Basil, Ocimum basilicum, yield in northern latitudinal aquaponic growing conditions
Metadata
- Cite key: abbeyBasilOcimumBasilicum2022
- Item type: Journal article (Applied Studies)
- Authors: M. Abbey, N.O. Anderson, C. Yue, M. Schermann, N. Phelps, P. Venturelli, Z. Vickers
- Affiliation: University of Minnesota (Horticultural Science; Applied Economics; Fisheries, Wildlife & Conservation Biology; Food Science and Nutrition); Ball State University (Biology)
- Journal: Journal of the World Aquaculture Society 53 (2022) 77-94
- Date: 01/2022
- Date added: 2026-07-29
- DOI: 10.1111/jwas.12819
- Funding: Minnesota Agricultural Experiment Station; University of Minnesota Global Food Ventures MNDrive
- URL: https://doi.org/10.1111/jwas.12819
- PDF:
Abbey et al. - 2022 - Basil, Ocimum basilicum, yield in northern latitud.pdf
Opinion
Methodologically ambitious (5 parallel aquaponic configurations + a shared soilless control, 13 months, GLIMMIX with lognormal fit) but the reporting is sloppy: a fresh-weight figure disagrees with its own table by 10x, a tank water volume is stated two incompatible ways for the same tanks, one treatment’s air-temperature daily integral is given twice with different numbers, and the Methods and Results sections disagree on the winter feed total. None of these are fatal to the paper’s central, well-supported finding (koi suppress basil yield relative to other fish and to the soilless control), but several secondary system-description numbers cannot be trusted at face value. Worth citing for the yield baseline and the koi effect; verify any water-volume or A-frame-yield figure against the original before reuse.
Abstract
The need for yield research is increasing with the popularity of aquaponic food production systems where plants and fish are grown together in a recirculating system. Our objective was to compare the yield of three basil cultivars (“Elenora,” “Genovese,” “Nufar,” Ocimum basilicum) in four configurations (greenhouses: floating raft DWS, ebb & flow, A-frame; warehouse: floating raft; 1-8 tank replications/treatment) as well as a soilless control to establish a baseline for aquaponic producers in northern latitudes. We also tested whether basil yield varied among fish types: yellow perch, Perca flavescens, goldfish, Carassius auratus, tilapia, Oreochromis spp., and koi, Cyprinus carpio. Fish tank water temperature, nitrite, and alkalinity levels differed significantly over seasons (summer, winter). There was a significant decrease in basil production in all aquaponic systems that used koi (greenhouse, warehouse). No significant difference in basil yield was found among greenhouse and aquaponic systems for perch, tilapia, or goldfish nor among the basil “Elenora,” “Genovese,” and “Nufar.” Considerable variation occurred in yield although most of it occurred in fresh weights, being reduced significantly in dry weights. The effects of growing koi fish on reducing aquaponic production of basil is noteworthy, although future research is needed to the exact cause.
Summary
Basil (“Elenora,” “Genovese” Compact Improved, “Nufar”) was grown for 13 months (Jan 2016-Feb 2017) at the University of Minnesota’s St. Paul campus in five parallel systems: floating-raft DWC with yellow perch, A-frame ebb-and-flow with koi, tray ebb-and-flow with goldfish, tray ebb-and-flow with tilapia, and a warehouse floating-raft system with koi — each compared against a shared soilless-medium (potted, fertigated, non-aquaponic) control. Basil was co-grown in the same tanks/pots as day-neutral strawberries and lettuce (reported in companion papers). Harvests were single-cut, every ~4 weeks in summer and ~6 weeks in winter, with at least 6 harvests per treatment; fresh and dry weights were recorded per plant and per m2. Fresh and dry weight yields differed significantly among greenhouse locations, among fish types, and between aquaponic and soilless treatments (GLIMMIX, lognormal fit), but not among the three basil cultivars. The soilless control produced the highest yields; among aquaponic treatments, floating-raft DWC (perch) came closest to the control, while both koi treatments (A-frame and warehouse) produced markedly and significantly lower fresh and dry weights than every other treatment. The authors attribute the koi effect tentatively to the fish species itself (not tank design, feeding rate, or lighting, which were comparable across non-koi treatments) and flag the warehouse’s cooler air temperature (20-21C) as a plausible additional factor specific to that one treatment.
Experiment data
- Location: University of Minnesota, St. Paul campus (44.9883, -93.1810); 4 treatments in greenhouses, 1 (warehouse/koi) in a retrofitted walk-in cooler
- Design: 5 aquaponic treatments (fish type x system type) vs. 1 shared soilless-medium control; randomized block design (A-frame randomized by PVC line instead); split-plot GLIMMIX, koi treatments analyzed separately
- Replicates / n: Tanks per treatment: floating raft DWC 8; A-frame (koi) 2; tray ebb&flow (goldfish) 1; tray ebb&flow (tilapia) 1; warehouse (koi) 1. Soilless control: 32 plants/cultivar
- Duration: 13 months (Jan 2016-Feb 2017); >=6 harvests/treatment, ~4-wk (summer) to ~6-wk (winter) cycles
- Organisms: Basil (Ocimum basilicum) cvs. “Elenora”/“Genovese”/“Nufar” (pooled, ns); Yellow Perch (Perca flavescens), Koi (Cyprinus carpio), Goldfish (Carassius auratus), Tilapia (Oreochromis spp.)
- Statistics: GLIMMIX (lognormal distribution, lowest AIC/BIC), split-plot; SAS v.25
- Basil yield: fresh wt range 144-1,390 g/m2 (9.6-69.5 g/plant) across treatments; dry wt range 22.5-162 g/m2 (1.5-8.1 g/plant); soilless control highest, both koi treatments lowest (p<=.001)
- Water quality (aquaponics): treatment and season both significant for most fish-water parameters (Tables 1-2); nitrate lowest in floating-raft DWC (22.42 ppm, covered raft surface implicated)
Fish x system treatments
| Trial | System | Fish | Tanks |
|---|---|---|---|
| T1 | Floating raft DWC | Yellow perch | 8 |
| T2 | A-frame ebb & flow | Koi | 2 |
| T3 | Tray ebb & flow | Goldfish | 1 |
| T4 | Tray ebb & flow | Tilapia | 1 |
| T5 | Warehouse floating raft (Nelson & Pade F5) | Koi | 1 |
Control (shared, all 5 trials): soilless medium — potted (754 cm3 pots, Sunshine LC8), fertigated (125 ppm N, 20N-4.4P-16.6K), non-aquaponic. Not true recirculating hydroponics — flagged wherever it stands in for a HYD column.
Yield
This paper: Fresh weight (g/m2; g/plant), Table 4: Soilless 1,390.0+/-498.0 (69.5+/-24.9); Floating raft DWC/perch 577.9+/-623.9 (64.0+/-69.1); Tray e&f/goldfish 447.0+/-339.0 (59.6+/-45.2); A-frame/koi UNCLEAR g/m2 (Table 4 prints 4,469.3+/-770.4; body text p.88 states 446.9 for the same treatment, a ~10x gap) / 23.9+/-41.2 g/plant (uncontested); Tray e&f/tilapia 393.0+/-478.5 (52.4+/-63.8); Warehouse/koi 144.0+/-181.5 (9.6+/-12.1). Dry weight, same order: 162.0+/-72.0 (8.1+/-3.6); 53.3+/-52.4 (5.9+/-5.8); 57.8+/-39.0 (7.7+/-5.2); 39.3+/-69.2 (2.1+/-3.7); 42.0+/-54.8 (5.6+/-7.3); 22.5+/-27.0 (1.5+/-1.8). All differences among treatments p<=.001 (Table 3); cultivar ns.
Compared with:
- todo Kiferle Maggini Pardossi 2013 — “Genovese” hydroponic greenhouse, 27.1-64.1 g/plant fresh, 2.96-6.03 g/plant dry (Table 6, secondary/comparative citation in this paper)
- todo Walters Currey 2015 — “Genovese” 34.3 g/plant fresh/2.6 dry; “Nufar” 43.7/3.6 (Table 6)
- todo Walters Currey 2018 — “Nufar” 30.5 g/plant fresh/2.5 dry (Table 6)
- todo Saha Monroe Day 2016 — “Nufar” hydroponic 96.6/9.6 g/plant; aquaponic 150.2/15.9 g/plant (Table 6) — considerably higher than this paper’s own aquaponic yields, cited as context
- todo Rakocy et al. 2007 — unnamed cvs., outdoor aquaponics, 6/18/25 g/plant fresh (field/batch/staggered harvests) — lower range than this paper (Table 6)
- todo Hochmuth Davis Laughlin Simonne 2003 — unnamed cv. hydroponic greenhouse, 15-63 g/plant early yield (Table 6)
Water quality
This paper (Table 2, pooled or seasonal per treatment):
- Aq pH (pooled): Floating raft DWC 8.32; A-frame 7.20; Ebb&flow (goldfish+tilapia tanks pooled) 7.19; Warehouse 7.56
- TAN/Ammonia (pooled, ppm): 0.78 / 0.40 / 0.48 / 0.42 respectively
- NO3-N (pooled, ppm): 22.42 (floating raft, significantly lower than all others) / 70.77 / 70.77 / 72.94
- NO2-N and Water temperature: only seasonal (summer/winter) means reported in Table 2, no single pooled trial mean — recorded as ranges in trials.csv with “seasonal means only” note
- Alkalinity and TDS: reported in Table 2 but have no dedicated trials.csv column — see NO COLUMN notes in Experimental Remarks
- DO: only a target threshold (>6 ppm) is stated; no measured mean anywhere — NR
Linked claims
- Koi may suppress basil yield in aquaponic systems
- Cooler air temperature reduces basil yield
- Basil cultivar choice does not significantly affect aquaponic yield
- N deficiency causes leaf chlorosis in aquaponic basil
Citations to chase
- todo Kiferle Maggini Pardossi 2013 — hydroponic “Genovese” basil, N nutrition and rosmarinic acid
- todo Walters Currey 2015 — hydroponic basil systems/cultivars comparison
- todo Walters Currey 2018 — nutrient solution concentration and DLI effects on hydroponic basil
- todo Saha Monroe Day 2016 — basil growth/yield/quality/nutrition under soilless systems, aquaponic vs. hydroponic
- todo Rakocy Bailey Shultz Danaher 2007 — organic waste from aquaculture as hydroponic nutrient source, outdoor basil
- todo Hochmuth Davis Laughlin Simonne 2003 — organic nutrient sources, hydroponic basil
Extraction notes
Re-extracted from a full read of the PDF under the updated (severity-tiered) schema. Five real contradictions/inconsistencies survived scrutiny (3 BLOCK, 1 MATERIAL, 3 MINOR — see below), so quality: suspect was added.
⚠️BLOCK — A-frame/koi fresh-weight yield (g/m2), ~10x table-vs-text gap. Table 4 (p.89) prints 4,469.3+/-770.4 g/m2 for A-frame ebb&flow/koi. Results text (p.88) states the same treatment’s g/m2 value was “the third lowest” at “446.9 g/m2” — only consistent with the treatment ranking (Warehouse 144 < Tray/tilapia 393 < A-frame ~447 < Tray/goldfish 447 < Floating raft 577.9 < Soilless 1,390) if 446.9 is used; 4,469.3 would make it the highest value, contradicting the same sentence. No way to determine which is the typo. Per the schema’s own BLOCK example (“a yield reported 10x apart in table and text”), the cell is UNCLEAR; both candidates recorded. Because this is the paper’s primary outcome, Statistically analysed is also set to UNCLEAR for T2, per the BLOCK-on-primary-outcome rule (the ANOVA significance itself, p<=.01, is not in question — only this specific magnitude).
⚠️BLOCK — floating raft DWC daily temperature integral, stated twice with different values. p.83-84, same subsection, same treatment: “a 23.6+/-0.8C daily integral, based on a temperature setpoint of 23.5C” and, later with no reconciling statement, “The daily temperature integral was 22.3+/-0.9C.” I considered whether the second figure could instead be a restatement of water temperature (it sits near the water-quality-monitoring sentence, and is numerically close to Table 2’s water-temp range of 22.04-22.43) rather than a duplicate of the air-temperature claim — but “daily temperature integral” is used consistently elsewhere in this paper (soilless, A-frame, tray e&f) to mean greenhouse air temperature paired with a setpoint, so I did not apply that reading. Treated as a genuine duplicate with no basis to prefer either; Average room Temperature = UNCLEAR for T1.
⚠️BLOCK — floating raft/A-frame/tray tank water volume, two figures for the same tanks. p.83: “the tanks were filled with ~378.541 L (~100 gals.)” … “The same tanks and volumes were also used in the A-frame ebb and flow and the tray ebb and flows systems. … the water volume/tank was ~550 L or 0.55 m3.” Both describe the actual fill volume of the identical tank type (696.5 L nominal capacity is a separate, uncontested spec). No textual basis to prefer 378.5 L or 550 L. Water volume in the system = UNCLEAR for T1-T4. T5 (warehouse) has its own distinct, singly-stated 416 L tank — unaffected.
⚠️MATERIAL — total feed, Methods (2,100 mg/tank winter) vs. Results (4,200 mg/tank winter). p.86 (Methods): “the winter weight gain … was based on 2,100 mg feed/tank.” p.89 (Results): “Feed totals for the perch … over the 4- and 6-week production seasonal periods were 2,800 and 4,200 mg/tank which meant the 30 fish/tank consumed, on average, a total of 93.2 and 140 mg.” Recomputing to check (not to derive): 4,200/30 = 140 (matches); 2,800/30 = 93.3 (matches “93.2”); the same 2,800/4,200 mg/tank basis also reproduces the paper’s 90-goldfish figures (2,800/90=31.1, 4,200/90=46.7, matching “31/46.6”) and 32-koi figures (2,800/32=87.5, 4,200/32=131.25, matching “87.5 and 131.2”). All six of the Results section’s own sub-figures are self-consistent only with 4,200 mg/tank for winter, giving a clear basis to prefer it over the Methods section’s 2,100. Total Feed (kg) recorded from the Results-derived figure (0.0028 kg/tank summer, 0.0042 kg/tank winter, per production cycle — not a full-trial cumulative total, since the paper never states how many summer vs. winter harvests occurred). The paper separately states an approximate “50 mg/tank/day” feeding rate elsewhere, which does not precisely reconcile with either period total; treated as a rough approximation, not a further contradiction.
⚠️MINOR — floating raft DWC plant-count arithmetic. p.81: “12 plants/tank (4 basil+4 lettuce+4 strawberries/tank) x 8 tanks = 72 total” — 12x8=96, not 72 (8 tanks is independently confirmed on p.83, “eight aluminum tanks”). Doesn’t affect the independently-stated Plants/m2 (9.03) figure.
⚠️MINOR — two small rounding restatements of dry weight (g/plant). Tray e&f/goldfish: Table 4 gives 7.7+/-5.2, Results text (p.89) restates “7.68.” Soilless control: Table 4 gives 8.1+/-3.6, Discussion text restates “8.14.” Simple rounding, Table 4 values retained.
Checked and NOT flagged (not a contradiction, per the new carve-out): A-frame ebb&flow (p.85) states the same shared-greenhouse daily integral used for the tray systems (“21.7+/-0.4C … set-point 21.5C,” p.84) and then adds “although daily temperatures were 23.5+/-0.9C.” Read as a genuine A-frame-specific microclimate note (the A-frame structure sits closer to the glazing) rather than a duplicate contradictory claim — signalled by the explicit “although” and by the fact this second figure is not given in the otherwise-identical tray ebb&flow passage. Recorded 23.5+/-0.9C as T2’s Average room Temperature (the more specific figure); the shared 21.7+/-0.4C is noted as the general greenhouse value common to T2/T3/T4.
[not reported] fields, grouped:
- Fish: stocking density (kg/m3), FCR, SGR, protein/N/P/K of feed for goldfish (brand gives fat/fiber/moisture/ash but not protein %), % of body weight ration, fish size initial/final, fish survival rate, fish weight gain, exact fish trial duration in days (only “13-month period, Jan 2016-Feb 2017” given, exact start/end day not stated), FUE AP/HYD, WUE, Fish Category, Water type, Water classification, Daily Water exchange rate, DO (measured — only a >6 ppm target given), EC (paper reports TDS instead, a different unit — not converted)
- Plant: Plant height, Leaf count, SPAD, Tissue nitrate (AP and HYD) — no tissue analysis performed in this paper
- System: Water recycle rate given only for floating raft DWC (~4 L/min); NR for the other 4 treatments (pump models given without a restated flow rate)
[unclear] fields: none beyond the three UNCLEAR cells created by the BLOCK contradictions above (A-frame yield g/m2 and Statistically analysed; floating raft Average room Temperature; T1-T4 Water volume in the system).
New tags introduced: Meta/Fish/Yellow-Perch, Meta/Fish/Koi, Meta/Fish/Goldfish, Meta/Fish/Tilapia, Meta/Plant/Basil, Meta/Region/North-America (all checked against the CLAUDE.md-listed existing facets; none of these four fish species or Basil existed yet under Meta/Fish/ or Meta/Plant/ in the facets enumerated in CLAUDE.md, so they’re new leaves under the existing hierarchies, not new facets).
New wikilink targets: Basil (Ocimum basilicum), Yellow Perch (Perca flavescens), Koi (Cyprinus carpio), Goldfish (Carassius auratus), Tilapia (Oreochromis spp.), Basil yield, Water quality (aquaponics), Koi may suppress basil yield in aquaponic systems, Cooler air temperature reduces basil yield, Basil cultivar choice does not significantly affect aquaponic yield, N deficiency causes leaf chlorosis in aquaponic basil — none pre-existed in this (currently empty) notes/ folder; candidates for merge if a fuller vault already has these under different names.
Source: Abbey et al. - 2022 - Basil, Ocimum basilicum, yield in northern latitud.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
abbeyBasilOcimumBasilicum2022-T1
Fish
| Field | Value |
|---|---|
| Fish | Yellow perch (Perca flavescens) |
| Protein | 35 (% min crude protein, Zeigler Finfish Bronze 3mm/5mm pellets; also 5% min crude fat, 4% max crude fiber) |
| Feed routine | Twice daily, 08:00-09:00 and 15:00-16:00 (fish feeding/monitoring); water quality checked in 50% of tanks each day, alternating |
| Feed regime | Zeigler Finfish Bronze pellets (Florida Tropical Fish Farmer’s Coop), 3mm then 5mm as fish aged, 35% min crude protein/5% min crude fat/4% max crude fiber; ~50 mg/tank/day maintenance diet (nitrogen-source diet, not growth-maximizing) |
| Total Feed (kg) | ~0.0028 kg/tank (summer, ~4-wk cycle); ~0.0042 kg/tank (winter, ~6-wk cycle) - per production cycle, NOT a full-trial cumulative total |
Water
| Field | Value |
|---|---|
| Water recycle | Y |
| Water volume in the system | UNCLEAR (378.5 L vs 550 L, same tank type as T2-T4; see remarks) |
| Daily Water exchange rate | ~4 L/min (Danner Supreme 700 GPH pump recirculation rate; not a stated fractional daily exchange rate) |
| Aq pH | 8.32 |
| pHOptimal | 7 |
| Water temperature | 22.04-22.43 (seasonal means only, no pooled trial mean reported) |
| TAN / NH4-N | 0.78 |
| NO2-N | 0.49-0.67 (seasonal means only, no pooled trial mean reported) |
| NO3-N | 22.42 |
Plant
| Field | Value |
|---|---|
| Plant | Basil (Ocimum basilicum), cvs. ‘Elenora’/‘Genovese’/‘Nufar’ (pooled, cultivar ns) |
| Details | Co-grown with day-neutral strawberries (‘Albion’/‘Portola’/‘Evie 2’) and lettuce (‘Rex’/‘Skyphos’/‘Salanova Series Mix’) in the same tanks/pots (reported in companion papers, not extracted here); single-cut harvest at marketable maturity, cut at top of soilless media/rockwool plug; >=6 harvests/treatment, ~4-wk (summer) to ~6-wk (winter) cycles; fresh weight then oven-dried 77C for 7 days for dry weight; transplant seedling FW (66.59+/-24.81 mg pooled, ns among cultivars) ignored as negligible |
| Plant Category | Herbs (culinary herb, Lamiaceae) |
| Days Plant after transplant | ~28 (summer) to ~42 (winter) days per harvest cycle (>=6 harvests/treatment; derived from stated ~4-wk/~6-wk harvest intervals) |
| Plants/m2 | 9.03 |
| Plant fresh weight | 64.0 +/- 69.1 g/plant |
| Plant dry matter | 5.9 +/- 5.8 g/plant |
System & Setup
| Field | Value |
|---|---|
| System type | Floating raft DWC (greenhouse) |
| Media Details | Rockwool plugs in 12cm net cups (max 16 plants/4x4 grid raft); 2 floating rafts/tank (60x60x5.5cm Owens Corning FOAMULAR 150 R-10); 8 aluminum tanks (193x77.5x75cm, 6.5cm walls, 696.516 L nominal capacity); 2 hemispherical plastic biofilters/tank (68x47x26cm), 2cm granite gravel, ammonium chloride inoculated |
| Biological system already in use | Y (Biofilters (2/tank) established and inoculated with ammonium chloride (1g/biofilter) prior to fish introduction; low-density goldfish (~25-30/tank) initially used to cycle the biofilter, later replaced with the tested yellow perch (~25-30/tank)) |
| Air supplement | Y (Third pump outlet discharged directly to fish tank for additional aeration/circulation, separate from the two biofilter-directed outlets) |
| Climate control | Y (Computerized greenhouse (St. Paul Plant Growth Facilities), Argus Controller for automatic air temp recording; electric generator power backup; HPS-HID supplemental lighting during winter/cloudy days) |
| Artificial Lighting | Y (400W HPS-HID lamps, min. 150 umol/m2/s at plant level, 16hr photoperiod (06:00-22:00), supplemented during winter months/cloudy days) |
| Equipment | Danner Supreme 700 GPH magnetic drive pump (~4 L/min to biofilters + direct tank discharge); automatic bell siphon per biofilter tank (gravel water level 2-15cm cycling); Argus Controller (air temp logging) |
| Control Parameters | pH 6.5-7.5 (optimum 7); DO >6 ppm target; ammonia/nitrite/nitrate <0.75 ppm each (feeding suspended if exceeded); temperature setpoint 23.5C |
| Combination | Yellow perch and basil (co-grown with day-neutral strawberries and lettuce, reported in companion papers); floating raft DWC vs. soilless control |
Site
| Field | Value |
|---|---|
| Region | North America |
| Country | USA |
| Lat | 44.9883 |
| Long | -93.1810 |
| Average room Temperature | UNCLEAR (23.6+/-0.8C vs 22.3+/-0.9C, both stated for the same treatment with no reconciling statement; see remarks) |
Results & Statistics
| Field | Value |
|---|---|
| Measured Unit | g fw/m2; g fw/plant (fresh weight); g dw/m2; g dw/plant (dry weight) |
| Statistic Details | GLIMMIX (lognormal distribution, lowest AIC/BIC), split-plot design; fish type = treatment, cultivar = variable; koi treatments (A-frame, warehouse) analyzed separately from non-koi; Tukey HSD for water-quality mean separations (Table 2); ANOVA for treatment comparisons (Table 3); SAS v.25 |
| Statistically analysed | Y |
| Replicates (n) | 8 |
| AP | 577.9 |
| HYD | 1390.0 |
Experimental Remarks: TRIAL DEFINITION: T1 = Floating raft DWC (greenhouse) with yellow perch (25-30 fingerlings/tank, initially goldfish for biofilter cycling, replaced with 30 perch/tank), 8 tank replicates. Paired control = shared soilless-medium (potted, fertigated, non-aquaponic) treatment recorded in HYD columns - NOT true hydroponics (no recirculating nutrient solution), flagged wherever it stands in for HYD. | BLOCK Average room Temperature: p.83-84 states the floating raft DWC greenhouse daily temperature integral TWICE with different values in the same subsection: ‘23.6+/-0.8C daily integral, based on a temperature setpoint of 23.5C’ and later ‘The daily temperature integral was 22.3+/-0.9C,’ with no reconciling statement. Considered whether the second figure was a restatement of water temperature (numerically close to Table 2’s water-temp range 22.04-22.43C) rather than air temperature, but ‘daily temperature integral’ is used consistently elsewhere (soilless, A-frame, tray e&f) to mean greenhouse air temperature paired with a setpoint - so this reading was not applied. No basis to prefer either candidate; both recorded, cell = UNCLEAR. | BLOCK Water volume in the system: p.83 gives the SAME tank type (used identically in floating raft DWC, A-frame ebb&flow, and tray ebb&flow) two incompatible fill-volume figures: ‘’131.2’) per-fish figures elsewhere in the same paragraph. All of Results’ own sub-figures are self-consistent only with 4,200 mg/tank for winter; Results-derived figure preferred over Methods’ 2,100. Total Feed (kg) recorded as 0.0028 kg/tank (summer, ~4-wk cycle) / 0.0042 kg/tank (winter, ~6-wk cycle) - a per-production-cycle amount, NOT a full 13-month cumulative total (paper never states number of summer vs winter cycles). A separate ~50 mg/tank/day approximate feeding rate is also given elsewhere; does not precisely reconcile with either period total, treated as a rough approximation not a further contradiction. | MINOR plant-count arithmetic: p.81 states ‘12 plants/tank (4 basil+4 lettuce+4 strawberries/tank) x 8 tanks = 72 total’ - 12x8=96, not 72 (8 tanks independently confirmed p.83 ‘eight aluminum tanks’). Does not affect the independently-stated Plants/m2 (9.03) figure used here. | UNIT CONVERSION ONLY: coordinates 44 59’17.8”N/93 10’51.6”W -> decimal 44.9883/-93.1810; harvest interval ‘~4 weeks (summer)/~6 weeks (winter)’ -> ~28/378.541 L (’46.6’) and koi (2,800/32=87.5, 4,200/32=131.25100 gals.)’ to allow for floating rafts, and later ‘the water volume/tank was ’31’, 4,200/90=46.7550 L or 0.55 m3.’ Both describe actual fill volume (distinct from the uncontested 696.516 L nominal tank capacity). No textual basis to prefer either; both candidates recorded, cell = UNCLEAR for T1-T4 (T5/warehouse has its own distinct, singly-stated 416 L tank, unaffected). | MATERIAL Total Feed: Methods section (p.86, Statistical analyses) states winter weight gain ‘was based on 2,100 mg feed/tank,’ but Results (p.89) states winter feed totals were ‘4,200 mg/tank,’ recomputing to 140 mg/fish for 30 perch/tank (2,800/30=93.3’93.2’ summer; 4,200/30=140 matches winter exactly); the same 4,200/2,800 mg/tank basis also reproduces the paper’s own goldfish (2,800/90=31.142 days (Days Plant after transplant), prefixed ''. | NO COLUMN: Alkalinity (pooled 46.72-59.79 ppm GH/KH, Table 2) and TDS (1.27 ppt summer, winter missing - too many missing data points) reported but have no dedicated trials.csv column. Alkalinity/TDS both significant by season (p<=.001); all alkalinity values within acceptable 40-65 ppm GH/KH range per Getter 2015, no treatment required. | NOT DERIVED, left NR: Fish stocking density (kg/m3 - only fish/tank counts and 20-30 fish/tank ranges given); FCR; SGR; feed N/P/K; % of body weight ration (fed a fixed ~50mg/tank/day maintenance diet, not by body-weight %); fish size initial/final; fish biomass created; fish survival rate; fish weight gain; exact fish trial duration in days (only ‘13-month period, Jan 2016-Feb 2017’ given); FUE AP/HYD; WUE; Fish Category; Water type; Water classification; Dissolved Oxygen (only a >6 ppm target stated, no measured mean); EC (paper reports TDS instead, a different unit, not converted); SPAD; Plant height; Leaf count; Tissue nitrate AP/HYD (no tissue analysis performed in this paper). | Daily Water exchange rate: ~4 L/min is the pump recirculation flow rate (Danner Supreme 700 GPH, split to 2 biofilters + 1 direct discharge), not a stated fractional daily exchange rate - recorded with this caveat; NR for T2-T5 (pump models given without a restated flow rate). | Secondary/comparative citations (Table 6, todo): Kiferle Maggini Pardossi 2013 (Genovese hydroponic, 27.1-64.1 g/plant fresh/2.96-6.03 dry); Walters & Currey 2015 (Genovese 34.3/2.6; Nufar 43.7/3.6); Walters & Currey 2018 (Nufar 30.5/2.5); Saha Monroe Day 2016 (Nufar hydroponic 96.6/9.6; aquaponic 150.2/15.9 - higher than this paper’s own aquaponic yields); Rakocy et al. 2007 (outdoor aquaponics, unnamed cvs., 6/18/25 g/plant); Hochmuth Davis Laughlin Simonne 2003 (hydroponic, 15-63 g/plant). All secondary, cited as context, not this paper’s own data.
abbeyBasilOcimumBasilicum2022-T2
Fish
| Field | Value |
|---|---|
| Fish | Koi (Cyprinus carpio) |
| Protein | 35 (grower pellets, min protein); 52 (starter crumble, min protein) - see Feed regime |
| Feed routine | Twice daily, 08:00-09:00 and 15:00-16:00 (fish feeding/monitoring); water quality checked in 50% of tanks each day, alternating |
| Feed regime | Same feed as tilapia (Skretting): starter crumble #2 (52% min protein, 16% min oil, 9% max moisture, 3% max fiber, 12% max ash, 18.2% digestible energy) then 1.5/2.5/3.5mm pellets (35% min protein, 7% min oil, 9% max moisture, 5% max fiber, 9% max ash, 13% digestible energy) as fish grew; rate ~50 mg/tank, varied by weekly fish size/density |
| Total Feed (kg) | ~0.0028 kg/tank (summer, ~4-wk cycle); ~0.0042 kg/tank (winter, ~6-wk cycle) - per production cycle, NOT a full-trial cumulative total |
Water
| Field | Value |
|---|---|
| Water recycle | Y |
| Water volume in the system | UNCLEAR (378.5 L vs 550 L, same tank type as T1/T3/T4; see remarks) |
| Daily Water exchange rate | NR (pump model given, flow rate not restated for this treatment) |
| Aq pH | 7.20 |
| pHOptimal | 7 |
| Water temperature | 22.17-24.19 (seasonal means only, no pooled trial mean reported) |
| TAN / NH4-N | 0.40 |
| NO2-N | 3.49-7.30 (seasonal means only, no pooled trial mean reported) |
| NO3-N | 70.77 |
Plant
| Field | Value |
|---|---|
| Plant | Basil (Ocimum basilicum), cvs. ‘Elenora’/‘Genovese’/‘Nufar’ (pooled, cultivar ns) |
| Details | Co-grown with day-neutral strawberries (‘Albion’/‘Portola’/‘Evie 2’) and lettuce (‘Rex’/‘Skyphos’/‘Salanova Series Mix’) in the same tanks/pots (reported in companion papers, not extracted here); single-cut harvest at marketable maturity, cut at top of soilless media/rockwool plug; >=6 harvests/treatment, ~4-wk (summer) to ~6-wk (winter) cycles; fresh weight then oven-dried 77C for 7 days for dry weight; transplant seedling FW (66.59+/-24.81 mg pooled, ns among cultivars) ignored as negligible |
| Plant Category | Herbs (culinary herb, Lamiaceae) |
| Days Plant after transplant | ~28 (summer) to ~42 (winter) days per harvest cycle (>=6 harvests/treatment; derived from stated ~4-wk/~6-wk harvest intervals) |
| Plants/m2 | 18 |
| Plant fresh weight | 23.9 +/- 41.2 g/plant |
| Plant dry matter | 2.1 +/- 3.7 g/plant |
System & Setup
| Field | Value |
|---|---|
| System type | A-frame ebb and flow (greenhouse) |
| Media Details | Rockwool plugs, A-frame PVC plant lines (6 tubes/A-frame, randomized by line); 2 aluminum tanks (same 193x77.5x75cm spec as floating raft DWC); 4 hemi-cylindrical plastic biofilters/A-frame (68x47x26cm), 3-4cm lava rock, ammonium carbonate inoculated; airlift pumps tank-to-biofilter, Danner Supreme 700 GPH mag drive pump biofilter-to-A-frame lines |
| Biological system already in use | Y (4 hemi-cylindrical lava-rock biofilters/A-frame, ammonium carbonate inoculated (shared start-up method with tray ebb&flow system)) |
| Climate control | Y (Same greenhouse computerized climate control as floating raft DWC/tray ebb&flow (Argus Controller, generator backup, HPS-HID supplemental lighting)) |
| Artificial Lighting | Y (400W HPS-HID lamps, min. 150 umol/m2/s at plant level, 16hr photoperiod, supplemented during winter/cloudy days) |
| Equipment | Airlift pumps (fish tank to biofilter); Danner Supreme 700 GPH mag drive pump (biofilter to A-frame lines); 4 hemi-cylindrical lava-rock biofilters/A-frame |
| Control Parameters | pH 6.5-7.5 (optimum 7); DO >6 ppm target; ammonia/nitrite/nitrate <0.75 ppm each (feeding suspended if exceeded); temperature setpoint 21.5C (A-frame daily temps ran warmer, 23.5+/-0.9C) |
| Combination | Koi and basil (co-grown with day-neutral strawberries and lettuce, reported in companion papers); A-frame ebb and flow vs. soilless control |
Site
| Field | Value |
|---|---|
| Region | North America |
| Country | USA |
| Lat | 44.9883 |
| Long | -93.1810 |
| Average room Temperature | 23.5 +/- 0.9C (A-frame-specific microclimate value; shared greenhouse general value 21.7 +/- 0.4C, setpoint 21.5C, noted in remarks - not a contradiction, see remarks) |
Results & Statistics
| Field | Value |
|---|---|
| Measured Unit | g fw/m2; g fw/plant (fresh weight); g dw/m2; g dw/plant (dry weight) |
| Statistic Details | GLIMMIX (lognormal distribution, lowest AIC/BIC), split-plot design; fish type = treatment, cultivar = variable; koi treatments (A-frame, warehouse) analyzed separately from non-koi; Tukey HSD for water-quality mean separations (Table 2); ANOVA for treatment comparisons (Table 3); SAS v.25 |
| Replicates (n) | 2 |
| AP | UNCLEAR (Table 4: 4,469.3 +/- 770.4 g/m2; Results text p.88: 446.9 g/m2 - ~10x gap, see remarks; per-plant figure 23.9 g/plant uncontested) |
| HYD | 1390.0 |
Experimental Remarks: TRIAL DEFINITION: T2 = A-frame ebb & flow (greenhouse) with koi, 2 tank replicates (~20-30 fish/tank, decreasing as fish grew). Paired control = shared soilless-medium control (see T1 remarks for description/caveats). | BLOCK A-frame/koi fresh-weight yield, ~10x table-vs-text gap: Table 4 (p.89) prints 4,469.3+/-770.4 g/m2 for this treatment. Results text (p.88) states the same treatment’s g/m2 value was ‘the third lowest’ at ‘446.9 g/m2’ - only consistent with the treatment ranking (Warehouse 144 < Tray/tilapia 393 < A-frame ~447 < Tray/goldfish 447 < Floating raft 577.9 < Soilless 1,390) if 446.9 is used; 4,469.3 would make A-frame the HIGHEST value, directly contradicting the same sentence. No textual basis to determine which is the typo; both candidates recorded (Table 4: 4,469.3+/-770.4; Results text: 446.9), cell (AP, and the m2 component of any composite yield figure) = UNCLEAR. Because this is the paper’s primary outcome, Statistically analysed is also set to UNCLEAR for this row (the underlying ANOVA significance, p<=.01, is not itself in question - only this specific magnitude). The per-plant fresh weight (23.9+/-41.2 g/plant) is uncontested and not affected - recorded normally in Plant fresh weight. | Water volume in the system: same tank type as T1 (floating raft DWC/A-frame/tray share tanks) - UNCLEAR, see T1 remarks for the 378.5 L vs 550 L conflict. | Average room Temperature: p.85 states the same shared-greenhouse daily integral used for the tray systems (‘21.7+/-0.4C, set-point 21.5C,’ shared with T3/T4) and then adds ‘although daily temperatures were 23.5+/-0.9C.’ Read as a genuine A-frame-specific microclimate note (structure sits closer to glazing) rather than a duplicate/contradictory claim - signalled by the explicit ‘although’ and because this second figure is NOT given in the otherwise-identical tray ebb&flow passage. NOT flagged as a contradiction (setpoint/general-value + specific-microclimate-value carve-out). Recorded 23.5+/-0.9C (the more specific, A-frame value) in Average room Temperature; the shared 21.7+/-0.4C noted here as the general greenhouse value common to T2/T3/T4. | Total Feed: same 2,800(summer)/4,200(winter) mg/tank basis as T1 (this treatment’s per-fish recomputation at 30 koi/tank: 2,800/30=93.3, 4,200/30=140 - matches the paper’s own perch/koi/tilapia-at-30-fish/tank figures); see T1 remarks for the Methods-vs-Results MATERIAL resolution. | Koi fed same feed as tilapia (Skretting): starter crumble #2 (52% min protein, 16% min oil, 9% max moisture, 3% max fiber, 12% max ash, 18.2% digestible energy) then 1.5/2.5/3.5mm pellets (35% min protein, 7% min oil, 9% max moisture, 5% max fiber, 9% max ash, 13% digestible energy) as fish grew; rate ~50mg/tank, varied by weekly size/density. | NR fields, unit conversions, NO COLUMN items: as T1 (pooled Aq pH 7.20/ammonia 0.40/nitrate 70.77 taken from Table 2’s ‘A-frame’ row specifically, distinct from the shared ‘Ebb and flow’ row used for T3/T4).
abbeyBasilOcimumBasilicum2022-T3
Fish
| Field | Value |
|---|---|
| Fish | Goldfish (Carassius auratus) |
| Protein | NR (Pentair tropical flake food ztf25 data sheet gives fat/fiber/moisture/ash but not crude protein %) |
| Feed routine | Twice daily, 08:00-09:00 and 15:00-16:00 (fish feeding/monitoring); water quality checked in 50% of tanks each day, alternating |
| Feed regime | Pentair Aquatic Ecosystems tropical flake food (#ZTF25): 15% min fat, 2% max fiber, 12% max moisture, 8% max ash (protein % not stated on data sheet); ~50 mg/tank/day maintenance diet |
| Total Feed (kg) | ~0.0028 kg/tank (summer, ~4-wk cycle); ~0.0042 kg/tank (winter, ~6-wk cycle) - per production cycle, NOT a full-trial cumulative total |
Water
| Field | Value |
|---|---|
| Water recycle | Y |
| Water volume in the system | UNCLEAR (378.5 L vs 550 L, same tank type as T1/T2/T4; see remarks) |
| Aq pH | 7.19 |
| pHOptimal | 7 |
| Water temperature | 22.36-24.24 (seasonal means only, no pooled trial mean reported) |
| TAN / NH4-N | 0.48 |
| NO2-N | 2.54-7.25 (seasonal means only, no pooled trial mean reported) |
| NO3-N | 70.77 |
Plant
| Field | Value |
|---|---|
| Plant | Basil (Ocimum basilicum), cvs. ‘Elenora’/‘Genovese’/‘Nufar’ (pooled, cultivar ns) |
| Details | Co-grown with day-neutral strawberries (‘Albion’/‘Portola’/‘Evie 2’) and lettuce (‘Rex’/‘Skyphos’/‘Salanova Series Mix’) in the same tanks/pots (reported in companion papers, not extracted here); single-cut harvest at marketable maturity, cut at top of soilless media/rockwool plug; >=6 harvests/treatment, ~4-wk (summer) to ~6-wk (winter) cycles; fresh weight then oven-dried 77C for 7 days for dry weight; transplant seedling FW (66.59+/-24.81 mg pooled, ns among cultivars) ignored as negligible |
| Plant Category | Herbs (culinary herb, Lamiaceae) |
| Days Plant after transplant | ~28 (summer) to ~42 (winter) days per harvest cycle (>=6 harvests/treatment; derived from stated ~4-wk/~6-wk harvest intervals) |
| Plants/m2 | 8 |
| Plant fresh weight | 59.6 +/- 45.2 g/plant |
| Plant dry matter | 7.7 +/- 5.2 g/plant |
System & Setup
| Field | Value |
|---|---|
| System type | Tray ebb and flow (greenhouse) |
| Media Details | Rockwool plugs in web trays (50s/72s plug trays) on shelving racks, 2 shelves/rack, 2 tubs/shelf (123x94x18cm, 6 web trays/tub); 1 aluminum tank (same spec as floating raft DWC); 1 rectangular biofilter tub/tank (123x186x18cm), 3-4cm lava rock, ammonium carbonate inoculated; upper shelf natural+HPS-HID light, lower shelf supplemental LED (GrowPan or Green Power LED) |
| Biological system already in use | Y (1 rectangular biofilter tub/tank, 3-4cm lava rock, ammonium carbonate inoculation to ~1 mg/L ammonia with initial goldfish start) |
| Climate control | Y (Same greenhouse computerized climate control as floating raft DWC/A-frame (Argus Controller, generator backup, HPS-HID supplemental lighting)) |
| Artificial Lighting | Y (Upper shelf: natural light + 400W HPS-HID supplemental; lower shelf: supplemental LED (Sunshine Systems GrowPan 450-470/630nm 300W, or Green Power LED 450-470/660nm 300W); 16hr photoperiod) |
| Equipment | 1 rectangular biofilter tub/tank (3-4cm lava rock); shelving rack system (2 shelves, 2 tubs/shelf, 6 web trays/tub); pump model not specified for this system |
| Control Parameters | pH 6.5-7.5 (optimum 7); DO >6 ppm target; ammonia/nitrite/nitrate <0.75 ppm each (feeding suspended if exceeded); temperature setpoint 21.5C |
| Combination | Goldfish and basil (co-grown with day-neutral strawberries and lettuce, reported in companion papers); tray ebb and flow vs. soilless control |
Site
| Field | Value |
|---|---|
| Region | North America |
| Country | USA |
| Lat | 44.9883 |
| Long | -93.1810 |
| Average room Temperature | 21.7 +/- 0.4C (setpoint 21.5C; shared tray ebb&flow greenhouse value) |
Results & Statistics
| Field | Value |
|---|---|
| Measured Unit | g fw/m2; g fw/plant (fresh weight); g dw/m2; g dw/plant (dry weight) |
| Statistic Details | GLIMMIX (lognormal distribution, lowest AIC/BIC), split-plot design; fish type = treatment, cultivar = variable; koi treatments (A-frame, warehouse) analyzed separately from non-koi; Tukey HSD for water-quality mean separations (Table 2); ANOVA for treatment comparisons (Table 3); SAS v.25 |
| Statistically analysed | Y |
| Replicates (n) | 1 |
| AP | 447.0 |
| HYD | 1390.0 |
Experimental Remarks: TRIAL DEFINITION: T3 = Tray ebb & flow (greenhouse) with goldfish, 1 tank (2-rack system; goldfish remained in one tank while the other was later switched to tilapia, recorded as T4). Paired control = shared soilless-medium control (see T1). | Water volume in the system: same shared tank type as T1/T2/T4 - UNCLEAR, see T1 remarks (378.5 L vs 550 L). | Average room Temperature: 21.7+/-0.4C, setpoint 21.5C - the shared tray ebb&flow greenhouse value (same greenhouse as T4 and, for the general/non-microclimate value, T2’s A-frame); no A-frame-style ‘although’ caveat appears in this passage, so only the single value is recorded (not flagged, per the setpoint+measured-mean carve-out). | Total Feed: 2,800(summer)/4,200(winter) mg/tank, same basis as T1/T2 (goldfish at 90/tank recomputes to 31.1~‘31’/46.7~‘46.6’ mg/fish, matching the paper’s own figures); see T1 for the Methods-vs-Results MATERIAL resolution. | Feed: Pentair Aquatic Ecosystems tropical flake food ztf25, ‘15% min fat, 2% max fiber, 12% max moisture, 8% max ash’ - brand data sheet does NOT state a crude-protein percentage, so Protein = NR for this trial specifically (all other fish types in this paper have a stated protein %). Rate ~50mg/tank/day. | MINOR dry-weight rounding: Table 4 gives 7.7+/-5.2 g/plant; Results text (p.89) restates ‘7.68’ - simple rounding, Table 4 value retained (recorded in Plant dry matter). | NR fields/unit conversions/NO COLUMN: as T1.
abbeyBasilOcimumBasilicum2022-T4
Fish
| Field | Value |
|---|---|
| Fish | Tilapia (Oreochromis spp.) |
| Protein | 35 (grower pellets, min protein); 52 (starter crumble, min protein) - see Feed regime |
| Feed routine | Twice daily, 08:00-09:00 and 15:00-16:00 (fish feeding/monitoring); water quality checked in 50% of tanks each day, alternating |
| Feed regime | Skretting: starter crumble #2 (52% min protein, 16% min oil, 9% max moisture, 3% max fiber, 12% max ash, 18.2% digestible energy) then 1.5/2.5/3.5mm pellets (35% min protein, 7% min oil, 9% max moisture, 5% max fiber, 9% max ash, 13% digestible energy) as fish grew; ~50 mg/tank/day maintenance diet |
| Total Feed (kg) | ~0.0028 kg/tank (summer, ~4-wk cycle); ~0.0042 kg/tank (winter, ~6-wk cycle) - per production cycle, NOT a full-trial cumulative total |
Water
| Field | Value |
|---|---|
| Water recycle | Y |
| Water volume in the system | UNCLEAR (378.5 L vs 550 L, same tank type as T1-T3; see remarks) |
| Aq pH | 7.19 |
| pHOptimal | 7 |
| Water temperature | 22.36-24.24 (seasonal means only, no pooled trial mean reported; same ‘ebb and flow’ Table 2 row as T3) |
| TAN / NH4-N | 0.48 |
| NO2-N | 2.54-7.25 (seasonal means only, no pooled trial mean reported) |
| NO3-N | 70.77 |
Plant
| Field | Value |
|---|---|
| Plant | Basil (Ocimum basilicum), cvs. ‘Elenora’/‘Genovese’/‘Nufar’ (pooled, cultivar ns) |
| Details | Co-grown with day-neutral strawberries (‘Albion’/‘Portola’/‘Evie 2’) and lettuce (‘Rex’/‘Skyphos’/‘Salanova Series Mix’) in the same tanks/pots (reported in companion papers, not extracted here); single-cut harvest at marketable maturity, cut at top of soilless media/rockwool plug; >=6 harvests/treatment, ~4-wk (summer) to ~6-wk (winter) cycles; fresh weight then oven-dried 77C for 7 days for dry weight; transplant seedling FW (66.59+/-24.81 mg pooled, ns among cultivars) ignored as negligible |
| Plant Category | Herbs (culinary herb, Lamiaceae) |
| Days Plant after transplant | ~28 (summer) to ~42 (winter) days per harvest cycle (>=6 harvests/treatment; derived from stated ~4-wk/~6-wk harvest intervals) |
| Plants/m2 | 8 |
| Plant fresh weight | 52.4 +/- 63.8 g/plant |
| Plant dry matter | 5.6 +/- 7.3 g/plant |
System & Setup
| Field | Value |
|---|---|
| System type | Tray ebb and flow (greenhouse) |
| Media Details | Rockwool plugs in web trays on shelving racks (same rack system design/greenhouse as T3, separate tank switched from goldfish to tilapia prior to basil planting); 1 aluminum tank; 1 rectangular biofilter tub/tank, 3-4cm lava rock |
| Biological system already in use | Y (1 rectangular biofilter tub/tank, 3-4cm lava rock, ammonium carbonate inoculation (tank started as goldfish for cycling before switch to tilapia)) |
| Climate control | Y (Same greenhouse computerized climate control as T1-T3 (Argus Controller, generator backup, HPS-HID supplemental lighting)) |
| Artificial Lighting | Y (Upper shelf: natural light + 400W HPS-HID supplemental; lower shelf: supplemental LED (GrowPan or Green Power LED); 16hr photoperiod) |
| Equipment | 1 rectangular biofilter tub/tank (3-4cm lava rock); shelving rack system; pump model not specified for this system |
| Control Parameters | pH 6.5-7.5 (optimum 7); DO >6 ppm target; ammonia/nitrite/nitrate <0.75 ppm each (feeding suspended if exceeded); temperature setpoint 21.5C |
| Combination | Tilapia and basil (co-grown with day-neutral strawberries and lettuce, reported in companion papers); tray ebb and flow vs. soilless control |
Site
| Field | Value |
|---|---|
| Region | North America |
| Country | USA |
| Lat | 44.9883 |
| Long | -93.1810 |
| Average room Temperature | 21.7 +/- 0.4C (setpoint 21.5C; shared tray ebb&flow greenhouse value) |
Results & Statistics
| Field | Value |
|---|---|
| Measured Unit | g fw/m2; g fw/plant (fresh weight); g dw/m2; g dw/plant (dry weight) |
| Statistic Details | GLIMMIX (lognormal distribution, lowest AIC/BIC), split-plot design; fish type = treatment, cultivar = variable; koi treatments (A-frame, warehouse) analyzed separately from non-koi; Tukey HSD for water-quality mean separations (Table 2); ANOVA for treatment comparisons (Table 3); SAS v.25 |
| Statistically analysed | Y |
| Replicates (n) | 1 |
| AP | 393.0 |
| HYD | 1390.0 |
Experimental Remarks: TRIAL DEFINITION: T4 = Tray ebb & flow (greenhouse) with tilapia, 1 tank (2-rack system; this tank was switched from goldfish to tilapia prior to basil planting; the other tank remained goldfish, recorded as T3). Paired control = shared soilless-medium control (see T1). | Water volume in the system: same shared tank type as T1/T2/T3 - UNCLEAR, see T1 remarks (378.5 L vs 550 L). | Average room Temperature: 21.7+/-0.4C, setpoint 21.5C - shared tray ebb&flow greenhouse value (same greenhouse as T3); not flagged (setpoint+measured-mean carve-out). | Total Feed: 2,800(summer)/4,200(winter) mg/tank, same basis as T1-T3 (tilapia at 30/tank recomputes identically to the perch/koi-at-30/tank figures: 93.3/140 mg/fish); see T1 for the Methods-vs-Results MATERIAL resolution. | Feed: as fingerlings, Skretting starter crumble #2 (52% min protein, 16% min oil, 9% max moisture, 3% max fiber, 12% max ash, 18.2% digestible energy), then 1.5/2.5/3.5mm Skretting pellets (35% min protein, 7% min oil, 9% max moisture, 5% max fiber, 9% max ash, 13% digestible energy) as fish grew; rate ~50mg/tank/day. | NR fields/unit conversions/NO COLUMN: as T1.
abbeyBasilOcimumBasilicum2022-T5
Fish
| Field | Value |
|---|---|
| Fish | Koi (Cyprinus carpio) |
| Protein | 35 (grower pellets, min protein); 52 (starter crumble, min protein) - see Feed regime |
| Feed routine | Twice daily, 08:00-09:00 and 15:00-16:00 (fish feeding/monitoring); water quality checked in 50% of tanks each day, alternating |
| Feed regime | Same feed as koi/tilapia elsewhere (Skretting): starter crumble #2 (52% min protein) then 1.5/2.5/3.5mm pellets (35% min protein); 50 mg/tank/day stated directly for this treatment |
| Total Feed (kg) | ~0.0028 kg/tank (summer, ~4-wk cycle); ~0.0042 kg/tank (winter, ~6-wk cycle) - per production cycle, NOT a full-trial cumulative total |
Water
| Field | Value |
|---|---|
| Water recycle | Y |
| Water volume in the system | 416 L (Nelson & Pade F5 fish tank; distinct, singly-stated figure, separate from settlement tanks which have no stated volume - NR) |
| Aq pH | 7.56 |
| pHOptimal | 7 |
| Water temperature | 17.11-25.11 (seasonal means only, no pooled trial mean reported; ONLY treatment where season significantly affects water temp, ‘most likely because the warehouse is underground in a modified cooler’) |
| TAN / NH4-N | 0.42 |
| NO2-N | 0.11-3.17 (seasonal means only, no pooled trial mean reported) |
| NO3-N | 72.94 |
Plant
| Field | Value |
|---|---|
| Plant | Basil (Ocimum basilicum), cvs. ‘Elenora’/‘Genovese’/‘Nufar’ (pooled, cultivar ns) |
| Details | Co-grown with day-neutral strawberries (‘Albion’/‘Portola’/‘Evie 2’) and lettuce (‘Rex’/‘Skyphos’/‘Salanova Series Mix’) in the same tanks/pots (reported in companion papers, not extracted here); single-cut harvest at marketable maturity, cut at top of soilless media/rockwool plug; >=6 harvests/treatment, ~4-wk (summer) to ~6-wk (winter) cycles; fresh weight then oven-dried 77C for 7 days for dry weight; transplant seedling FW (66.59+/-24.81 mg pooled, ns among cultivars) ignored as negligible |
| Plant Category | Herbs (culinary herb, Lamiaceae) |
| Days Plant after transplant | ~28 (summer) to ~42 (winter) days per harvest cycle (>=6 harvests/treatment; derived from stated ~4-wk/~6-wk harvest intervals) |
| Plants/m2 | 15 |
| Plant fresh weight | 9.6 +/- 12.1 g/plant |
| Plant dry matter | 1.5 +/- 1.8 g/plant |
System & Setup
| Field | Value |
|---|---|
| System type | Warehouse floating raft, Nelson & Pade F5 (Fantastically Fun Fresh Food Factory), retrofitted walk-in cooler (7.19 x 4.87 x 2.74 m) |
| Media Details | Rockwool plugs in 5.8cm net pots (15 pots/raft, 2 rafts/tub, 0.9x1.5m tubs); 1 fish tank (416 L) with separate biofilter and 2 settlement tanks |
| Biological system already in use | Y (F5 system biofilter, separate from fish tank and 2 settlement tanks (Nelson & Pade commercial design); no further specification given) |
| Climate control | Y (Supplemental cooling maintained 20-21C day/night in retrofitted walk-in cooler; electric generator backup; galvanized interior walls) |
| Artificial Lighting | Y (LED triple-band light bar (Agrivolution), single-chip blue/green/red diodes, PAR 400-700nm, telescoped vertically as plants grew; 16hr photoperiod, min. 150 umol/m2/s; chosen over HPS-HID specifically to avoid overheating the cooler beyond its cooling capacity) |
| Equipment | Nelson & Pade F5 commercial system (fish tank, biofilter, 2 settlement tanks, grow beds); Agrivolution LED triple-band light bar |
| Control Parameters | pH 6.5-7.5 (optimum 7); DO >6 ppm target; ammonia/nitrite/nitrate <0.75 ppm each (feeding suspended if exceeded); target growing temperature 20-21C day/night |
| Combination | Koi and basil (co-grown with day-neutral strawberries and lettuce, reported in companion papers); warehouse floating raft (F5) vs. soilless control |
Site
| Field | Value |
|---|---|
| Region | North America |
| Country | USA |
| Lat | 44.9883 |
| Long | -93.1810 |
| Average room Temperature | 20-21C (constant day/night, maintained by supplemental cooling; distinct, uncontested figure for this treatment) |
Results & Statistics
| Field | Value |
|---|---|
| Measured Unit | g fw/m2; g fw/plant (fresh weight); g dw/m2; g dw/plant (dry weight) |
| Statistic Details | GLIMMIX (lognormal distribution, lowest AIC/BIC), split-plot design; fish type = treatment, cultivar = variable; koi treatments (A-frame, warehouse) analyzed separately from non-koi; Tukey HSD for water-quality mean separations (Table 2); ANOVA for treatment comparisons (Table 3); SAS v.25 |
| Statistically analysed | Y |
| Replicates (n) | 1 |
| AP | 144.0 |
| HYD | 1390.0 |
Experimental Remarks: TRIAL DEFINITION: T5 = Warehouse floating raft, Nelson & Pade F5 (‘Fantastically Fun Fresh Food Factory’) commercial system, in a retrofitted walk-in cooler; koi, 1 tank (32 koi/tank). Paired control = shared soilless-medium control (see T1). | Water volume in the system: 416 L - this treatment’s F5 fish tank is distinct from, and singly/uncontested-ly stated (unlike the shared floating-raft/A-frame/tray tank type in T1-T4); separate from its settlement tanks, which have no stated volume (NR). | Average room Temperature: 20-21C, constant day/night, maintained by supplemental cooling - distinct, singly-stated figure for this treatment (uncontested; not part of the T1 floating-raft-DWC BLOCK). Paper explicitly identifies this as the coolest treatment and flags it as a plausible contributing factor (alongside koi species) to this treatment’s markedly lower yield. | Total Feed: 2,800(summer)/4,200(winter) mg/tank, same basis as T1-T4 (32 koi/tank recomputes to 87.5/131.25~‘131.2’ mg/fish, matching the paper’s own figure); see T1 for the Methods-vs-Results MATERIAL resolution. | Feed: same as koi/tilapia elsewhere (Skretting, 52% min protein starter then 35% min protein grower pellets), rate ‘50 mg/tank/day’ stated directly for this treatment (p.85). | Water quality: this is the ONLY treatment where season x treatment water-temperature difference is itself significant (summer 25.11 vs winter 17.11C, Table 2) - ‘most likely because of the warehouse being underground in a modified cooler’ (p.86); recorded as a range per the ‘seasonal means only’ convention, consistent with the other treatments. | Lighting: LED (Agrivolution triple-band bar, blue/green/red single-chip diodes, PAR 400-700nm, telescoping height) used instead of HPS-HID specifically because HPS-HID would have overheated the warehouse beyond the cooling system’s capacity - this is a system-specific deviation from T1-T4’s HPS-HID lighting. | NR fields/unit conversions/NO COLUMN: as T1 (Alkalinity pooled 72.94 ppm GH/KH, TDS 1.20 ppt summer/winter missing, Table 2).