Evaluating Performance Limitations in Aquaponic vs. Hydroponic: Dynamics of Nutrient Release by Fish and Accumulation Rate in Plants
Metadata
- Cite key: mehdiEvaluatingPerformanceLimitations2026
- Item type: Journal Article
- Authors: S.E.H. Mehdi, A. Sharma, S. Shahzad, W. Kang, S. Pandey, B.-J. Park, H.-S. Kim, S.-E. Oh
- Affiliation: Department of Food Biotechnology and Environmental Science, Kangwon National University, 192-1 Hyoja-dong, Chuncheon-si 24341, Gangwon-do, Republic of Korea
- Journal: Water 18 (2026) 742
- Date: 03/2026
- Date added: 2026-07-13
- DOI: 10.3390/w18060742
- Funding: Cooperative Research Program for Agriculture Science and Technology Development (Project No. RS-2023-00227531), Rural Development Administration, Republic of Korea; National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIT) (No. 2023R1A2C1004608)
- URL: https://doi.org/10.3390/w18060742
- PDF:
Mehdi et al. - 2026 - Evaluating Performance Limitations in Aquaponic vs. Hydroponic Dynamics of Nutrient Release by Fish.pdf
Opinion
A clean, well-instrumented small-scale comparison (daily water chemistry, ICP-OES tissue analysis, explicit batch tests for fish excretion and plant uptake rates) that is refreshingly honest about its own framing: the authors state outright that the HP arm is an “optimized benchmark,” not a matched control, so this is a diagnostic comparison rather than a factorial experiment. The batch-test-derived design parameters (NH4+-N release, per-plant nutrient uptake) are the paper’s most reusable contribution. The main weakness is internal reporting precision: several water-quality and plant means are restated with two different SE values between the first Results narrative (§3.3/3.2) and the summary Table 4 (means always match; only dispersion differs), which raises a small question mark over which SE is correct without threatening the paper’s substantive conclusions.
Abstract
Aquaponics (AP) is the combination of aquaculture and hydroponic systems, developed based on waste to wealth theory. This study compared the plant growth and overall productivity of an aquaponic system (AP) with a controlled hydroponic system (HP) to assess the AP system’s performance and identification of the performance-limiting factors. This comparative study spanned over a 35-day period, supported by batch tests for the nutrient accumulation rate in plants and the NH4+-N excretion rate by fish as a baseline for the system design. HP performed better in terms of plant growth, showing a mean plant fresh weight (g) of 165.6 ± 3.01 while AP showed 147.0 ± 4.6. Nutrient accumulation was better in HP for K and P; however, Ca2+, Mg2+, and Fe accumulation was higher in AP plants. The AP system supported a better fish growth of 31.95 ± 3.21% (FCR 1.29 ± 0.1, SGR 0.79 ± 0.06, and PER 2.24 ± 0.18) and a moderate plant biomass production. Further system design modifications and integrations are required to optimize the nutrient availability and sustainability of the AP systems.
Summary
The authors ran a 35-day, triplicate (n=3) side-by-side comparison of a koi-fish-fed aquaponic DWC lettuce system (20 kg/m3 stocking density, no macro/micronutrient fertilizer except a corrective iron dose in week 2) against an “optimized” hydroponic DWC benchmark fertilized with a commercial two-part nutrient solution. Ahead of the main trial they ran two preliminary batch tests — a 36-hour fish NH4+-N excretion test and a 6-day lettuce nutrient-uptake test — to generate baseline mass-balance design parameters, not to characterize the 35-day system itself. Daily water chemistry (pH, DO, EC, turbidity, NH4+-N, NO2-N, NO3-N, PO4-P) was tracked in situ, and plant tissue (Ca, Mg, K, P, Fe) was analyzed by ICP-OES at harvest alongside fresh weight, leaf count, root length and SPAD chlorophyll. HP plants grew larger (higher fresh weight, root length, P and K tissue content, driven by much higher dissolved NO3-N and EC), while AP plants had higher SPAD, Fe, Mg and Ca tissue content and higher dissolved oxygen and PO4-P — the AP Fe advantage being explicitly attributed to the corrective iron supplementation rather than an intrinsic AP benefit. Fish performance was strong (FCR 1.29, SGR 0.79%/day, 0% mortality) despite the lack of nutrient supplementation. The authors conclude that the dominant limitation on AP productivity was insufficient dissolved ionic strength and K/P availability from fish waste alone, and propose the excretion/uptake batch-test rates as reusable design parameters for future fish-to-plant ratio and remineralization decisions.
Experiment data
- Location: Kangwon National University, Chuncheon-si, Gangwon-do, Republic of Korea (lab-based, small-scale)
- Design: Single aquaponic treatment (triplicate, n=3 independent systems) vs. single hydroponic “optimized benchmark” (not a matched/nutrient-equivalent control); paired two-tailed t-tests, p<0.05. Two preliminary batch tests (36 h fish excretion; 6-day plant uptake) preceded the 35-day main trial.
- Replicates / n: 3 independent systems per treatment
- Duration: 35 days (main trial); 36 h (fish excretion batch test); 6 days (plant uptake batch test)
- Organisms: Koi (Cyprinus rubrofuscus var. koi) / Romaine lettuce (Lactuca sativa L. var. longifolia)
- Statistics: Paired two-tailed t-tests, p<0.05 threshold; Pearson correlation (NH4-N vs NO3-N). Software: [not reported]
- Plant fresh weight: AP 147.0 ± 4.6 g/plant vs HP 165.6 ± 3.05 g/plant (p<0.05)
- Feed Conversion Rate (FCR): 1.29 ± 0.1 (AP fish)
- Specific Growth Rate (SGR): 0.79 ± 0.06 %/day (AP fish)
Fish growth
This paper: Koi (20 fish, 20 kg/m3 stocking density) grew from a system biomass of 1.19 ± 0.005 kg to 1.57 ± 0.03 kg over 35 days (31.95 ± 3.21% increase; per-fish weights not given). FCR = 1.29 ± 0.1, SGR = 0.79 ± 0.06%/day, PER = 2.24 ± 0.18, mortality 0%. Both the SGR and FCR figures internally cross-check against the stated formulas and inputs (verified by recomputation, not entered as new cells).
Compared with:
- todo Sallenave 2016 — FCR 1.9–4.6 for koi at 1.4–2.8 kg/m3 stocking density, much worse than this paper’s 1.29 at 20 kg/m3. (p. 11)
- todo Hussain et al. 2015 — FCR 1.76 at 20.3 kg/m3 for tilapia in a lettuce-based AP system. (p. 11-12)
- todo Somerville et al. 2014 — SGR below 1%/day reported for koi-based aquaponics, consistent with this paper’s 0.79%/day. (p. 12)
Plant growth and nutrient accumulation
This paper: HP plants outperformed AP on fresh weight (165.6 ± 3.05 vs 147.0 ± 4.6 g/plant), root length (25.83 ± 0.50 vs 21.8 ± 0.44 cm), and tissue P/K; AP plants had higher SPAD (45.47 ± 0.81 vs 37.40 ± 0.90) and tissue Fe/Mg/Ca. Leaf count was numerically higher in HP (21.0 ± 0.58 vs 20.2 ± 0.48) but the paper’s own statistics (Table 4, same superscript letter; explicit statement p.11 “did not show significant differences, p>0.05”) say this difference is not significant, despite §3.2’s prose framing it as HP performing “better.” Unit productivity: AP 5.96 kg/m2/cycle (lettuce) and 6.17 kg/m3/cycle (fish); HP 6.71 kg/m2/cycle.
Interveinal chlorosis (Fe deficiency) appeared in AP plants in week 2; FeSO4·7H2O was added as a reactive corrective measure, not a planned nutrient input. The paper is explicit that the resulting AP Fe tissue advantage reflects successful correction of an induced deficiency, not an intrinsic AP nutrient benefit.
Compared with:
- todo Klinger and Naylor 2012 — reports Ca2+ and Mg2+ as major deficient nutrients in AP systems generally; this paper’s AP system had sufficient Ca/Mg, a contrast the authors note explicitly. (p. 13)
- todo Kasozi et al. 2019 — review of iron supplementation/management in aquaponics, cited as context for the Fe deficiency episode. (p. 12)
- todo Lages Barbosa et al. 2015 — cited for grow-bed configuration comparison (10 plants/fish design ratio). (p. 3)
Water quality dynamics
This paper: AP maintained lower EC (0.92 vs 2.55 mS/cm), lower dissolved N species (NH4+-N 0.83 vs 12.81 mg/L; NO3-N 81.12 vs 280.6 mg/L), similar pH (~6.76 both), higher DO (6.32 vs 5.24 mg/L), higher turbidity (5.94 vs 1.92 NTU) and higher PO4-P (19.2 vs 13.8 mg/L) than HP. AP showed a negative NH4-N/NO3-N correlation (r=-0.564, p=0.0095), consistent with active nitrification; HP showed a positive correlation (r=0.686, p=0.0009), which the authors interpret as parallel nutrient depletion rather than biological transformation. See Extraction notes for a reporting-precision issue affecting several of these means’ stated dispersion.
Compared with:
- todo Wongkiew et al. 2017 — nitrogen transformation review in aquaponic systems, cited for the nitrification-driven AP correlation interpretation. (p. 9)
- todo Groenveld et al. 2019 — fertigation/nitrogen-use-efficiency optimization in an integrated aquaculture-agriculture system, cited on pH management. (p. 12)
Linked claims
- HP plants outyield AP plants when HP is nutrient-optimized
- Fish waste alone under-supplies potassium and phosphorus for plant growth in aquaponics
- Iron deficiency is a recurring problem in fertilizer-free aquaponic systems
- Aquaponic systems show lower electrical conductivity than hydroponic systems
Citations to chase
- todo Sallenave (2016) — FCR range for koi at low stocking density, comparison point for this paper’s FCR
- todo Hussain et al. (2015) — FCR for tilapia at comparable 20.3 kg/m3 stocking density in lettuce-based AP
- todo Somerville et al. (2014) — SGR benchmark for koi-based aquaponics
- todo Klinger and Naylor (2012) — Ca2+/Mg2+ deficiency claim in AP systems, contradicted by this paper’s own data
- todo Kasozi et al. (2019) — review of iron supplementation/management in aquaponics
- todo Wongkiew et al. (2017) — nitrogen transformation review, aquaponic systems
- todo Groenveld et al. (2019) — fertigation/N-use-efficiency in integrated aquaculture-agriculture
Extraction notes
⚠️MATERIAL — NO2-N (AP) mean: §3.3 (p.7): “the concentrations in the AP system were 0.45 ± 0.03 mg/L for the 35-day study period.” Table 4 (p.9): “NO2−–N 0.49 ± 0.03 b” (AP). §3.4 Discussion (p.10-11): “AP’s values of 0.83 ± 0.07, 0.49 ± 0.03, and 81.12 ± 3.10 mg L−1” (NH4/NO2/NO3 in that order). §4 Discussion (p.13): “nitrite mean concentrations in the AP system were below 0.49 ± 0.03 mg/L.” Three independent restatements agree on 0.49; only the first Results mention (§3.3) says 0.45. Recorded 0.49 ± 0.03 (majority + reproduced in the paper’s own later synthesis). Affects: NO2-N cell only; does not change any other conclusion.
⚠️MATERIAL — systematic SE-magnitude mismatch, §3.3/§3.2 vs. Table 4/§3.4 (consolidated, one likely root cause, six parameters): For six water-quality/plant-growth parameters, the mean is byte-identical (or trivially rounded) between the first Results narrative and Table 4/the Discussion synthesis, but the stated dispersion (both places explicitly label it “SE,” Table 4 caption and §2.6 both say results are mean ± SE, n=3) differs by roughly the same factor (~1.5-2x) in every case, always with Table 4’s SE the larger one: DO AP 6.33±0.09 (§3.3, p.7) vs 6.32±0.19 (Table 4 p.9; §3.4 p.11, matches Table 4). DO HP 5.24±0.09 (§3.3) vs 5.24±0.14 (Table 4; §3.4). EC AP 0.92±0.01 mS/cm (§3.3) vs 0.92±0.03 (Table 4; §3.4). EC HP 2.55±0.05 (§3.3) vs 2.55±0.09 (Table 4; §3.4). Turbidity AP 5.94±0.34 NTU (§3.3) vs 5.94±0.54 (Table 4; §3.4). Turbidity HP 1.93±0.10 (§3.3, trivial mean rounding to 1.92 elsewhere) vs 1.92±0.18 (Table 4; §3.4). The consistent direction and rough scaling (close to sqrt(3), the SD/SE ratio at n=3) suggests one of the two sources conflated SD and SE despite both being labeled SE, but the paper never says so and neither figure is stated as wrong. Table 4’s values are used for all six cells because they are independently reproduced verbatim in the Discussion narrative (2 sources vs. 1). Section 3.3’s original figures are recorded here as the alternate. Scoring note: these six discrepancies are counted as ONE MATERIAL flag for the quality score below, since they share what looks like a single root cause rather than being six independent reporting errors — flagged explicitly so the user can recount them individually (which would push the paper to suspect) if they judge that more appropriate.
⚠️MINOR — SPAD (HP): §3.2/Figure 2 (p.6): “37.40 ± 0.96.” Table 4 (p.9) and §3.4 (p.11): “37.40 ± 0.90.” Same direction of discrepancy as above but smaller magnitude and does not fit the ~sqrt(3) pattern (ratio ~1.07), so treated as a separate, trivial rounding-level issue. Recorded 37.40 ± 0.90 (Table 4/majority). No cell-choice consequence beyond the SD value itself.
⚠️MINOR — Plant fresh weight (HP): Abstract: “165.6 ± 3.01.” §3.2 (p.6) and Table 4 (p.9): “165.6 ± 3.05.” Recorded 165.6 ± 3.05 (majority, more precise sourcing). Trivial, no interpretive impact.
⚠️MINOR — Total Fish Weight Gain: Table 3 (p.6): “0.38 ± 0.034 kg.” Running text (p.6): “weight gain of 0.38 ± 0.03 kg.” Recorded 0.38 ± 0.034 (Table 3, more decimal precision). No impact.
⚠️MINOR — Leaf count narrative framing: §3.2 (p.6) states leaf number “showed better results for HP” (21 ± 0.58 vs 20.2 ± 0.48), but Table 4 marks both AP and HP with the same superscript letter “a” and §3.4 (p.11) explicitly states the leaf-count difference is not significant (p>0.05). Not a numeric contradiction — the two raw values are stated consistently everywhere — just an interpretive overstatement in §3.2’s prose. No cell affected; noted for anyone citing “HP has more leaves” from this paper.
⚠️CHECK — plant nutrient accumulation (Table 4), unit and tissue basis both unclear, p.9. Table 4’s “Plant Nutrient Accumulation” block (P, K, Fe, Mg2+, Ca2+) gives no unit in its header or caption, unlike Table 2’s batch-test rates which are explicitly labelled mg/L·d⁻¹·plant⁻¹. Methods §2.3 (p.5) states dried root and leaf samples were separately collected, oven-dried, then “crushed to a 40 mesh size for further macro- and micronutrient…analysis” and digested for ICP-OES, but does not say whether root and leaf were combined into one composite sample or whether Table 4 reports only one tissue fraction — Table 4 gives a single value per nutrient per system either way. Recorded in plant_measurements.csv with Unit UNCLEAR (mg/kg DW would be the standard ICP-OES convention but is not confirmed in text) rather than assumed. Added to REVIEW.md by the batch merge step.
[not reported]:
- Fish Category, Water classification, pHOptimal (system-specific target — a general 6.5-8.5 nitrification/lettuce range is cited from the literature, ref. [17], not measured by this study)
- Feed protein/N/P/K composition (%), % of body weight ration basis is stated (1%/day of initial weight, not adjusted for growth) but no compositional breakdown of the commercial feed
- Fish size initial/final (per fish) — only system-level total biomass given (1.19 ± 0.005 kg / 1.57 ± 0.03 kg for 20 fish); NOT divided per-fish (no derivation)
- Fish weight gain (per fish) — same reason; system-level “Fish biomass created” is recorded instead
- Fish survival rate — only “Fish Mortality (%) = 0” (Table 3) is stated; the complementary 100% survival figure is not itself written, so left NR rather than computed
- Total Feed (kg) — not stated as an absolute total; inputs exist to estimate it (FCR 1.29 x weight gain 0.38 kg ~ 0.49 kg) but this is not derived per the no-derivation rule
- Tissue nitrate AP/HYD — this paper does not measure plant tissue nitrate; it measures dissolved water NO3-N and plant tissue Ca/Mg/K/P/Fe instead
- Lat/Long — university address given, but no coordinates stated anywhere in the paper (not looked up externally per the prime directive)
- FUE AP/HYD, WUE — not calculated
- Statistics software — never named (only “paired two-tailed t-tests, p<0.05” and Pearson r/p are given)
- Plant height, Plant dry matter — height was not among the measured traits (only mass, leaf count, root length, chlorophyll); dry mass was measured per Methods (§2.3) for tissue-nutrient analysis but no dry-matter % or dry-weight value is reported in Results/Table 4
NO COLUMN (no dedicated trials.csv field):
- HP-side water quality, since the schema’s water-quality columns (Aq pH, DO, EC, Water temperature, TAN/NH4-N, NO2-N, NO3-N) are aquaponic-loop fields with no paired HYD column: NH4+-N (HP) 12.81 ± 1.57 mg/L; NO2-N (HP) 1.37 ± 0.30 mg/L; NO3-N (HP) 280.6 ± 7.9 mg/L; EC (HP) 2.55 ± 0.09 mS/cm; DO (HP) 5.24 ± 0.14 mg/L; pH (HP) 6.75 ± 0.06
- Turbidity (no dedicated column): AP 5.94 ± 0.54 NTU, HP 1.92 ± 0.18 NTU (Table 4)
- PO4-P (no dedicated column): AP 19.2 ± 1.7 mg/L, HP 13.8 ± 1.2 mg/L (Table 4)
- Root length (no dedicated column): AP 21.8 ± 0.44 cm, HP 25.83 ± 0.50 cm (Table 4)
- HP-side plant growth for singular (non-AP/HYD-paired) columns: Leaf count HP 21.0 ± 0.58; Plant fresh weight HP 165.6 ± 3.05 g/plant (also recorded via the AP/HYD unit-productivity columns, see TRIAL DEFINITION in trials.csv); Chlorophyll SPAD HP 37.40 ± 0.90 (also captured in plant_measurements.csv)
- Unit productivity: AP 5.96 kg/m2/cycle (lettuce) and 6.17 kg/m3/cycle (fish); HP 6.71 kg/m2/cycle — recorded in trials.csv’s generic AP/HYD columns for the lettuce area-yield figures (fish volumetric UP has no home and stays here)
- PER (protein efficiency ratio) = 2.24 ± 0.18 — no dedicated column
- % biomass increase (system-level, distinct from daily feed ration %): 31.95 ± 3.21% over 35 days
- HRT (hydraulic retention time, fish tank) = 3 h — distinct metric from the “Water recycle” flow-rate (L/min) column, which the paper does not state
- Daily water top-up (distinct from a %-of-volume exchange rate): ~2 L/day (AP), ~1 L/day (HP), to offset evaporation/transpiration/sampling losses
- NH4-N vs NO3-N correlation: AP r=-0.564 (p=0.0095), HP r=0.686 (p=0.0009) (Figure 7)
- Plant nutrient accumulation Table 4 values (P, K, Fe, Mg2+, Ca2+, mg/kg) are captured in
plant_measurements.csvinstead, per SCHEMA.md (plant analytes belong there, not trials.csv)
Judgment calls / tags: Meta/Fish/Koi and Meta/Region/East-Asia reused as-is from existing vault entries for Korean koi-aquaponics papers (kimComparisonWaterQuality2023, leeComparativeStudyGrowth2019). Meta/Plant/Lettuce reused (romaine is L. sativa, matching the vault’s existing generic Lettuce leaf). No new tags introduced. Plant Category left NR — the paper never applies a categorical label like “leafy vegetable” to the lettuce itself (unlike some other vault papers that do).
Source: Mehdi et al. - 2026 - Evaluating Performance Limitations in Aquaponic vs. Hydroponic Dynamics of Nutrient Release by Fish.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
mehdiEvaluatingPerformanceLimitations2026-T1
Fish
| Field | Value |
|---|---|
| Fish | Koi (Cyprinus rubrofuscus var. koi) |
| Initial Stock density | 20 |
| FCR | 1.29 +/- 0.1 |
| SGR | 0.79 +/- 0.06 |
| % of body weight | 1 (%/day of initial body weight, not adjusted for growth over the trial, Methods 2.2, p.3) |
| Feed routine | Fed twice daily at 09:00 and 18:00, commercial feed (Methods 2.2, p.3) |
| Feed regime | Fixed ration, 1%/day of initial body weight (not ad libitum, not adjusted for growth) |
| Fish biomass created (kg) | 0.38 +/- 0.034 |
| Fish trial duration (days) | 35 |
Water
| Field | Value |
|---|---|
| Water volume in the system | 60 (fish tank only, L; separate 15 L up-flow mechanical filter and 30 L biofilter per system, not summed by the paper into one total, Methods 2.2, p.3) |
| Water type | Dechlorinated water (granular activated carbon), Methods 2.2, p.3 |
| Aq pH | 6.76 +/- 0.06 a (ns vs HP 6.75 +/- 0.06 a, Table 4) |
| Dissolved Oxigen | 6.32 +/- 0.19 a (Table 4; 6.33 +/- 0.09 in SS3.3 — see WARN-MATERIAL) |
| EC | 0.92 +/- 0.03 b mS/cm (Table 4; 0.92 +/- 0.01 in SS3.3 — see WARN-MATERIAL) |
| Water temperature | 23-25 (maintained range, stated jointly with room temperature, not a separately measured trial mean; Methods 2.2, p.3) |
| TAN / NH4-N | 0.83 +/- 0.07 b |
| NO2-N | 0.49 +/- 0.03 b (see WARN-MATERIAL) |
| NO3-N | 81.12 +/- 3.10 b |
Plant
| Field | Value |
|---|---|
| Plant | Romaine lettuce (Lactuca sativa var. longifolia) |
| Details | 150 plants/system on styrofoam DWC trays; 3-week-old saplings at transplant; 10 plants/fish ratio derived from preliminary batch tests; 35-day cycle, single harvest at D35 |
| Days Plant after transplant | 35 |
| Plants/m2 | 40.5 |
| SPAD (aquaponics) | 45.47 +/- 0.81 a |
| Leaf count | 20.2 +/- 0.48 a (ns vs HP 21.0 +/- 0.58 a, Table 4; see WARN-MINOR) |
| Plant fresh weight | 147.0 +/- 4.6 b |
System & Setup
| Field | Value |
|---|---|
| System type | Deep-water culture (DWC), floating styrofoam trays; recirculating fish tank + mechanical filter + biofilter loop, HRT 3 h (fish tank), Methods 2.2, p.3 |
| Media Details | Fish tank 60 L; up-flow mechanical filter 15 L (sponge media); aerobic/biofilter tank 30 L (foam sponge, 10-20 ppi pore size); DWC grow bed area ~3.7 m2 total; styrofoam floating trays, foam plugs (Methods 2.2, p.3) |
| Biological system already in use | N (System was newly set up and matured for 1 month pre-experiment to establish biofilter microbial communities (Methods 2.2, p.3); not a reuse of a system from earlier published work.) |
| Air supplement | Y (Air pumps installed in fish tank and plant beds to maintain DO above 5 mg/L (AP, Methods 2.2, p.3); air stones + air pump also aerate the HP DWC unit.) |
| Iron supplemented | Y (Corrective FeSO4.7H2O supplementation administered in week 2 of the 35-day trial after interveinal chlorosis (Fe deficiency) appeared in AP plants; a reactive, one-time remedial measure, not a planned/routine input (Methods 2.2, p.3-4).) |
| Remineralization | N (AP system explicitly did not incorporate macro- or micronutrient fertilizers; the sole exception is the corrective iron dose recorded separately under Iron supplemented (Methods 2.2, p.3-4).) |
| pH Buffers | Y (KOH used to adjust AP pH (to avoid Na buildup, considered risky for fish); NaOH used for HP pH correction in limited amounts (Methods 2.2, p.3-4).) |
| Climate control | Y (Room and water temperature maintained at 23-25C throughout (Methods 2.2, p.3); specific climate-control equipment/mechanism not described.) |
| Artificial Lighting | Y (12 h photoperiod (07:00-19:00) combining natural sunlight and LED lights, applied identically to AP and HP (Methods 2.2, p.3).) |
| Nutrient supplemented | N (AP system received no macro/micronutrient fertilizer, relying on fish-waste-derived nutrients alone except the corrective iron dose (see Iron supplemented). HP (the benchmark, not the AP treatment this field describes) received a commercial two-part fertilizer, Daeyu Mulpure No.1 A+B, diluted 400x (Table 1), Methods 2.2, p.3-4.) |
| Equipment | Orion 920A ion-selective electrode, Thermo Scientific (NH4+-N); Metrohm-861 Advanced Compact Ion Chromatography (NO3-N, NO2-N, PO4-P); HI-6522 HANNA Instruments (EC); HI9147 HANNA Instruments (DO, temperature); S20 SevenEasy, Mettler Toledo (pH); HI98703 HANNA Instruments (turbidity); ICP-OES iCAP 6300, Thermo-Scientific (plant tissue Ca/K/Mg/Fe/P); peristaltic pumps (AP water transfer); submersible pump (HP circulation); air pumps + air stones (aeration, both systems) |
| Control Parameters | DO maintained above 5 mg/L in fish tank and plant beds (both systems); pH corrected with KOH (AP, to avoid Na buildup) or NaOH (HP); HRT 3 h for AP fish tank; feed 1%/day of initial body weight, twice daily (9:00, 18:00); 12 h photoperiod (7:00-19:00), natural sunlight + LED; room and water temperature maintained at 23-25C; daily top-up ~2 L (AP) / ~1 L (HP) to offset evaporation/transpiration/sampling losses |
| Combination | Koi (Cyprinus rubrofuscus var. koi) and romaine lettuce (Lactuca sativa var. longifolia) in a small-scale lab triplicate DWC aquaponic system, compared against a single ‘optimized’ hydroponic DWC benchmark (not a matched/nutrient-equivalent control); two preliminary batch tests (36 h fish NH4+-N excretion; 6-day plant nutrient uptake) preceded the 35-day main trial to generate baseline design parameters |
Site
| Field | Value |
|---|---|
| Region | East Asia |
| Country | South Korea |
| Average room Temperature | 23-25 (maintained range, stated jointly with water temperature, not a separately measured trial mean; Methods 2.2, p.3) |
Results & Statistics
| Field | Value |
|---|---|
| Measured Unit | g/plant (fresh weight); cm (root length); count (leaf number); SPAD units (chlorophyll); mg/kg DW, unit not explicit (plant tissue Ca/K/Mg/Fe/P — see plant.csv); mg/L (dissolved water parameters); mS/cm (EC); NTU (turbidity) |
| Statistic Details | Paired two-tailed t-tests, p<0.05; Pearson correlation (NH4-N vs NO3-N); statistics software not named (Methods 2.6) |
| Statistically analysed | Y |
| Replicates (n) | 3 |
| AP | 5.96 (kg/m2/cycle, lettuce) |
| HYD | 6.71 (kg/m2/cycle, lettuce) |
Experimental Remarks: TRIAL DEFINITION: T1 = single aquaponic (AP) treatment, triplicate (n=3) independent 60 L fish tank + 15 L mechanical filter + 30 L biofilter DWC systems, koi-fed, no macro/micronutrient fertilizer except one corrective iron dose in week 2. Paired ‘control’ = HP, a single optimized hydroponic DWC benchmark (also n=3 replicate systems) fertilized with a commercial two-part nutrient solution (Daeyu Mulpure No.1 A+B, 400x dilution) — the authors explicitly frame HP as an optimized benchmark, not a nutrient-matched control, so this is a diagnostic comparison rather than a factorial experiment (Abstract; Methods 2.2, p.3-4). Only one AP arm exists in this paper (trials: 1). | WARN-MATERIAL NO2-N (AP) mean, p.7-13. SS3.3 (p.7): ‘the concentrations in the AP system were 0.45 +/- 0.03 mg/L for the 35-day study period.’ Table 4 (p.9): ‘NO2-N 0.49 +/- 0.03 b’ (AP). SS3.4 Discussion (p.10-11): ‘AP’s values of 0.83 +/- 0.07, 0.49 +/- 0.03, and 81.12 +/- 3.10 mg/L’ (NH4/NO2/NO3 in that order). SS4 Discussion (p.13): ‘nitrite mean concentrations in the AP system were below 0.49 +/- 0.03 mg/L.’ Three independent restatements agree on 0.49; only the first Results mention (SS3.3) says 0.45. Recorded 0.49 +/- 0.03 (majority, reproduced in the paper’s own later synthesis). Affects: NO2-N cell only. | WARN-MATERIAL systematic SE-magnitude mismatch, SS3.3/SS3.2 vs Table 4/SS3.4 (one likely root cause, six parameters, counted as ONE material flag — see below). For six water-quality/plant-growth parameters the mean is identical between the first Results narrative and Table 4/Discussion, but the stated SE differs by roughly the same ~1.5-2x factor every time, always with Table 4’s SE larger: DO AP 6.33+/-0.09 (SS3.3) vs 6.32+/-0.19 (Table 4, SS3.4). DO HP 5.24+/-0.09 (SS3.3) vs 5.24+/-0.14 (Table 4). EC AP 0.92+/-0.01 (SS3.3) vs 0.92+/-0.03 (Table 4). EC HP 2.55+/-0.05 (SS3.3) vs 2.55+/-0.09 (Table 4). Turbidity AP 5.94+/-0.34 (SS3.3) vs 5.94+/-0.54 (Table 4). Turbidity HP 1.93+/-0.10 (SS3.3) vs 1.92+/-0.18 (Table 4). The consistent direction and rough scaling (~sqrt(3), the SD/SE ratio at n=3) suggests one source conflated SD and SE despite both being labelled SE (Table 4 caption, SS2.6), but the paper never says so. Table 4’s values used for all six cells (independently reproduced verbatim in Discussion, 2 sources vs 1); SS3.3’s figures given here as the alternate. Scored as ONE material flag since it looks like a single root cause, not six independent errors — noted so the user can recount individually if they judge that more appropriate. | WARN-MINOR SPAD (HP), p.6 vs p.9/p.11: 37.40+/-0.96 (SS3.2/Fig.2) vs 37.40+/-0.90 (Table 4, SS3.4). Same direction as above but smaller magnitude (~1.07x), doesn’t fit the sqrt(3) pattern; treated as a separate trivial rounding issue. Recorded 37.40+/-0.90 (majority). No column impact (HP SPAD has no dedicated cell; see NO COLUMN). | WARN-MINOR Plant fresh weight (HP), Abstract vs p.6/p.9: 165.6+/-3.01 (Abstract) vs 165.6+/-3.05 (SS3.2, Table 4). Recorded 165.6+/-3.05 (majority, more precise sourcing). No column impact (HP fresh weight has no dedicated cell; see NO COLUMN). | WARN-MINOR Total Fish Weight Gain, Table 3 vs running text, p.6: 0.38+/-0.034 kg (Table 3) vs 0.38+/-0.03 kg (running text). Recorded 0.38+/-0.034 (Table 3, more decimal precision) in Fish biomass created (kg). | WARN-MINOR leaf count narrative framing, p.6 vs p.11. SS3.2 states leaf number ‘showed better results for HP’ (21+/-0.58 vs 20.2+/-0.48 AP), but Table 4 marks both AP and HP with the same superscript letter and SS3.4 (p.11) explicitly states the difference is not significant (p>0.05). Not a numeric contradiction — raw values agree everywhere — just an interpretive overstatement in SS3.2’s prose. No cell affected. | NOT DERIVED, left NR: Fish size initial/final and Fish weight gain per fish (only system-level total biomass given for 20 fish collectively — 1.19+/-0.005 kg to 1.57+/-0.03 kg — not divided per fish); Fish survival rate (only ‘Fish Mortality (%) = 0’ is stated, Table 3; the complementary 100% survival figure is not itself written, so left NR rather than computed); Total Feed (kg) (FCR 1.29 and weight gain 0.38 kg would allow a ~0.49 kg estimate, but this is not stated as a figure and is not derived per the no-derivation rule); feed Protein/N/P/K composition (% of body weight ration — 1%/day of initial weight — is stated, but no compositional breakdown of the commercial feed is given); Water recycle L/min (peristaltic/submersible pumps used, HRT 3 h stated for the AP fish tank, but no flow rate in L/min given; computing one from HRT and tank volume would be derivation); Daily Water exchange rate as a % of system volume (only an absolute daily top-up volume is stated — ~2 L/day AP, ~1 L/day HP — to offset evaporation/transpiration/sampling losses, not a % exchange rate, and no summed system volume exists to compute one against); pHOptimal (a general 6.5-8.5 nitrification/lettuce range is cited from cited literature, ref [17], not measured or targeted by this study); Plant height (not among the measured traits — mass, leaf count, root length, chlorophyll only); Plant dry matter % (dry mass was measured per Methods 2.3 for tissue-nutrient digestion, but no dry-matter % or dry-weight figure is reported in Results/Table 4); Tissue nitrate AP/HYD (this paper measures dissolved water NO3-N and plant tissue Ca/Mg/K/P/Fe, not plant tissue nitrate); FUE AP/HYD, WUE (not calculated); Lat/Long (university address given — Kangwon National University, Chuncheon-si — but no coordinates stated anywhere in the paper); Fish Category, Water classification, Plant Category (no categorising term used by the paper itself); Statistics software (never named — only ‘paired two-tailed t-tests, p<0.05’ and Pearson r/p given, SS2.6). | NO COLUMN: HP-side water quality with no paired HYD column in the schema (Aq pH, DO, EC, Water temperature, TAN/NH4-N, NO2-N, NO3-N are AP-loop-only fields): NH4+-N (HP) 12.81+/-1.57 mg/L; NO2-N (HP) 1.37+/-0.30 mg/L; NO3-N (HP) 280.6+/-7.9 mg/L; EC (HP) 2.55+/-0.09 mS/cm; DO (HP) 5.24+/-0.14 mg/L; pH (HP) 6.75+/-0.06 (Table 4). Turbidity (no dedicated column): AP 5.94+/-0.54 NTU, HP 1.92+/-0.18 NTU (Table 4). PO4-P (no dedicated column): AP 19.2+/-1.7 mg/L, HP 13.8+/-1.2 mg/L (Table 4). Root length (no dedicated column): AP 21.8+/-0.44 cm, HP 25.83+/-0.50 cm (Table 4). HP-side plant growth for singular non-AP/HYD-paired columns: Leaf count HP 21.0+/-0.58; Plant fresh weight HP 165.6+/-3.05 g/plant; SPAD HP 37.40+/-0.90 (Table 4). PER (protein efficiency ratio) = 2.24+/-0.18 — no dedicated column. System-level fish % biomass increase over 35 days = 31.95+/-3.21% (distinct from the daily feed-ration %, which IS recorded in ’% of body weight’). HRT (fish tank) = 3 h — distinct metric from the ‘Water recycle’ flow-rate column, which the paper does not state. Fish volumetric unit productivity = 6.17 kg/m3/cycle — the lettuce area-yield figures (5.96 AP / 6.71 HP kg/m2/cycle) are recorded in the generic AP/HYD columns below, but this fish-side volumetric figure has no matching column. NH4-N vs NO3-N correlation: AP r=-0.564 (p=0.0095), HP r=0.686 (p=0.0009) (Figure 7). Batch-test design parameters (Table 2, not from the 35-day main trial, so not entered as trial-mean cells): NH4+-N release by fed fish = 4.13+/-0.08 mg/L over 36 h; NH4+-N release, unfed control = 0.55+/-0.01 mg/L over 36 h; plant nutrient uptake rates (6-day batch test, mg/L.d-1.plant-1) NO3-N 0.42+/-0.01, PO4-P 2.78+/-0.09, Ca2+ 0.13+/-0.006, K+ 0.46+/-0.01, Mg2+ 0.0101+/-0.0004. Plant tissue nutrient accumulation (P, K, Fe, Mg2+, Ca2+) from Table 4 is captured in plant_measurements.csv per SCHEMA.md, not here. | UNIT CONVERSION ONLY: none required; all values transcribed in the units printed.
Plant Measurements
| Trial | System | Category | Analyte | Value | Unit | Sig. | Location |
|---|---|---|---|---|---|---|---|
| mehdiEvaluatingPerformanceLimitations2026-T1 | AP | mineral | Phosphorus (P) | 8961.5 ± 236.4 | UNCLEAR | b | Table 4, p.9 (‘Plant Nutrient Accumulation’) |
| mehdiEvaluatingPerformanceLimitations2026-T1 | HYD | mineral | Phosphorus (P) | 11346.2 ± 395.6 | UNCLEAR | a | Table 4, p.9 (‘Plant Nutrient Accumulation’) |
| mehdiEvaluatingPerformanceLimitations2026-T1 | AP | mineral | Potassium (K) | 16517.7 ± 586.8 | UNCLEAR | b | Table 4, p.9 (‘Plant Nutrient Accumulation’) |
| mehdiEvaluatingPerformanceLimitations2026-T1 | HYD | mineral | Potassium (K) | 19161.1 ± 791.3 | UNCLEAR | a | Table 4, p.9 (‘Plant Nutrient Accumulation’) |
| mehdiEvaluatingPerformanceLimitations2026-T1 | AP | mineral | Iron (Fe) | 301.1 ± 11.8 | UNCLEAR | a | Table 4, p.9 (‘Plant Nutrient Accumulation’) |
| mehdiEvaluatingPerformanceLimitations2026-T1 | HYD | mineral | Iron (Fe) | 232.6 ± 8.4 | UNCLEAR | b | Table 4, p.9 (‘Plant Nutrient Accumulation’) |
| mehdiEvaluatingPerformanceLimitations2026-T1 | AP | mineral | Magnesium (Mg) | 3009.3 ± 125.8 | UNCLEAR | a | Table 4, p.9 (‘Plant Nutrient Accumulation’) |
| mehdiEvaluatingPerformanceLimitations2026-T1 | HYD | mineral | Magnesium (Mg) | 2115.7 ± 77.5 | UNCLEAR | b | Table 4, p.9 (‘Plant Nutrient Accumulation’) |
| mehdiEvaluatingPerformanceLimitations2026-T1 | AP | mineral | Calcium (Ca) | 4110.6 ± 160.1 | UNCLEAR | a | Table 4, p.9 (‘Plant Nutrient Accumulation’) |
| mehdiEvaluatingPerformanceLimitations2026-T1 | HYD | mineral | Calcium (Ca) | 3285.6 ± 131.0 | UNCLEAR | b | Table 4, p.9 (‘Plant Nutrient Accumulation’) |