Comparative Analysis of Horizontal and Vertical Decoupled Aquaponic Systems for Basil Production and Effect of Light Supplementation by LED

Metadata

• Cite key: fernandezcabanasComparativeAnalysisHorizontal2020
• Item type: Journal Article
• Authors: V.M. Fernández-Cabanás, L. Pérez-Urrestarazu, A. Juárez, N.T. Kaufman, J.A. Gross
• Affiliation: Urban Greening and Biosystems Engineering Research Group, Departamento de Ciencias Agroforestales / Area of Agro-Forestry Engineering, Universidad de Sevilla, ETSIA, Seville, Spain (authors 1–2); Department of Animal Sciences, University of California, Davis, CA, USA (authors 3–5)
• Journal: Agronomy 10 (2020) 1414 (14 pp.)
• Date: 09/2020
• Date added: [not reported] — Zotero export contains multiple duplicate entries for this item with different “date added” timestamps (2024-03-05 and 2026-07-13); no single value can be attributed with confidence
• DOI: 10.3390/agronomy10091414
• Funding: “This research received no external funding.” Acknowledgments separately state the University of Seville (Plan Propio de Investigación) financed V.M. Fernández-Cabanás’s research stay at UC Davis — recorded as-is, not treated as a contradiction (internal institutional travel funding vs. an external competitive research grant are different things).
• URL: https://doi.org/10.3390/agronomy10091414
• PDF: Fernández-Cabanás et al. - 2020 - Comparative Analysis of Horizontal and Vertical Decoupled Aquaponic Systems for Basil Production and.pdf

Opinion

A well-designed, adequately replicated (24 tray locations, 20 monitored plants/tray, randomized LED assignment, blind data collection, GLM + Tukey/Games-Howell) comparison that isolates two confounded variables cleanly: tray orientation (horizontal vs. vertical) and lighting supplementation. The economic costing (Table 8) is a nice practical addition. The main limitation for this vault is structural rather than a fault of the paper: the “vertical” system’s per-plant morphometrics are only ever compared to horizontal at the single top shelf (level 3), never as a whole-rack aggregate, while the area-yield comparison (Table 3, the headline 160% figure) is a whole-rack aggregate — the two “vertical” numbers used elsewhere in the literature are not measuring the same population of plants. Worth citing for anyone modelling vertical-vs-horizontal aquaponic productivity, but flag which vertical value is being reused.

Abstract

Aquaponic facilities allow the joint production of fish and plants in a reduced space. This hydroponic technique, combined with vertical farming, is not well studied. The main objective of this work was to compare basil production between horizontal and vertical decoupled aquaponic systems and assess the utilization of supplemental lighting in a greenhouse environment. Six independent vertical racks were arranged with hydroponic trays at three heights. Three racks were affixed with LED lighting on the lowest levels and three with supplemental lighting on the intermediate level. Immediately adjacent to the vertical systems, two independent horizontal systems containing three trays were constructed to compare basil production. After 35 days of growth post transplanting, the total production per tray and weight, height, number of leaves, and nodes of 20 selected plants per tray were determined. Records of the intensity of illumination photosynthetic photon flux density (PPFD)) were recorded at dawn (8:00), noon (12:00), and dusk (18:00) at randomly predetermined positions associated with the 20 selected plants per tray. The total basil production in the experiment was 58.79 kg, with an average production per unit area of 2.43 and 0.94 kg m−2 for vertical and horizontal systems, respectively. Productivity per unit area in the vertical systems was 160% greater than in horizontal systems. A significant effect of lighting, the height of the tray, and plant position inside the tray was found on plant growth parameters. Optimization of light source distribution and tray orientation can enhance the productive performance in vertical aquaponic systems. Electricity cost associated with supplemental lighting per kg of production increment was 21.84 and 12.25 $ kg−1 for the bottom and intermediate levels of the vertical system, respectively, the latter being economically the most profitable.

Summary

The authors built a decoupled aquaponic system at UC Davis in which effluent from three white/green sturgeon RAS units was piped (without return flow) into six vertical, three-tier flood-and-drain racks and two horizontal flood-and-drain systems, both growing basil (cv. Nufar) for 35 days after transplant. Within the vertical racks, the two lower shelves (levels 1 and 2) were randomly assigned either natural sunlight only or supplemental LED lighting, while the top shelf (level 3) always received natural light only; horizontal trays also received natural light only. Basil production per unit floor area was 160% higher in vertical racks than in horizontal systems (2.427 vs. 0.939 kg/m2, p=0.0037), and LED supplementation significantly increased plant weight, height, leaf count and node count at the shaded lower levels, with a stronger effect at level 1 than level 2. A simple cost analysis found that lighting the intermediate shelf (level 2) alone was the most cost-effective option ($12.25perkgofextraproduction),whilelightingthebottomshelfalonewasleastcost-effective($21.84/kg). No fish growth, survival, or water-quality trial means beyond a target nitrate range were reported — the fish/RAS side of the system functioned purely as a nutrient source for the plant-side comparison.

---

Experiment data

• Location: 280 m2 polycarbonate greenhouse, UC Davis Campus, Davis, CA, USA (38.5386, -121.7817)
• Design: Randomized, replicated comparison of decoupled aquaponic basil production in horizontal (2 racks x 3 trays) vs. vertical (6 racks x 3 shelf levels) flood-and-drain systems; within vertical racks, levels 1 and 2 randomized to natural sunlight vs. LED supplementation (one of each per rack per level), level 3 always natural sunlight only; blocked for greenhouse location, shelf height, and lighting
• Replicates / n: 24 tray locations total (6 horizontal + 18 vertical); 20 individually monitored plants per tray; 4800 seedlings transplanted overall
• Duration: 19 August – 8 October 2019 overall; 14 days germination-to-transplant; 35 days post-transplant to harvest
• Organisms: Basil (Ocimum basilicum) cv. Nufar (Fusarium-resistant Italian large-leaf type); nutrient source: White sturgeon (Acipenser transmontanus) and Green sturgeon (Acipenser medirostris) RAS effluent (decoupled — no fish growth/performance data reported)
• Statistics: Generalized linear model (GLM); Shapiro-Wilk normality test; Kruskal-Wallis (non-normal data); Tukey HSD and Games-Howell post-hoc; STATGRAPHICS Centurion XVIII v18.1.12; significance at p<0.05
• Basil production per unit area: 2.427 ± 0.155 kg/m2 (vertical, aggregate) vs. 0.939 ± 0.246 kg/m2 (horizontal), p=0.0037 — 160% higher in vertical
• Basil production per unit area: by level x lighting (Table 1): Level 1 Sunlight 0.451, Level 1 LED 0.750, Level 2 Sunlight 0.685, Level 2 LED 1.191, Level 3 (natural only) 0.981 kg/m2
• Plant fresh weight: Horizontal 12.98 ± 0.30 g/plant vs. Vertical (level 3 only) 13.97 ± 0.33 g/plant, p=0.0293 (Table 6); Level 1/2 sub-means (Table 4): 7.13 / 8.53 / 11.52 / 17.04 g for Level1-Sun / Level2-Sun / Level1-LED / Level2-LED
• Feed Conversion Rate (FCR): [not reported] — fish growth was not measured in this paper (decoupled system, fish are a nutrient source only)

---

Horizontal vs. vertical productivity

This paper: Vertical racks produced 2.427 ± 0.155 kg basil m⁻² vs. 0.939 ± 0.246 kg m⁻² in horizontal systems (p=0.0037, Table 3), a 160% increase, matching the abstract’s rounded 2.43/0.94 figures. The authors note this is lower than the theoretical 200% (3x) increment expected from stacking three trays, attributable to reduced yield on shaded lower shelves.

Compared with:

• todo Neocleous et al. 2010 — horizontal vs. vertical hydroponic lettuce/strawberry; vertical lettuce had more marketable heads/area but lower average weight and marketable %; vertical strawberry had higher yield with no quality difference. (p. 10, secondary)
• todo Ramírez-Arias et al. 2018 — vertical (bucket column) strawberry productivity 3.23-fold and 1.8-fold vs. two horizontal setups, per-plant production not significantly different. (p. 10, secondary)
• todo Touliatos, Dodd & McAinsh 2016 — indoor vertical lettuce 13.8x more productive per floor area than horizontal (artificial light only); bottom vertical layer produced 43% less shoot fresh weight than top layer; PPFD dropped from 224 to 122 µmol m⁻²s⁻¹ top-to-bottom. (p. 11, secondary)
• todo Rakocy et al. 2004 (UVI system) — outdoor commercial-scale aquaponic basil, 1.8 kg m⁻² (vs. 0.68 kg m⁻² in soil); higher than this paper’s horizontal yield, attributed to climate and a 3+1 cut harvest regime vs. this paper’s single harvest. (p. 12, secondary)
• todo Filep et al. 2016 — horizontal aquaponic basil + carp, 180-day, 3-harvest yields of 939/1130/1213 g m⁻² with 600 W supplemental lighting, consistent with UVI figures. (p. 12, secondary)
• todo Saha, Monroe & Day 2016 — basil, crayfish-based floating-raft aquaponics vs. hydroponics: height 89.9±4.5 vs. 78.7±3.9 cm (+14%); fresh weight 150.2±18 vs. 96.6±10.4 g/plant (+56%); dry weight 15.9±2 vs. 9.6±1 g (+65%), all favoring aquaponics. (p. 12, secondary)
• todo Roosta & Afsharipoor 2012 — carp-based aquaponic/hydroponic basil ratios; best result (~240 g/plant) with pure hydroponics, ~50 g/plant with pure aquaponic solution — contradicts the aquaponics-favoring direction of Saha et al. (p. 12, secondary)

LED lighting effect

This paper: At level 1, LED supplementation raised production from 0.451 to 0.750 kg m⁻² (p=0.002); at level 2, from 0.685 to 1.191 kg m⁻² (n.s., p=0.1106, though the pooled level x lighting analysis in Table 2 found both main effects significant, p=0.0070 lighting / p=0.0229 height). Per-plant weight, height and leaf count were all significantly higher with LED at both levels (Table 4, p=0.0000 for level x lighting interaction on all three). Economically, lighting level 2 alone was most cost-effective ($12.25perkgproductionincrement)vs.level1alone($21.84/kg) vs. both levels ($15.70/kg)(Table8),givenelectricityat$0.1524/kWh and basil at $13.58/kg.

Compared with:

• todo Liaros, Botsis & Xydis 2016 — estimated basil DLI requirement ~15 mol m⁻²d⁻¹ (≈300 µmol m⁻²s⁻¹ at 14 h photoperiod); found sweet basil commercially viable indoors in Greece despite high light needs, due to higher crop value. (p. 11, secondary)

Extraction notes

Not a review — the paper collects and statistically analyses its own greenhouse data (GLM, Tukey/Games-Howell, n=24 tray locations, 20 monitored plants/tray); classified experiment.

Decoupled system, fish side out of scope: the three RAS units (white/green sturgeon, <1 yr, 2000 L tanks, well water, bead filters) are described only as a nutrient source; no fish growth, survival, FCR, SGR, or stocking density is reported anywhere in the paper. All fish-performance columns in trials.csv are NR, not NA, because fish were present and reared but their performance data simply were not reported in this paper (as opposed to a plant-only study where fish are absent by design).

WARN-CHECK — Plant height / leaf count / fresh weight aggregation level. Table 6 is the only place the paper directly compares Horizontal against a “Vertical” value for per-plant weight/height/leaves/nodes, and there “Vertical” means level 3 (top shelf) only (13.97±0.33 g weight; 33.55±0.49 cm height; 24.06±0.55 leaves, n.s. p=0.0745; 6.18±0.05 nodes). Table 4 separately reports level 1 and level 2 means split further by lighting (Sunlight vs. Sunlight+LED): weight 7.13/8.53/11.52/17.04 g, height 20.04/26.13/20.48/33.23 cm, leaves 9.30/11.28/16.82/24.65 for Level1-Sun/Level2-Sun/Level1-LED/Level2-LED respectively; Table 5 gives node-count main effects only (Level1=4.49, Level2=5.22; Sunlight=4.55, LED=5.16). No single reported mean represents “the whole vertical rack” for these per-plant metrics — only the area-yield metric (kg/m2, Table 3) is aggregated across the entire vertical system. Recorded Table 6’s level-3-only figures as the trials.csv basis for the Vertical trial because it is the only genuine Horizontal-vs-Vertical comparison point for these variables; all Level 1/2 sub-condition means are listed in each trial’s Experimental Remarks as candidates. Added to REVIEW.md.

[not reported] fields, grouped:

• Fish performance: Initial Stock density, FCR, SGR, Fish size initial/final, Fish weight gain, Fish survival rate, Fish trial duration, Total Feed (kg), Fish biomass created — none measured/reported (decoupled nutrient-source RAS, not a fish trial)
• Feed composition: N (%), K (%) of feed — only crude protein (46%) and phosphorus (1.1%) given
• Water quality trial means: Aq pH (only a 6.0–6.5 target/setpoint given, recorded in pHOptimal instead per the setpoint-vs-measured convention), EC, Water temperature, TAN/NH4-N, NO2-N — measured with a Bluelab combo meter per Methods but no summary values reported in Results
• Plant: SPAD, Plant dry matter — not measured
• Daily Water exchange rate — only a twice-weekly absolute replenishment volume (110 L/rack) is given for vertical racks, not restated for horizontal, and not expressible as a daily % without derivation
• Average room Temperature — not reported (a cooling system between levels 1–2 is mentioned as affecting north-end air temperature, but no value given)
• Fish Category, Water classification — species/source given but no categorical label used by the paper (consistent with vault precedent, e.g. Alcarraz 2018)

No [unclear] fields — nothing in this paper was ambiguous enough to warrant that tag; gaps were clean absences ([not reported]).

NO COLUMN items (no home in trials.csv/plant.csv, kept here and in each trial’s Experimental Remarks):

• Economic analysis (Table 8): lighting cost $6.53/m2persupplementedlevel(14h/dayx35daysx65Wx4bars/shelf,$0.1524/kWh); cost per kg production increment $21.84(Level1-LED)/$12.25 (Level2-LED) / $15.70(bothlevels);basilpriceassumed$13.58/kg
• PPFD regression: ln(Weight) = -0.422383 + 0.571854 x ln(PPFD), R2=0.52, p=0.0000
• PPFD trial means by tray/light/position/time-of-day (Table 7) — full time-series/positional matrix, not reducible to one trial mean; range across all conditions/times ≈ 17–468 µmol s⁻¹m⁻² (range only, no single trial mean reported)
• Total basil production across the whole experiment: 58.79 kg (abstract)

No water panel excluded from plant.csv — this paper’s only water-quality figure is the nitrate target/achieved value (57.5±18.4 mg/L, “in all systems”), which was routed to trials.csv NO3-N per schema; nothing water-related was discarded.

MINOR: Pot dimensions given as “0.102 × 0.102 × 0.089 m3” (p. 4–5) — almost certainly intends three linear dimensions in meters with m3 as the resulting product’s unit label, an unusual but not internally inconsistent notation (67.19 plants/m2 recomputes correctly from 200 plants / (1.22×2.44 m2) tray area, confirming the paper’s own “around 67 plants m−2” figure — recomputed for verification only, not entered as a cell). Does not affect any extracted cell.

Severity tally: 0 BLOCK, 0 MATERIAL, 1 CHECK (plant morphometric aggregation level), 1 MINOR (pot-dimension unit notation) → quality: ok (0 BLOCK, ≤2 MATERIAL).

Citations to chase

• todo Neocleous, D.; Kaittanis, C.; Seraphides, N.; Polycarpou, P. (2010) — horizontal vs. vertical soilless lettuce/strawberry, Cyprus
• todo Ramírez-Arias, J.A. et al. (2018) — horizontal vs. vertical hydroponic strawberry at high density
• todo Touliatos, D.; Dodd, I.C.; McAinsh, M. (2016) — vertical farming lettuce yield per unit area
• todo Rakocy, J.E.; Shultz, R.C.; Bailey, D.S.; Thoman, E.S. (2004) — UVI aquaponic tilapia/basil, batch vs. staggered cropping
• todo Filep, R.M. et al. (2016) — pilot aquaponic carp/basil system
• todo Saha, S.; Monroe, A.; Day, M.R. (2016) — basil growth/quality under soilless (hydroponic vs. crayfish aquaponic) systems
• todo Roosta, H.R.; Afsharipoor, S. (2012) — cultivation media effects on strawberry under hydroponic/aquaponic systems
• todo Liaros, S.; Botsis, K.; Xydis, G. (2016) — technoeconomic evaluation of urban basil plant factories

---

Source: Fernández-Cabanás et al. - 2020 - Comparative Analysis of Horizontal and Vertical Decoupled Aquaponic Systems for Basil Production and.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

fernandezcabanasComparativeAnalysisHorizontal2020-T1

Fish

Field	Value
Fish	White sturgeon (Acipenser transmontanus) and green sturgeon (Acipenser medirostris), <1 yr
Protein	46
P	1.1
Feed routine	Fed once daily (200 g/tank/day); time of day not stated
Feed regime	Sinking complete feed for trout/steelhead (Skretting, Tooele, UT, USA), 46% crude protein, 12% crude fat, 3% crude fiber, 1.1% phosphorus

Water

Field	Value
Water recycle	69.4 L/min
Water volume in the system	473 L (sump, plant loop only; excludes 2000 L fish tanks, decoupled)
Water type	Fish effluent (decoupled, sourced from sturgeon RAS reared on well water)
pHOptimal	6.0-6.5 (target range, maintained with phosphoric+citric acid pH Down dosing; no measured trial mean reported)
NO3-N	57.5 +/- 18.4 mg/L (NO3-N, LaMotte Nitrate Nitrogen tablet kit, target ~60 ppm, stated as applying ‘in all systems’)

Plant

Field	Value
Plant	Basil (Ocimum basilicum), cv. Nufar (Fusarium-resistant Italian large-leaf type)
Details	Pelleted seeds sown into propagation trays with RO water; transplanted 14 days after germination into Q plugs media pots; harvested 35 days after transplant
Plant Category	Specialty herb (p.4, “popular specialty herb”)
Days Plant after transplant	35
Plants/m2	67 (paper states “around 67 plants m-2”; 200 plants/tray on a 1.22x2.44 m tray)
Plant height	31.58 +/- 0.45 cm (Table 6, Horizontal)
Leaf count	22.73 +/- 0.50 (Table 6, Horizontal)
Plant fresh weight	12.98 +/- 0.30 g/plant (Table 6, Horizontal)

System & Setup

Field	Value
System type	Horizontal flood-and-drain tray system (2 independent horizontal racks, 3 trays each), decoupled aquaponics, natural sunlight only
Media Details	Q plugs rooting media in pots labelled 0.102 x 0.102 x 0.089 m3 (p.4-5, likely 3 linear dimensions in m with m3 as product unit); 200 pots/tray in 20x10 grid; tray 1.22x2.44x0.15 m (Botanicare, Chandler, AZ)
Biological system already in use	Y (Nutrients sourced from 3 pre-existing recirculating aquaculture systems (RAS) at UC Davis Center for Aquatic Biology and Aquaculture (CABA) rearing white and green sturgeon (<1 yr) in 2000 L tanks on well water with bead-filter mechanical/biological filtration; decoupled — effluent transferred to plant loop only, no return flow to fish)
Iron supplemented	Y (0.2 g chelated Fe-DTPA (13% Fe) dissolved weekly per sump, raising Fe concentration by 0.055 mg/L; total Fe applied per system 0.182 g vs. estimated plant uptake need 0.174 g)
pH Buffers	Y (Phosphoric + citric acid blend (pH Down, General Hydroponics, Sebastopol, CA) used to lower pH to target 6.0-6.5 range)
Climate control	Y (280 m2 polycarbonate greenhouse at UC Davis; a cooling system located between levels 1 and 2 of vertical racks affected north-end air temperature; no setpoints given)
Artificial Lighting	N (N — horizontal trays received natural sunlight only, no LED supplementation (p.4))
Nutrient supplemented	Y (1% KNO3 foliar feed applied twice weekly by manual sprayer (potassium supplementation without affecting solution pH); plus Fe-DTPA (see Iron column))
Equipment	Botanicare 1.22x2.44x0.15 m flood-and-drain trays; 473 L sump; ActiveAqua 1100 GPH (4164 L/h) submersible pump; Bluelab combo meter (pH/EC/temp); LaMotte Nitrate Nitrogen tablet test kit; MQ-500 Apogee quantum meter (PPFD); Infinity 2.0 LED bars (vertical racks levels 1-2 only)
Control Parameters	Target NO3 ~60 ppm (achieved 57.5+/-18.4 mg/L); target pH 6.0-6.5; flood irrigation 2x/day (07:00, 14:00) for 15 min to ~0.02 m depth; supplemental lighting 14 h/day 06:30-20:30 where applied (vertical racks levels 1-2 only)
Combination	Basil (Ocimum basilicum cv. Nufar) grown on decoupled sturgeon RAS effluent; horizontal vs. vertical (3-tier) flood-and-drain racks; vertical racks additionally randomized for LED supplementation at levels 1-2 vs. natural sunlight only

Site

Field	Value
Region	North America
Country	USA
Lat	38.5386
Long	-121.7817

Results & Statistics

Field	Value
Measured Unit	kg/m2 (production per unit area, Table 3); g/plant (fresh weight), cm (height), count (leaves, nodes) (Table 6)
Statistic Details	Generalized linear model (GLM); Shapiro-Wilk normality test; Kruskal-Wallis for non-normal data; Tukey HSD and Games-Howell post-hoc; p<0.05; STATGRAPHICS Centurion XVIII v18.1.12
Statistically analysed	Y
Replicates (n)	6 (2 racks x 3 trays); 20 monitored plants/tray
AP	939 g fw/m2 (UNIT CONVERSION from 0.939 kg/m2, Table 3)

Experimental Remarks: TRIAL DEFINITION: T1 = Horizontal decoupled aquaponic system (2 independent horizontal racks x 3 trays each = 6 tray replicates), natural sunlight only, no LED. Paired comparison = T2 (Vertical system, aggregate across levels 1-3). No hydroponic-only control exists in this paper (all treatments receive sturgeon RAS effluent); HYD columns are NA throughout. | WARN-CHECK Plant height/leaf count/fresh weight aggregation level: Table 6 gives a single Horizontal-vs-Vertical comparison, but ‘Vertical’ there means level 3 (top shelf) ONLY (13.97+/-0.33 g weight; 33.55+/-0.49 cm height; 24.06+/-0.55 leaves, n.s. p=0.0745; 6.18+/-0.05 nodes), not a whole-rack aggregate. Table 4 separately reports level 1 and level 2 means split further by lighting (Sunlight vs. Sunlight+LED): weight 7.13/8.53/11.52/17.04 g; height 20.04/26.13/20.48/33.23 cm; leaves 9.30/11.28/16.82/24.65 for Level1-Sun/Level2-Sun/Level1-LED/Level2-LED respectively (Table 4); Table 5 gives node-count main effects only (Level1=4.49, Level2=5.22; Sunlight=4.55, LED=5.16), no per-cell interaction means. No single reported mean represents ‘the whole vertical system’ for per-plant weight/height/leaf/node (only production kg/m2, Table 3, is aggregated across the entire vertical system). Recorded Table 6’s level-3-only value as the basis because it is the only place the paper directly compares Horizontal against a Vertical value for these per-plant metrics; all Level 1/2 sub-condition means listed here as candidates. Added to REVIEW.md. | NO COLUMN: Basil production per unit area is the paper’s headline metric but trials.csv has no dedicated yield-per-area column: 0.939 +/- 0.246 kg/m2 (horizontal, Table 3) vs. 2.427 +/- 0.155 kg/m2 (vertical, Table 3), significant p=0.0037 (Table 3); abstract reports rounded 0.94 and 2.43 kg/m2 respectively (rounding, not a conflict). | NOT DERIVED, left NR: fish stocking density, FCR, SGR, fish size initial/final, fish weight gain, fish survival rate, fish trial duration, Total Feed (kg) (only a 200 g/day/tank feeding rate given, no total or duration to sum over), SPAD, plant dry matter (never measured), Water temperature/EC/DO (measured with a Bluelab meter per Methods but no summary values reported anywhere in Results), Daily Water exchange rate (only twice-weekly absolute replenishment volumes given, not a % rate), Nitrogen % and Potassium % of feed (only crude protein 46% and phosphorus 1.1% given). | Aq pH left NR (no measured trial mean reported); the stated 6.0-6.5 is a target/setpoint maintained by pH-Down dosing, recorded in pHOptimal instead per the setpoint-vs-measured convention. | UNIT CONVERSION ONLY: water recycle 4164 L/h -> 69.4 L/min; coordinates 38 deg 32’18.8” N / 121 deg 46’54.2” W -> decimal 38.5386 / -121.7817. Production 0.939 kg/m2 -> 939 g/m2 (AP column). | Funding statement says “This research received no external funding,” but Acknowledgments separately credit University of Seville Plan Propio de Investigacion for financing V.M. Fernandez-Cabanas’s research stay at UC Davis — recorded both, not treated as a contradiction (internal institutional travel funding vs. external competitive grant funding are different things). | MINOR: pot dimensions given as “0.102 x 0.102 x 0.089 m3” (p.4-5) — likely intends meters per dimension with m3 as the resulting product unit label, an unusual but not numerically inconsistent notation; does not affect any extracted cell. Plants/m2 (67, ‘around 67 plants m-2’) and tray dimensions (1.22x2.44 m, 200 plants/tray) identical for horizontal and vertical trays (same equipment, p.4). | Fish Category NR — paper names species (white/green sturgeon, <1 yr) but does not assign a category label (cf. vault precedent, e.g. Alcarraz 2018). | Water exchange for horizontal racks specifically: the paper states vertical racks each received ‘an additional 110 L of sturgeon effluent twice a week’ (p.5) but does NOT explicitly restate this figure for the horizontal systems (only that horizontal systems used ‘the same equipment … sumps, pumps, and irrigation periods’, p.4) — recorded as NR for horizontal-specific replenishment volume rather than assumed equal. | NO COLUMN: economic analysis (Table 8) — lighting cost $6.53/m2persupplementedlevel,costperkgproductionincrement$21.84 (Level1-LED) / $12.25(Level2-LED)/$15.70 (both levels), basil price assumed $13.58/kg,electricity$0.1524/kWh; PPFD regression ln(Weight) = -0.422383 + 0.571854 x ln(PPFD), R2=0.52 (p=0.0000); PPFD trial means by tray/light/position/time-of-day (Table 7) not reducible to a single trial mean, range only across conditions ~17-468 umol/s/m2; total basil production across the whole experiment 58.79 kg (abstract).

fernandezcabanasComparativeAnalysisHorizontal2020-T2

Fish

Field	Value
Fish	White sturgeon (Acipenser transmontanus) and green sturgeon (Acipenser medirostris), <1 yr
Protein	46
P	1.1
Feed routine	Fed once daily (200 g/tank/day); time of day not stated
Feed regime	Sinking complete feed for trout/steelhead (Skretting, Tooele, UT, USA), 46% crude protein, 12% crude fat, 3% crude fiber, 1.1% phosphorus

Water

Field	Value
Water recycle	69.4 L/min
Water volume in the system	473 L (sump, plant loop only; excludes 2000 L fish tanks, decoupled)
Water type	Fish effluent (decoupled, sourced from sturgeon RAS reared on well water)
pHOptimal	6.0-6.5 (target range, maintained with phosphoric+citric acid pH Down dosing; no measured trial mean reported)
NO3-N	57.5 +/- 18.4 mg/L (NO3-N, LaMotte Nitrate Nitrogen tablet kit, target ~60 ppm, stated as applying ‘in all systems’)

Plant

Field	Value
Plant	Basil (Ocimum basilicum), cv. Nufar (Fusarium-resistant Italian large-leaf type)
Details	Pelleted seeds sown into propagation trays with RO water; transplanted 14 days after germination into Q plugs media pots; harvested 35 days after transplant
Plant Category	Specialty herb (p.4, “popular specialty herb”)
Days Plant after transplant	35
Plants/m2	67 (paper states “around 67 plants m-2”; 200 plants/tray on a 1.22x2.44 m tray)
Plant height	33.55 +/- 0.49 cm (Table 6, Vertical level 3 only; see WARN-CHECK)
Leaf count	24.06 +/- 0.55 (Table 6, Vertical level 3 only, n.s. p=0.0745; see WARN-CHECK)
Plant fresh weight	13.97 +/- 0.33 g/plant (Table 6, Vertical level 3 only; see WARN-CHECK)

System & Setup

Field	Value
System type	Vertical flood-and-drain rack system (6 independent racks, 3 shelves/levels each at 0.61/1.22/1.83 m height), decoupled aquaponics; levels 1-2 randomized to LED or natural sunlight (one of each per rack per level); level 3 always natural sunlight only
Media Details	Q plugs rooting media in pots labelled 0.102 x 0.102 x 0.089 m3 (p.4-5, likely 3 linear dimensions in m with m3 as product unit); 200 pots/tray in 20x10 grid; tray 1.22x2.44x0.15 m (Botanicare, Chandler, AZ)
Biological system already in use	Y (Nutrients sourced from 3 pre-existing recirculating aquaculture systems (RAS) at UC Davis Center for Aquatic Biology and Aquaculture (CABA) rearing white and green sturgeon (<1 yr) in 2000 L tanks on well water with bead-filter mechanical/biological filtration; decoupled — effluent transferred to plant loop only, no return flow to fish)
Iron supplemented	Y (0.2 g chelated Fe-DTPA (13% Fe) dissolved weekly per sump, raising Fe concentration by 0.055 mg/L; total Fe applied per system 0.182 g vs. estimated plant uptake need 0.174 g)
pH Buffers	Y (Phosphoric + citric acid blend (pH Down, General Hydroponics, Sebastopol, CA) used to lower pH to target 6.0-6.5 range)
Climate control	Y (280 m2 polycarbonate greenhouse at UC Davis; a cooling system located between levels 1 and 2 of vertical racks affected north-end air temperature; no setpoints given)
Artificial Lighting	Y (Four Infinity 2.0 (Thrive Agritech, New York, NY) LED bars per shelf (1.2 m length, 65 W each, 145 umol/s output, full-spectrum white with blue+red peaks) on randomly assigned shelves at levels 1 and 2 (one LED-supplemented and one sunlight-only shelf per rack per level, randomized by number generator); operated 14 h/day, 06:30-20:30. Level 3 always natural sunlight only, no LED.)
Nutrient supplemented	Y (1% KNO3 foliar feed applied twice weekly by manual sprayer (potassium supplementation without affecting solution pH); plus Fe-DTPA (see Iron column))
Equipment	Botanicare 1.22x2.44x0.15 m flood-and-drain trays; 473 L sump; ActiveAqua 1100 GPH (4164 L/h) submersible pump; Bluelab combo meter (pH/EC/temp); LaMotte Nitrate Nitrogen tablet test kit; MQ-500 Apogee quantum meter (PPFD); Infinity 2.0 LED bars (vertical racks levels 1-2 only)
Control Parameters	Target NO3 ~60 ppm (achieved 57.5+/-18.4 mg/L); target pH 6.0-6.5; flood irrigation 2x/day (07:00, 14:00) for 15 min to ~0.02 m depth; supplemental lighting 14 h/day 06:30-20:30 where applied (vertical racks levels 1-2 only)
Combination	Basil (Ocimum basilicum cv. Nufar) grown on decoupled sturgeon RAS effluent; horizontal vs. vertical (3-tier) flood-and-drain racks; vertical racks additionally randomized for LED supplementation at levels 1-2 vs. natural sunlight only

Site

Field	Value
Region	North America
Country	USA
Lat	38.5386
Long	-121.7817

Results & Statistics

Field	Value
Measured Unit	kg/m2 (production per unit area, Table 3); g/plant (fresh weight), cm (height), count (leaves, nodes) (Table 6)
Statistic Details	Generalized linear model (GLM); Shapiro-Wilk normality test; Kruskal-Wallis for non-normal data; Tukey HSD and Games-Howell post-hoc; p<0.05; STATGRAPHICS Centurion XVIII v18.1.12
Statistically analysed	Y
Replicates (n)	18 (6 racks x 3 levels); 20 monitored plants/tray
AP	2427 g fw/m2 (UNIT CONVERSION from 2.427 kg/m2, Table 3)

Experimental Remarks: TRIAL DEFINITION: T2 = Vertical decoupled aquaponic system (6 independent racks x 3 shelf levels [0.61/1.22/1.83 m] = 18 tray replicates), levels 1 and 2 randomized to LED supplementation or natural sunlight only (one of each per rack per level), level 3 always natural sunlight only. Paired comparison = T1 (Horizontal system). No hydroponic-only control exists in this paper; HYD columns are NA throughout. | WARN-CHECK Plant height/leaf count/fresh weight aggregation level: Table 6 gives a single Horizontal-vs-Vertical comparison, but ‘Vertical’ there means level 3 (top shelf) ONLY (13.97+/-0.33 g weight; 33.55+/-0.49 cm height; 24.06+/-0.55 leaves, n.s. p=0.0745; 6.18+/-0.05 nodes), not a whole-rack aggregate. Table 4 separately reports level 1 and level 2 means split further by lighting (Sunlight vs. Sunlight+LED): weight 7.13/8.53/11.52/17.04 g; height 20.04/26.13/20.48/33.23 cm; leaves 9.30/11.28/16.82/24.65 for Level1-Sun/Level2-Sun/Level1-LED/Level2-LED respectively (Table 4); Table 5 gives node-count main effects only (Level1=4.49, Level2=5.22; Sunlight=4.55, LED=5.16), no per-cell interaction means. No single reported mean represents ‘the whole vertical system’ for per-plant weight/height/leaf/node (only production kg/m2, Table 3, is aggregated across the entire vertical system). Recorded Table 6’s level-3-only value as the basis because it is the only place the paper directly compares Horizontal against a Vertical value for these per-plant metrics; all Level 1/2 sub-condition means listed here as candidates. Added to REVIEW.md. | NO COLUMN: Production per unit area 2.427 +/- 0.155 kg/m2 (Table 3, vertical aggregate across all levels/lighting), 160% greater than horizontal (Table 3, p=0.0037); by level x lighting (Table 1): Level1-Sun 0.451, Level1-LED 0.750, Level2-Sun 0.685, Level2-LED 1.191, Level3-Sun(only) 0.981 kg/m2; simulated rack economics (Table 8). | NOT DERIVED, left NR: fish stocking density, FCR, SGR, fish size initial/final, fish weight gain, fish survival rate, fish trial duration, Total Feed (kg) (only a 200 g/day/tank feeding rate given, no total or duration to sum over), SPAD, plant dry matter (never measured), Water temperature/EC/DO (measured with a Bluelab meter per Methods but no summary values reported anywhere in Results), Daily Water exchange rate (only twice-weekly absolute replenishment volumes given, not a % rate), Nitrogen % and Potassium % of feed (only crude protein 46% and phosphorus 1.1% given). | Aq pH NR (measured trial mean not reported); target 6.0-6.5 recorded in pHOptimal. | UNIT CONVERSION ONLY: water recycle 4164 L/h -> 69.4 L/min; coordinates 38 deg 32’18.8” N / 121 deg 46’54.2” W -> decimal 38.5386 / -121.7817. Production 2.427 kg/m2 -> 2427 g/m2 (AP column). | Vertical-rack-specific effluent replenishment: ‘an additional 110 L of sturgeon effluent twice a week’ per rack (p.5) to offset evapotranspiration losses; not necessarily equal for horizontal (see T1 remarks, NR there). | Funding statement says “This research received no external funding,” but Acknowledgments separately credit University of Seville Plan Propio de Investigacion for financing V.M. Fernandez-Cabanas’s research stay at UC Davis — recorded both, not treated as a contradiction (internal institutional travel funding vs. external competitive grant funding are different things). | Fish Category NR — paper names species (white/green sturgeon, <1 yr) but does not assign a category label (cf. vault precedent, e.g. Alcarraz 2018). | MINOR: pot dimensions given as “0.102 x 0.102 x 0.089 m3” (p.4-5) — likely intends meters per dimension with m3 as the resulting product unit label, an unusual but not numerically inconsistent notation; does not affect any extracted cell. Plants/m2 (67, ‘around 67 plants m-2’) and tray dimensions (1.22x2.44 m, 200 plants/tray) identical for horizontal and vertical trays (same equipment, p.4). | NO COLUMN: economic analysis (Table 8) — lighting cost $6.53/m2persupplementedlevel,costperkgproductionincrement$21.84 (Level1-LED) / $12.25(Level2-LED)/$15.70 (both levels), basil price assumed $13.58/kg,electricity$0.1524/kWh; PPFD regression ln(Weight) = -0.422383 + 0.571854 x ln(PPFD), R2=0.52 (p=0.0000); PPFD trial means by tray/light/position/time-of-day (Table 7) not reducible to a single trial mean, range only across conditions ~17-468 umol/s/m2; total basil production across the whole experiment 58.79 kg (abstract).