg⁻¹ f.w.” while every other mineral in the same table is “g kg⁻¹ d.w.” As printed, the reported NO3 values (523–3478 “g kg⁻¹”) are physically impossible (52%–348% of the tissue’s own fresh mass). The Discussion (p.7) states “considering the limit allowed for leafy vegetables, in our growing condition, we were less than that limit in both systems and cut periods” — true only if the values are mg kg⁻¹ f.w. (typical regulatory leafy-vegetable nitrate limits are low thousands of mg/kg), not g kg⁻¹. Recorded Tissue nitrate AP/HYD in mg/kg fw using the printed numeric values unchanged, treating this as a unit-label correction (not a derivation) supported by the paper’s own regulatory-comparison sentence. Affects both trial rows’ Tissue nitrate cells.
⚠️MATERIAL: Abstract states hydroponic and aquaponic “fresh yields of on average 41.69 and 38.38 g, respectively.” Table 1 gives H Cut I=38.05±1.67, H Cut II=38.75±2.83 (Table-stated Mean=38.4, internally consistent with a simple average of the two cuts) and AQ Mean=38.37 (matches the abstract’s 38.38 within rounding) — but no combination of Table 1’s hydroponic values reaches 41.69. The Table 1 per-cut values (used directly in the Plant fresh weight cells of both trial rows) are recorded as the defensible source since they are more granular and internally self-consistent; the abstract’s hydroponic mean is flagged as unreliable. Does not change any extracted cell, since per-cut Table 1 values are used directly rather than the abstract’s system-level mean.
⚠️MATERIAL: Results 2.3 (p.3) states nitrate “increased by 1171% and 380% in hydroponic and aquaponics conditions, respectively” between cuts. Recomputing from Table 3’s own Cut I/II values gives H: (3311.48−523.29)/523.29 = 532.8%, AQ: (3478.24−1565.55)/1565.55 = 122.2% — neither matches 1171%/380% under any combination tried, including a swapped assignment. All other percentage-change claims in the same paragraph (P −38%, K +19% H, Mg +108% AQ vs H, Ca +104% H, Cl +133% pooled) were independently recomputed and matched Table 3 closely, so this is not a general methodological offset — the nitrate percentages specifically appear to be a calculation error by the authors. No trials.csv cell is derived from the narrative percentage (the Tissue nitrate cells use the raw Table 3 readings), so no cell is affected, but flagged for reader awareness when citing this paper’s nitrate trend.
[not reported] fields (grouped): Fish Category; Water type; Water classification; FCR; SGR; feed N/P/K composition (only 42% crude protein given); Total Feed (kg); Fish biomass created; Fish survival rate; Fish weight gain; Fish size initial/final; Fish trial duration in days (only calendar dates given); Water recycle flow rate; Water volume in the system (component volumes given, not summed); Daily Water exchange rate; FUE AP/HYD; WUE; Dissolved Oxygen; TAN/NH4-N, NO2-N, NO3-N (water); Plant height; Average room Temperature (only an instantaneous noon reading during gas-exchange measurements is given, not a trial-mean); Biological system already in use; Iron supplemented; Remineralization; Artificial Lighting; Nutrient supplemented.
[unclear] fields: none — ambiguities encountered were resolvable as either “not a contradiction” (setpoint vs. monitored mean for pH/EC/temperature) or were escalated to the ⚠️ flags above rather than left silently unclear.
NO COLUMN items (in Experimental Remarks): Total Leaf Area (cm² plant⁻¹, Table 1); absolute Total Dry Weight (g DW plant⁻¹, Table 1); Specific Leaf Area (cm²/g, Table 1); full gas-exchange/chlorophyll-fluorescence panel (Pn, gs, E, Fv/Fm, ΦPSII, ETR, ΦNO, NPQ, WUEi, LUE, RWC — Table 2, no matching category in either trials.csv or plant.csv’s biochemistry/mineral/microbiology/proximate scheme); hydroponic paired plant-growth values (Leaf number, Total FW, Total DM) for each cut, since trials.csv growth columns are single (AQ-focused) rather than AP/HYD-paired.
Judgment call on Plant Category: p.2 background text mentions basil among “leafy vegetables” grown in aquaponics generally, but this is introductory/contextual, not the paper’s own categorisation of its basil crop — recorded NR per the prime directive rather than borrowing the general-discussion phrasing.
Judgment call on trial splitting: treated System x Cut as two aquaponic trials (T1 = Cut I, T2 = Cut II) rather than one, since the paper reports both cuts as independently replicated, separately tabulated treatment cells with a tested System x Cut interaction (not merely a restated duration). Fish-side data (tilapia, one continuous RAS, four weight classes) and the hydroponic Hoagland recipe are shared/identical across both rows since the same system ran continuously through both cuts.
Source: Modarelli et al. - 2023 - Hydroponic and Aquaponic Floating Raft Systems Elicit Differential Growth and Quality Responses to C.pdf
Data Tables
Structured data extracted from this paper into the vault's
trials.csv/plant_measurements.csvdatasets. Fields the paper didn't report are omitted. Download the full datasets (measurements).
Trial Parameters
modarelliHydroponicAquaponicFloating2023-T1
Fish
| Field | Value |
|---|---|
| Fish | Tilapia (Oreochromis niloticus) |
| Initial Stock density | 8.7 +/- 5.4 |
| Protein | 42 |
| % of body weight | 1.4 +/- 0.5 |
| Feed regime | Preformulated fish feed, Tilapia Grower 13-EF (Alltech Coopens, Helmond, NL), 42% protein content; daily feed target adjusted based on fish weight and stocking |
Water
| Field | Value |
|---|---|
| Aq pH | 6.9 |
| EC | 1.1 |
| Water temperature | 23 |
Plant
| Field | Value |
|---|---|
| Plant | Genovese basil (Ocimum basilicum L. cv. Sanremo) |
| Details | Cut I (first harvest), 35 days after planting (DAP); cut with scissors leaving 2 buds below |
| Days Plant after transplant | 35 |
| Plants/m2 | 20 |
| SPAD (aquaponics) | 19.41 +/- 0.94 |
| Leaf count | 33.33 +/- 0.71 |
| Plant fresh weight | 32.36 +/- 1.23 |
| Plant dry matter | 3.64 +/- 0.04 |
| Tissue nitrate AP | 1565.55 +/- 156.84 |
| Tissue nitrate HYD | 523.29 +/- 229.17 |
System & Setup
| Field | Value |
|---|---|
| System type | Floating raft system (title; p.7-8, floating hydroponic raft / floating raft units) |
| Media Details | Floating raft units, 2 m2 each (AQ, connected by 1 loop to the RAS); disconnected/monitored separately for H; two-week-old seedlings on polystyrene sowing trays, peat cube washed off before transplant |
| Air supplement | Y (Ambient air insufflation in RAS rearing tanks at 0.05 v/v/min (p.7)) |
| pH Buffers | Y (Hydroponic nutrient solution pH adjusted with nitric acid or KOH as needed (p.8); not stated whether the aquaponic loop was separately buffered) |
| Climate control | Y (Cold greenhouse shaded with 75% black shade net (p.7); no active heating/cooling system stated) |
| Equipment | RAS: 4x2800L rearing tanks; 800-L Superbead mechanical/biological filter (Air-aqua, Staphorst, NL); 400-L trickling filter (Scubla srl, Udine, Italy); 40-W UV steriliser (Air-aqua); Thermo Scientific Expert pH and Cond Testers; LCi T photosynthesis yield analyser (ADC Bioscientific Ltd.); Opti-Sciences Plant Stress Kit fluorimeter + Fv/Fm meter; SPAD-502 chlorophyll meter (Konica Minolta); Hach DR 2000 spectrophotometer; Dionex ICS-3000 ion chromatograph with IonPac CS12A/AS11-HC columns |
| Control Parameters | Water temperature setpoint 23 degC; pH ~6.9 and EC ~1100 uS/cm (1.1 dS/m) monitored daily, stated once generally for ‘the system’ (section 4.2) and re-stated as the hydroponic nutrient solution’s own composition (section 4.3) — read as a shared target/observed value, not a conflict (setpoint vs. monitored mean; target vs. achieved are not contradictions per schema); gas exchange measured at noon, ambient CO2 434 ppm, PPFD 1251.8 umol m-2 s-1, ~31.1 degC/45% RH; greenhouse shaded with 75% black net |
| Combination | Tilapia (Oreochromis niloticus) + Genovese basil cv. Sanremo; aquaponic vs hydroponic floating raft, two consecutive cuts |
Site
| Field | Value |
|---|---|
| Region | Southern Europe |
| Country | Italy |
| Lat | 40.8161 |
| Long | 14.3504 |
Results & Statistics
| Field | Value |
|---|---|
| Measured Unit | g FW plant-1; g DW plant-1; % DM (dry matter); n plant-1 (leaf number); SPAD units; mg/kg fw (tissue nitrate, unit-corrected, see remarks) |
| Statistic Details | Two-way ANOVA (System x Cut) with interaction, SPSS v.27; Tukey’s HSD post hoc p<0.05; n=3 replicates (6 plants pooled/replicate); cluster heatmap via ClustVis, Euclidean distance, complete linkage, log(x+1)-transformed data |
| Statistically analysed | Y |
| Replicates (n) | 3 |
Experimental Remarks: TRIAL DEFINITION: this row = Aquaponic floating raft treatment (AQ), Cut I (35 DAP), n=3 replicates (6 plants pooled/replicate, section 4.9). Paired control = Hydroponic floating raft treatment (H), Cut I, recorded in HYD-equivalent columns/remarks where a column exists (Tissue nitrate HYD has a dedicated column; other HYD growth figures are noted below since the growth columns are single/AQ-focused). This paper crosses System (H vs AQ) x Cut (I vs II, i.e. 35 vs 79 DAP) as a genuine 2x2 factorial with independently replicated, separately tabulated treatment cells (Tables 1-4 report all 4 cells with own SE and Tukey letters, plus significant Cut and System x Cut interaction terms) — treated as TWO aquaponic trials rather than one; see note’s ‘Judgment call on trial splitting’ for full reasoning. Fish (tilapia, Oreochromis niloticus, 4 weight classes, one continuous RAS unit) and the hydroponic Hoagland solution are shared/continuous across both cuts, not separately stocked per cut, so all fish-side and water-setpoint cells are identical between this row and the Cut II (79 DAP) row. | WARN-MATERIAL Tissue nitrate unit: Table 3 header states NO3 in ‘g kg-1 f.w.’ while every other mineral in the same table is ‘g kg-1 d.w.’; as printed, the reported NO3 values (523-3478 ‘g kg-1’) are physically impossible (52%-348% of the tissue’s own fresh mass). Discussion (p.7): ‘considering the limit allowed for leafy vegetables, in our growing condition, we were less than that limit in both systems and cut periods’ — true only if values are mg kg-1 f.w. (typical leafy-vegetable regulatory nitrate limits are low thousands of mg/kg), not g kg-1. Recorded Tissue nitrate AP/HYD in mg/kg fw using the printed numeric values unchanged (unit-label correction, not a derivation), supported by the paper’s own regulatory-limit sentence. Affects Tissue nitrate AP and Tissue nitrate HYD in both trial rows. | WARN-MATERIAL Abstract vs Table 1 (hydroponic mean fresh weight): abstract states H and AQ ‘fresh yields of on average 41.69 and 38.38 g, respectively.’ Table 1: H Cut I=38.05+/-1.67, H Cut II=38.75+/-2.83, Table-stated H Mean=38.4 (consistent with simple average of the two cuts); AQ Mean=38.37 (matches abstract’s 38.38 within rounding) — but no combination of Table 1’s hydroponic values reaches 41.69. Table 1’s per-cut values (used directly in this row’s Plant fresh weight cell) are recorded as the defensible source since they are more granular and internally self-consistent; abstract’s H mean flagged unreliable. Does not change any extracted cell, since per-cut Table 1 figures are used directly rather than the abstract’s system-level mean. | WARN-MATERIAL NO3 percent-change narrative (p.3, Results 2.3): text states nitrate ‘increased by 1171% and 380% in hydroponic and aquaponics conditions, respectively’ between cuts. Recomputing from Table 3’s own Cut I/II values gives H: (3311.48-523.29)/523.29=532.8%, AQ: (3478.24-1565.55)/1565.55=122.2% — neither matches 1171%/380% under any combination tried (incl. swapped assignment). All other narrated percentage-change claims in the same paragraph (P -38%, K +19% H, Mg +108% AQ vs H, Ca +104% H, Cl +133% pooled) were independently recomputed and matched Table 3 closely, so this is not a general methodological offset — the NO3 percentages specifically appear to be an authors’ calculation error. No trials.csv cell derives from the narrative percentage (Tissue nitrate cells use raw Table 3 readings directly), so no cell is affected, but flagged for reader awareness. | WARN-BLOCK (MISATTRIBUTION) Results 2.1, p.2: ‘In the second cut, the leaf area increased by 182% and 365% and by 166% and 208%, respectively, under hydroponic and aquaponic conditions’ — text attributes all four figures to ‘the leaf area.’ Recomputation shows 182%/365% match Total DW (Table 1: H 1.16->3.26 g DW plant-1 = +181.0%; AQ 1.18->5.5 g DW plant-1 = +366.1%) and 166%/208% match Total DM (Table 1: H 2.98->7.91% = +165.4%; AQ 3.64->11.22% = +208.2%), not Total Leaf Area (already correctly described three sentences earlier in the same paragraph as -57% H / +48% AQ). Total Leaf Area and absolute Total DW have no dedicated trials.csv column, so no cell is overwritten; this row’s Plant dry matter (%) cell is sourced directly from Table 1 (independently corroborated by this same recomputation), not from the mislabelled sentence, so it is NOT set UNCLEAR. Flagged because the vault’s rule treats any confirmed misattribution as BLOCK-severity and forces quality=suspect regardless of downstream cell impact. | UNIT CONVERSION ONLY: coordinates 40 deg 48’ 57.9” N / 14 deg 21’ 01.6” E -> decimal 40.8161 / 14.3504; EC 1100 uS/cm -> 1.1 dS/m. | NOT DERIVED, left NR: Fish trial duration (only calendar dates given, ‘24 June to 10 September 2021,’ and fish were farmed continuously across both cuts in 4 weight classes rather than as a single discrete stocking-to-harvest batch, so no single stated duration figure applies; computing days between dates would be derivation); Total Feed; Fish biomass created; Fish survival rate; Fish weight gain; Fish size initial/final; FCR; SGR; feed N/P/K composition (only 42% crude protein given); Water recycle flow rate; Daily Water exchange rate; Plant height; FUE AP/HYD; WUE; Dissolved Oxygen; TAN/NH4-N, NO2-N, NO3-N (water) — none stated. | Initial Stock density (8.7+/-5.4 kg/m3) is stated as a mean stocking density across ‘4 weight classes’ farmed together in one continuous RAS (p.7), not explicitly labelled ‘initial’; recorded in the only available stocking-density column with this caveat. | Water volume in the system: paper states component volumes (4 rearing tanks x 2800 L; 800-L Superbead filter; 400-L trickling filter) but never states a summed total system volume; summing would be derivation, so left NR. | Fish Category, Water type, Water classification: paper does not categorise these in its own words — NR. | Plant Category: p.2 background text mentions basil among ‘leafy vegetables’ grown in aquaponics generally, but this is introductory context, not an explicit categorisation of the study’s own basil crop — recorded NR per the prime directive. | Average room Temperature: paper gives only an instantaneous mean of 31.1 degC/45% RH recorded at noon during the gas-exchange measurement sessions (p.8, section 4.6), not a continuous trial-long greenhouse average — recorded NR rather than substituting a point measurement for a trial mean. | NO COLUMN: Total Leaf Area (cm2 plant-1, Table 1): H,I=2795.9+/-133.19, H,II=1206.26+/-147.76, AQ,I=1476+/-134.07, AQ,II=2186.86+/-7.36. Total DW (g DW plant-1, absolute, Table 1): H,I=1.16+/-0.07, H,II=3.26+/-0.27, AQ,I=1.18+/-0.03, AQ,II=5.5+/-0.37. SLA (cm2/g, Table 1): H,I=411.3+/-1.18, H,II=315.7+/-12.4, AQ,I=280.4+/-0.007, AQ,II=320.2+/-0.72. Gas exchange/fluorescence (Table 2, cultivation-system means, n=3): Pn H=2.97+/-0.87 vs AQ=3.01+/-0.98 umol CO2 m-2 s-1 (S ns); E H=3.36+/-0.32a vs AQ=2.93+/-0.35b mol H2O m-2 s-1 (S p=0.029); ETR H=32.68+/-4.79a vs AQ=25.99+/-4.95b umol m-2 s-1 (S p=0.039); Cut(C) main effect significant (p<=0.002) for Pn, gs, E, Fv/Fm, PhiPSII, ETR, PhiNO, WUEi, LUE, RWC. None of these fit trials.csv or plant.csv’s biochemistry/mineral/microbiology/proximate categories, retained here only as summary; full Table 2 in source PDF. HYD paired plant-growth values for Cut I (Table 1, System=H): Leaf number=36.12 +/- 2.89, Total FW=38.05 +/- 1.67 g plant-1, Total DM=2.98 +/- 0.06%. Half-strength Hoagland hydroponic nutrient solution (pH 6.9, EC 1100 uS/cm), reintegrated every second week; macro/meso nutrient concentrations for both systems tabulated in Supplementary Table S1 (not accessible in the main-text PDF supplied).
modarelliHydroponicAquaponicFloating2023-T2
Fish
| Field | Value |
|---|---|
| Fish | Tilapia (Oreochromis niloticus) |
| Initial Stock density | 8.7 +/- 5.4 |
| Protein | 42 |
| % of body weight | 1.4 +/- 0.5 |
| Feed regime | Preformulated fish feed, Tilapia Grower 13-EF (Alltech Coopens, Helmond, NL), 42% protein content; daily feed target adjusted based on fish weight and stocking |
Water
| Field | Value |
|---|---|
| Aq pH | 6.9 |
| EC | 1.1 |
| Water temperature | 23 |
Plant
| Field | Value |
|---|---|
| Plant | Genovese basil (Ocimum basilicum L. cv. Sanremo) |
| Details | Cut II (second harvest), 79 days after planting (DAP); cut with scissors leaving 2 buds below |
| Days Plant after transplant | 79 |
| Plants/m2 | 20 |
| SPAD (aquaponics) | 15.97 +/- 0.64 |
| Leaf count | 62.61 +/- 4.17 |
| Plant fresh weight | 44.38 +/- 4.44 |
| Plant dry matter | 11.22 +/- 0.11 |
| Tissue nitrate AP | 3478.24 +/- 321.61 |
| Tissue nitrate HYD | 3311.48 +/- 148.39 |
System & Setup
| Field | Value |
|---|---|
| System type | Floating raft system (title; p.7-8, floating hydroponic raft / floating raft units) |
| Media Details | Floating raft units, 2 m2 each (AQ, connected by 1 loop to the RAS); disconnected/monitored separately for H; two-week-old seedlings on polystyrene sowing trays, peat cube washed off before transplant |
| Air supplement | Y (Ambient air insufflation in RAS rearing tanks at 0.05 v/v/min (p.7)) |
| pH Buffers | Y (Hydroponic nutrient solution pH adjusted with nitric acid or KOH as needed (p.8); not stated whether the aquaponic loop was separately buffered) |
| Climate control | Y (Cold greenhouse shaded with 75% black shade net (p.7); no active heating/cooling system stated) |
| Equipment | RAS: 4x2800L rearing tanks; 800-L Superbead mechanical/biological filter (Air-aqua, Staphorst, NL); 400-L trickling filter (Scubla srl, Udine, Italy); 40-W UV steriliser (Air-aqua); Thermo Scientific Expert pH and Cond Testers; LCi T photosynthesis yield analyser (ADC Bioscientific Ltd.); Opti-Sciences Plant Stress Kit fluorimeter + Fv/Fm meter; SPAD-502 chlorophyll meter (Konica Minolta); Hach DR 2000 spectrophotometer; Dionex ICS-3000 ion chromatograph with IonPac CS12A/AS11-HC columns |
| Control Parameters | Water temperature setpoint 23 degC; pH ~6.9 and EC ~1100 uS/cm (1.1 dS/m) monitored daily, stated once generally for ‘the system’ (section 4.2) and re-stated as the hydroponic nutrient solution’s own composition (section 4.3) — read as a shared target/observed value, not a conflict (setpoint vs. monitored mean; target vs. achieved are not contradictions per schema); gas exchange measured at noon, ambient CO2 434 ppm, PPFD 1251.8 umol m-2 s-1, ~31.1 degC/45% RH; greenhouse shaded with 75% black net |
| Combination | Tilapia (Oreochromis niloticus) + Genovese basil cv. Sanremo; aquaponic vs hydroponic floating raft, two consecutive cuts |
Site
| Field | Value |
|---|---|
| Region | Southern Europe |
| Country | Italy |
| Lat | 40.8161 |
| Long | 14.3504 |
Results & Statistics
| Field | Value |
|---|---|
| Measured Unit | g FW plant-1; g DW plant-1; % DM (dry matter); n plant-1 (leaf number); SPAD units; mg/kg fw (tissue nitrate, unit-corrected, see remarks) |
| Statistic Details | Two-way ANOVA (System x Cut) with interaction, SPSS v.27; Tukey’s HSD post hoc p<0.05; n=3 replicates (6 plants pooled/replicate); cluster heatmap via ClustVis, Euclidean distance, complete linkage, log(x+1)-transformed data |
| Statistically analysed | Y |
| Replicates (n) | 3 |
Experimental Remarks: TRIAL DEFINITION: this row = Aquaponic floating raft treatment (AQ), Cut II (79 DAP), n=3 replicates (6 plants pooled/replicate, section 4.9). Paired control = Hydroponic floating raft treatment (H), Cut II, recorded in HYD-equivalent columns/remarks where a column exists (Tissue nitrate HYD has a dedicated column; other HYD growth figures are noted below since the growth columns are single/AQ-focused). This paper crosses System (H vs AQ) x Cut (I vs II, i.e. 35 vs 79 DAP) as a genuine 2x2 factorial with independently replicated, separately tabulated treatment cells (Tables 1-4 report all 4 cells with own SE and Tukey letters, plus significant Cut and System x Cut interaction terms) — treated as TWO aquaponic trials rather than one; see note’s ‘Judgment call on trial splitting’ for full reasoning. Fish (tilapia, Oreochromis niloticus, 4 weight classes, one continuous RAS unit) and the hydroponic Hoagland solution are shared/continuous across both cuts, not separately stocked per cut, so all fish-side and water-setpoint cells are identical between this row and the Cut I (35 DAP) row. | WARN-MATERIAL Tissue nitrate unit: Table 3 header states NO3 in ‘g kg-1 f.w.’ while every other mineral in the same table is ‘g kg-1 d.w.’; as printed, the reported NO3 values (523-3478 ‘g kg-1’) are physically impossible (52%-348% of the tissue’s own fresh mass). Discussion (p.7): ‘considering the limit allowed for leafy vegetables, in our growing condition, we were less than that limit in both systems and cut periods’ — true only if values are mg kg-1 f.w. (typical leafy-vegetable regulatory nitrate limits are low thousands of mg/kg), not g kg-1. Recorded Tissue nitrate AP/HYD in mg/kg fw using the printed numeric values unchanged (unit-label correction, not a derivation), supported by the paper’s own regulatory-limit sentence. Affects Tissue nitrate AP and Tissue nitrate HYD in both trial rows. | WARN-MATERIAL Abstract vs Table 1 (hydroponic mean fresh weight): abstract states H and AQ ‘fresh yields of on average 41.69 and 38.38 g, respectively.’ Table 1: H Cut I=38.05+/-1.67, H Cut II=38.75+/-2.83, Table-stated H Mean=38.4 (consistent with simple average of the two cuts); AQ Mean=38.37 (matches abstract’s 38.38 within rounding) — but no combination of Table 1’s hydroponic values reaches 41.69. Table 1’s per-cut values (used directly in this row’s Plant fresh weight cell) are recorded as the defensible source since they are more granular and internally self-consistent; abstract’s H mean flagged unreliable. Does not change any extracted cell, since per-cut Table 1 figures are used directly rather than the abstract’s system-level mean. | WARN-MATERIAL NO3 percent-change narrative (p.3, Results 2.3): text states nitrate ‘increased by 1171% and 380% in hydroponic and aquaponics conditions, respectively’ between cuts. Recomputing from Table 3’s own Cut I/II values gives H: (3311.48-523.29)/523.29=532.8%, AQ: (3478.24-1565.55)/1565.55=122.2% — neither matches 1171%/380% under any combination tried (incl. swapped assignment). All other narrated percentage-change claims in the same paragraph (P -38%, K +19% H, Mg +108% AQ vs H, Ca +104% H, Cl +133% pooled) were independently recomputed and matched Table 3 closely, so this is not a general methodological offset — the NO3 percentages specifically appear to be an authors’ calculation error. No trials.csv cell derives from the narrative percentage (Tissue nitrate cells use raw Table 3 readings directly), so no cell is affected, but flagged for reader awareness. | WARN-BLOCK (MISATTRIBUTION) Results 2.1, p.2: ‘In the second cut, the leaf area increased by 182% and 365% and by 166% and 208%, respectively, under hydroponic and aquaponic conditions’ — text attributes all four figures to ‘the leaf area.’ Recomputation shows 182%/365% match Total DW (Table 1: H 1.16->3.26 g DW plant-1 = +181.0%; AQ 1.18->5.5 g DW plant-1 = +366.1%) and 166%/208% match Total DM (Table 1: H 2.98->7.91% = +165.4%; AQ 3.64->11.22% = +208.2%), not Total Leaf Area (already correctly described three sentences earlier in the same paragraph as -57% H / +48% AQ). Total Leaf Area and absolute Total DW have no dedicated trials.csv column, so no cell is overwritten; this row’s Plant dry matter (%) cell is sourced directly from Table 1 (independently corroborated by this same recomputation), not from the mislabelled sentence, so it is NOT set UNCLEAR. Flagged because the vault’s rule treats any confirmed misattribution as BLOCK-severity and forces quality=suspect regardless of downstream cell impact. | UNIT CONVERSION ONLY: coordinates 40 deg 48’ 57.9” N / 14 deg 21’ 01.6” E -> decimal 40.8161 / 14.3504; EC 1100 uS/cm -> 1.1 dS/m. | NOT DERIVED, left NR: Fish trial duration (only calendar dates given, ‘24 June to 10 September 2021,’ and fish were farmed continuously across both cuts in 4 weight classes rather than as a single discrete stocking-to-harvest batch, so no single stated duration figure applies; computing days between dates would be derivation); Total Feed; Fish biomass created; Fish survival rate; Fish weight gain; Fish size initial/final; FCR; SGR; feed N/P/K composition (only 42% crude protein given); Water recycle flow rate; Daily Water exchange rate; Plant height; FUE AP/HYD; WUE; Dissolved Oxygen; TAN/NH4-N, NO2-N, NO3-N (water) — none stated. | Initial Stock density (8.7+/-5.4 kg/m3) is stated as a mean stocking density across ‘4 weight classes’ farmed together in one continuous RAS (p.7), not explicitly labelled ‘initial’; recorded in the only available stocking-density column with this caveat. | Water volume in the system: paper states component volumes (4 rearing tanks x 2800 L; 800-L Superbead filter; 400-L trickling filter) but never states a summed total system volume; summing would be derivation, so left NR. | Fish Category, Water type, Water classification: paper does not categorise these in its own words — NR. | Plant Category: p.2 background text mentions basil among ‘leafy vegetables’ grown in aquaponics generally, but this is introductory context, not an explicit categorisation of the study’s own basil crop — recorded NR per the prime directive. | Average room Temperature: paper gives only an instantaneous mean of 31.1 degC/45% RH recorded at noon during the gas-exchange measurement sessions (p.8, section 4.6), not a continuous trial-long greenhouse average — recorded NR rather than substituting a point measurement for a trial mean. | NO COLUMN: Total Leaf Area (cm2 plant-1, Table 1): H,I=2795.9+/-133.19, H,II=1206.26+/-147.76, AQ,I=1476+/-134.07, AQ,II=2186.86+/-7.36. Total DW (g DW plant-1, absolute, Table 1): H,I=1.16+/-0.07, H,II=3.26+/-0.27, AQ,I=1.18+/-0.03, AQ,II=5.5+/-0.37. SLA (cm2/g, Table 1): H,I=411.3+/-1.18, H,II=315.7+/-12.4, AQ,I=280.4+/-0.007, AQ,II=320.2+/-0.72. Gas exchange/fluorescence (Table 2, cultivation-system means, n=3): Pn H=2.97+/-0.87 vs AQ=3.01+/-0.98 umol CO2 m-2 s-1 (S ns); E H=3.36+/-0.32a vs AQ=2.93+/-0.35b mol H2O m-2 s-1 (S p=0.029); ETR H=32.68+/-4.79a vs AQ=25.99+/-4.95b umol m-2 s-1 (S p=0.039); Cut(C) main effect significant (p<=0.002) for Pn, gs, E, Fv/Fm, PhiPSII, ETR, PhiNO, WUEi, LUE, RWC. None of these fit trials.csv or plant.csv’s biochemistry/mineral/microbiology/proximate categories, retained here only as summary; full Table 2 in source PDF. HYD paired plant-growth values for Cut II (Table 1, System=H): Leaf number=46.67 +/- 3.49, Total FW=38.75 +/- 2.83 g plant-1, Total DM=7.91 +/- 0.13%. Half-strength Hoagland hydroponic nutrient solution (pH 6.9, EC 1100 uS/cm), reintegrated every second week; macro/meso nutrient concentrations for both systems tabulated in Supplementary Table S1 (not accessible in the main-text PDF supplied).
Plant Measurements
| Trial | System | Category | Analyte | Value | Unit | Sig. | Location |
|---|---|---|---|---|---|---|---|
| modarelliHydroponicAquaponicFloating2023-T1 | HYD | mineral | Phosphorus (P) | 4.33 ± 0.42 | g/kg d.w. | a | Table 3 |
| modarelliHydroponicAquaponicFloating2023-T2 | HYD | mineral | Phosphorus (P) | 3.47 ± 0.37 | g/kg d.w. | ab | Table 3 |
| modarelliHydroponicAquaponicFloating2023-T1 | AQ | mineral | Phosphorus (P) | 3.94 ± 0.22 | g/kg d.w. | ab | Table 3 |
| modarelliHydroponicAquaponicFloating2023-T2 | AQ | mineral | Phosphorus (P) | 2.54 ± 0.29 | g/kg d.w. | b | Table 3 |
| modarelliHydroponicAquaponicFloating2023-T1 | HYD | mineral | Potassium (K) | 72.06 ± 3.22 | g/kg d.w. | ab | Table 3 |
| modarelliHydroponicAquaponicFloating2023-T2 | HYD | mineral | Potassium (K) | 85.79 ± 4.7 | g/kg d.w. | a | Table 3 |
| modarelliHydroponicAquaponicFloating2023-T1 | AQ | mineral | Potassium (K) | 57.57 ± 3.65 | g/kg d.w. | bc | Table 3 |
| modarelliHydroponicAquaponicFloating2023-T2 | AQ | mineral | Potassium (K) | 54.24 ± 1.66 | g/kg d.w. | c | Table 3 |
| modarelliHydroponicAquaponicFloating2023-T1 | HYD | mineral | Sulphate (S) | 0.77 ± 0.05 | g/kg d.w. | b | Table 3 |
| modarelliHydroponicAquaponicFloating2023-T2 | HYD | mineral | Sulphate (S) | 0.96 ± 0.1 | g/kg d.w. | b | Table 3 |
| modarelliHydroponicAquaponicFloating2023-T1 | AQ | mineral | Sulphate (S) | 1.32 ± 0.04 | g/kg d.w. | a | Table 3 |
| modarelliHydroponicAquaponicFloating2023-T2 | AQ | mineral | Sulphate (S) | 0.73 ± 0.05 | g/kg d.w. | b | Table 3 |
| modarelliHydroponicAquaponicFloating2023-T1 | HYD | mineral | Calcium (Ca) | 7.64 ± 0.41 | g/kg d.w. | b | Table 3 |
| modarelliHydroponicAquaponicFloating2023-T2 | HYD | mineral | Calcium (Ca) | 15.59 ± 1.27 | g/kg d.w. | a | Table 3 |
| modarelliHydroponicAquaponicFloating2023-T1 | AQ | mineral | Calcium (Ca) | 16.5 ± 0.71 | g/kg d.w. | a | Table 3 |
| modarelliHydroponicAquaponicFloating2023-T2 | AQ | mineral | Calcium (Ca) | 15.13 ± 0.15 | g/kg d.w. | a | Table 3 |
| modarelliHydroponicAquaponicFloating2023-T1 | HYD | mineral | Magnesium (Mg) | 2.26 ± 0.1 | g/kg d.w. | b | Table 3 |
| modarelliHydroponicAquaponicFloating2023-T2 | HYD | mineral | Magnesium (Mg) | 3.21 ± 0.16 | g/kg d.w. | b | Table 3 |
| modarelliHydroponicAquaponicFloating2023-T1 | AQ | mineral | Magnesium (Mg) | 6.09 ± 0.32 | g/kg d.w. | a | Table 3 |
| modarelliHydroponicAquaponicFloating2023-T2 | AQ | mineral | Magnesium (Mg) | 5.3 ± 0.22 | g/kg d.w. | a | Table 3 |
| modarelliHydroponicAquaponicFloating2023-T1 | HYD | mineral | Ammonium (NH4) | 0.39 ± 0.01 | g/kg d.w. | c | Table 3 |
| modarelliHydroponicAquaponicFloating2023-T2 | HYD | mineral | Ammonium (NH4) | 0.94 ± 0.02 | g/kg d.w. | a | Table 3 |
| modarelliHydroponicAquaponicFloating2023-T1 | AQ | mineral | Ammonium (NH4) | 0.34 ± 0.01 | g/kg d.w. | c | Table 3 |
| modarelliHydroponicAquaponicFloating2023-T2 | AQ | mineral | Ammonium (NH4) | 0.57 ± 0.04 | g/kg d.w. | b | Table 3 |
| modarelliHydroponicAquaponicFloating2023-T1 | HYD | mineral | Sodium (Na) | 0.52 ± 0.11 | g/kg d.w. | No Tukey letters reported for Na. ANOVA: System p=0.132 ns, Cut p=0.079 ns, System x Cut p=0.592 ns | Table 3 |
| modarelliHydroponicAquaponicFloating2023-T2 | HYD | mineral | Sodium (Na) | 0.8 ± 0.14 | g/kg d.w. | No Tukey letters reported for Na. ANOVA: System p=0.132 ns, Cut p=0.079 ns, System x Cut p=0.592 ns | Table 3 |
| modarelliHydroponicAquaponicFloating2023-T1 | AQ | mineral | Sodium (Na) | 0.4 ± 0.09 | g/kg d.w. | No Tukey letters reported for Na. ANOVA: System p=0.132 ns, Cut p=0.079 ns, System x Cut p=0.592 ns | Table 3 |
| modarelliHydroponicAquaponicFloating2023-T2 | AQ | mineral | Sodium (Na) | 0.56 ± 0.08 | g/kg d.w. | No Tukey letters reported for Na. ANOVA: System p=0.132 ns, Cut p=0.079 ns, System x Cut p=0.592 ns | Table 3 |
| modarelliHydroponicAquaponicFloating2023-T1 | HYD | mineral | Chloride (Cl) | 6.07 ± 1.07 | g/kg d.w. | No Tukey letters reported for Cl. ANOVA: System p=0.013 ns, Cut p=0.00**, System x Cut p=0.2 ns | Table 3 |
| modarelliHydroponicAquaponicFloating2023-T2 | HYD | mineral | Chloride (Cl) | 14.36 ± 1.75 | g/kg d.w. | No Tukey letters reported for Cl. ANOVA: System p=0.013 ns, Cut p=0.00**, System x Cut p=0.2 ns | Table 3 |
| modarelliHydroponicAquaponicFloating2023-T1 | AQ | mineral | Chloride (Cl) | 4.2 ± 0.38 | g/kg d.w. | No Tukey letters reported for Cl. ANOVA: System p=0.013 ns, Cut p=0.00**, System x Cut p=0.2 ns | Table 3 |
| modarelliHydroponicAquaponicFloating2023-T2 | AQ | mineral | Chloride (Cl) | 9.57 ± 0.19 | g/kg d.w. | No Tukey letters reported for Cl. ANOVA: System p=0.013 ns, Cut p=0.00**, System x Cut p=0.2 ns | Table 3 |
| modarelliHydroponicAquaponicFloating2023-T1 | HYD | biochemistry | Chlorophyll a/b ratio | 1.35 ± 0.01 | ratio | b | Table 4 |
| modarelliHydroponicAquaponicFloating2023-T2 | HYD | biochemistry | Chlorophyll a/b ratio | 1.82 ± 0.04 | ratio | a | Table 4 |
| modarelliHydroponicAquaponicFloating2023-T1 | AQ | biochemistry | Chlorophyll a/b ratio | 1.78 ± 0.06 | ratio | a | Table 4 |
| modarelliHydroponicAquaponicFloating2023-T2 | AQ | biochemistry | Chlorophyll a/b ratio | 1.74 ± 0.04 | ratio | a | Table 4 |
| modarelliHydroponicAquaponicFloating2023-T1 | HYD | biochemistry | Total chlorophyll (a+b) | 1.76 ± 0.04 | mg/g FW | a | Table 4 |
| modarelliHydroponicAquaponicFloating2023-T2 | HYD | biochemistry | Total chlorophyll (a+b) | 0.95 ± 0.03 | mg/g FW | c | Table 4 |
| modarelliHydroponicAquaponicFloating2023-T1 | AQ | biochemistry | Total chlorophyll (a+b) | 1.25 ± 0.05 | mg/g FW | b | Table 4 |
| modarelliHydroponicAquaponicFloating2023-T2 | AQ | biochemistry | Total chlorophyll (a+b) | 1.12 ± 0.1 | mg/g FW | bc | Table 4 |
| modarelliHydroponicAquaponicFloating2023-T1 | HYD | biochemistry | Total carotenoids | 0.22 ± 0.0 | mg/g FW | c | Table 4 |
| modarelliHydroponicAquaponicFloating2023-T2 | HYD | biochemistry | Total carotenoids | 0.25 ± 0.01 | mg/g FW | b | Table 4 |
| modarelliHydroponicAquaponicFloating2023-T1 | AQ | biochemistry | Total carotenoids | 0.29 ± 0.0 | mg/g FW | a | Table 4 |
| modarelliHydroponicAquaponicFloating2023-T2 | AQ | biochemistry | Total carotenoids | 0.27 ± 0.01 | mg/g FW | ab | Table 4 |
| modarelliHydroponicAquaponicFloating2023-T1 | HYD | biochemistry | SPAD Index | 23.13 ± 0.75 | SPAD units | a | Table 4 |
| modarelliHydroponicAquaponicFloating2023-T2 | HYD | biochemistry | SPAD Index | 13.11 ± 0.88 | SPAD units | c | Table 4 |