Nitrogen transformation in engineered aquaponics with water celery (Oenanthe javanica) and koi carp (Cyprinus carpio): Effects of plant to fish biomass ratio

Metadata

Cite key: paudelNitrogenTransformationEngineered2020
Item type: Journal Article
Authors: S. Paudel
Affiliation: Department of Civil Engineering, Pulchowk Campus, Institute of Engineering, Tribhuvan University, Pulchowk, Lalitpur, Nepal; Department of Environmental Engineering, College of Science and Technology, Korea University, Sejong, Republic of Korea
Journal: Aquaculture 520 (2020) 734971
Date: Received 31/08/2019; revised 15/12/2019; accepted 15/01/2020; available online 16/01/2020
Date added: 2026-08-10
DOI: 10.1016/j.aquaculture.2020.734971
Funding: Not stated. The Acknowledgement section thanks the Department of Civil Engineering (Tribhuvan University) and Korea University for providing a working environment and laboratory facilities, and the Declaration of Competing Interest states no conflict of interest was reported — but no grant, funding body, or award number is identified anywhere in the paper.
URL: https://doi.org/10.1016/j.aquaculture.2020.734971
PDF: Paudel - 2020 - Nitrogen transformation in engineered aquaponics w.pdf

Opinion

A single-author, single-system engineering study built around a genuinely useful question (does raising the plant-to-fish biomass ratio reduce N2O emissions and improve nitrogen recovery in a zero-water-exchange aquaponic loop?) with a real methodological limitation baked into the design: there is only ONE physical bench-scale system, passed sequentially through four increasing P/F ratios in a fixed order (never randomized, never independently replicated). The paper’s own “n=21” is 21 repeated water samples taken every 2nd day within each phase from that same single system — temporal pseudo-replication, not the independent experimental units the ANOVA/Tukey HSD testing implicitly assumes. That does not make the trend reported (higher P/F ratio -> higher NUE, lower N2O, plant yield-fraction slightly declining) implausible or uninteresting, but it does mean the paper’s p-values should be read as “this single system’s phases differed” rather than “P/F ratio causes these differences across systems,” and it is why this note classifies the paper as quasi-experiment rather than experiment. Two smaller issues compound this: the P/F ratio labels for the two middle phases are given as 0.30/0.50 in the Methods text and three tables but as 0.29/0.49 in a fourth table and a figure caption (no consequence for any measured value, just the treatment label itself); and the paper’s own headline “17% reduction in N2O emission factor” claim does not reconcile with its own Table 4 numbers for that specific metric (the 17% figure matches the emission-rate column, not the emission-factor column it is verbally attached to) — see Extraction notes for the full arithmetic. On the data side, this is also the only paper encountered so far in this vault with no hydroponic-only control at all: the comparison here is aquaponics (at varying P/F ratios) against a fish-only aquaculture control, not against a plant-only hydroponic system, so the vault’s usual AP-vs-HYD yield framework does not really apply and every HYD cell in this paper’s rows is NA. Plant “yield” itself is also only ever reported as a whole-4-bed-system total in grams, never per-plant or per-m2, which limits comparability with the rest of the vault’s yield-centric lettuce/basil literature. Still a worthwhile paper for the N2O/nitrogen-budget angle, which is genuinely under-represented in aquaponics literature that focuses on biomass and water quality alone.

Abstract

Nitrogen plays critically important role in dual biomasses aquaponic production. This study examined the fate of nitrogen in an aquaponics at different plant-fish (P/F) biomass ratios. A floating raft hydroponic bed with water celery (Oenanthe javanica) as plant species coupled with an aquaculture tank reared with Koi carp (Cyprinus carpio) was evaluated using nitrogen mass balance techniques at P/F biomass ratio of 0.06–0.95 and feed conversion ratios (FCR) maintained at 1.73–1.85 at a constant hydraulic retention time of 24 h. The FCRs in the study were significantly lower than FCR (1.96) in the control (aquaculture only). Nitrogen use efficiency (NUE) in the aquaponics increased from 29.3 ± 3.5 to 39.1 ± 4.2% as the P/F biomass ratio increased from 0.06 to 0.95, respectively. Furthermore, the increase in P/F biomass ratio to 0.95 exhibited the highest nitrogen uptake rate by plants (~56 mg/day in the hydroponic bed) which resulted in a levelling-off of the nitrate concentration at 62.5 ± 4.2 mg/L in the aquaculture tank. The increase in P/F biomass ratio also lowered the nitrous oxide (N2O) emissions by 17% from the aquaponics compared to control. Overall, this study presents evidence that an aquaponics with higher P/F biomass ratio emits less N2O and has higher nutrient recovery with perpetual fish production in a zero water exchange-aquaponic system.

Summary

The author operated a single bench-scale, zero-water-exchange aquaponic system (a 200 L Koi carp tank feeding four 4 L floating-raft water celery beds in series) for 125 days, sequentially raising the plant-to-fish (P/F) biomass ratio through four fixed phases (0.06, 0.30, 0.50, 0.95) by replanting water celery cuttings of increasing mass at each phase change, and comparing fish and water-quality performance against a separate fish-only aquaculture control with no plants. As the P/F ratio rose, fish FCR improved slightly (1.73 to 1.85, all better than the fish-only control’s 1.96), nitrogen use efficiency rose from 29.3% to 39.1%, plant nitrogen uptake and root-zone denitrification both increased sharply (nitrogen removal rate in the hydroponic bed rising from 4.5 to 56 mg N/day), and nitrous oxide emissions from the fish tank fell relative to the control. A qPCR functional-gene assay taken at the end of the third phase found ammonia-oxidizing bacteria (amoA gene) far more abundant in the biofilter than on the plant roots, while the reverse pattern for general bacterial (16S rDNA) abundance suggested denitrifiers were relatively more common on the water celery root mats — supporting the paper’s interpretation that nitrification is biofilter-driven and denitrification is root-driven in this system. The paper frames its main contribution as evidence that raising a system’s plant-to-fish biomass ratio is a viable design lever for simultaneously improving nitrogen recovery and reducing greenhouse-gas emissions in a fully recirculating, zero-discharge aquaponic system, while flagging that the underlying mechanism (root-surface biofilm formation and root-zone anoxic microsites promoting denitrification) needs further targeted study across other plant/fish species and operating conditions.

Experiment data

Location: Not explicitly stated as a study site (see Extraction notes on the Region/Country judgment call); bench-scale laboratory system, author affiliated with Tribhuvan University (Nepal) and Korea University, Department of Environmental Engineering, Sejong Campus (Republic of Korea)

Design: Single bench-scale recirculating aquaponic system (200 L Koi tank + 4 L settler + 4 x 4 L floating-raft water celery beds in series, zero water exchange except <=0.50% evapotranspiration replenishment/2nd day) run continuously for 125 days through 4 sequential, fixed-order plant-to-fish (P/F) biomass ratio phases (0.06, 0.30, 0.50, 0.95), compared against a separately-run fish-only aquaculture control. No randomization or independent replication of treatment units is stated — see Extraction notes on the experiment/quasi-experiment judgment call.

Replicates / n: 1 physical system per phase (no independent replicate units); Table 1’s stated “n=21” is 21 repeated water-quality samples taken every 2nd day within each phase from the same single system, not independent experimental units

Duration: 125 days total operating period (after an initial 2-week fish-only acclimation period), covering all 4 phases + control; per-phase duration not separately stated

Organisms: Water celery (Oenanthe javanica) / Koi carp (Cyprinus carpio)

Statistics: One-way or two-way ANOVA followed by Tukey HSD (p<=0.05), SPSS v16.0

Nitrogen Use Efficiency (NUE): rose from 29.3 +/- 3.5% (P/F 0.06) to 39.1 +/- 4.2% (P/F 0.95)

Feed Conversion Rate (FCR): 1.73, 1.75, 1.80, 1.85 across the four phases vs. 1.96 in the fish-only control

N2O emission factor: 0.56 +/- 0.12, 0.55 +/- 0.21, 0.53 +/- 0.16, 0.50 +/- 0.10 g N2O-N/kg fish across the four phases vs. 0.68 +/- 0.17 in the control (Table 4)

Nitrogen transformation and the plant-to-fish biomass ratio

This paper: As the P/F biomass ratio increased from 0.06 to 0.95, the hydroponic bed’s nitrogen removal rate rose from 4.5 to 56.0 mg N/day, plant uptake rate from 2.55 to 34.5 mg N/day, and the root-zone denitrification rate from 1.55 to 20.5 mg N/day (Table 3). At the highest ratio, plant assimilation and denitrification together accounted for roughly 13.6% and 7.5% of the total nitrogen input (as feed) respectively. Total NUE (fish + plant assimilation) rose from 29.3% to 39.1% over the same range, while the fraction of nitrogen retained in the water fell from 50.5% to 27.8%, and total gaseous nitrogen loss (N2O + N2) rose from 17.75% to 30.16% (Section 4.4, Fig. 4). Specific nitrogen uptake by water celery tissue was in the range 2.41-2.98 mg N/g plant (an aggregate figure spanning all phases, not broken out per phase), which the author notes is lower than a comparable figure from a batch (non-continuous-flow) experiment with the same species.

Compared with:

#todo Santamaria et al. 1999 — batch-culture water celery specific N uptake of 7.8 mg N/g, higher than this paper’s continuous-flow range of 2.41-2.98 mg N/g; author attributes the gap to different kinetic conditions (contact time, bulk N concentration) between batch and continuous-flow systems (p.4)
#todo Endut et al. 2010 — varying hydraulic loading rate increased nitrogen uptake and plant growth rate; cited as a related design lever and as the source of the batch-vs-continuous kinetic explanation above (p.2, p.4)
#todo Hargreaves 1998 — theoretical fish nitrogen recovery of 25% in aquaculture ponds; this paper’s fish-biomass nitrogen recovery of 26.5-28.4% is described as “very close” to that theoretical value (p.6)
#todo Hu et al. 2014 — reported 32% total nitrogen loss in an intensive aquaculture system, compared against this paper’s 17.75-30.16% gaseous N loss range (p.6)
#todo Thakur and Lin 2003 — reported gaseous nitrogen loss of 5.2-36.0% in an intensive shrimp (Penaeus monodon) culture system, given as a wider benchmark range for the same comparison (p.6)
#todo Wongkiew et al. 2017b — reported that low hydraulic retention time (below 1 m3/m2/day) reduces nitrite oxidation and accelerates denitrification with nitrite accumulation; cited as background on HRT effects on nitrogen species (p.2)
#todo Hu et al. 2015 — effect of plant species on nitrogen recovery in aquaponics; cited for the idea that greater root biomass provides more surface area/habitat for nitrifying and denitrifying microbes, and for the assumption (adopted here) that N2O emission from the hydroponic bed itself is negligible at the long (1-day) HRT used (p.4-5)

Nitrous oxide (N2O) emissions

This paper: N2O was monitored (dissolved and gaseous) in 3 h intervals across a diurnal cycle at the end of each phase. Both forms rose for about 8 h after feeding, consistent with fish excretion timing, and both varied less at higher P/F ratios, with the highest ratio (0.95) showing the lowest and most stable N2O levels. Table 4 gives, by phase: N2O emission rate (mg/day) 1.53, 1.48, 1.41, 1.30 (vs. 1.56 in the fish-only control); emission factor (g N2O-N/kg fish) 0.56, 0.55, 0.53, 0.50 (vs. 0.68 in the control); and %N2O-N conversion of feed nitrogen input 0.56%, 0.54%, 0.51%, 0.47% (vs. 0.57% in the control) — all four metrics declining monotonically as P/F ratio rose. The author’s own headline figure, a “17% reduction” in N2O emission at the highest P/F ratio relative to control, is explicitly attached in the Results text to the emission-factor metric, but recomputing from Table 4’s emission-factor values gives a 26.5% reduction, not 17%; the 17% figure instead matches a recomputation from the emission-rate (mg/day) column (16.7%, rounds to 17%). This is flagged in detail in Extraction notes — it does not affect any trials.csv cell, since neither N2O metric has a dedicated schema column, but it matters for anyone citing the paper’s “17%” claim. One-way ANOVA/Tukey HSD found a significant difference in N2O concentration (both dissolved and gaseous) between the 0.50 and 0.95 P/F ratio phases (p<.05), which the author interprets as meaning the P/F ratio should exceed 0.5 to meaningfully mitigate N2O emission.

Compared with:

#todo Hu et al. 2013 — reported an N2O emission rate of 0.49 g N2O-N/kg fish from a comparable intensive aquaculture system, lower than this paper’s own fish-only control value of 0.68 g N2O-N/kg fish; author attributes the difference to a higher feeding rate used in this study (p.5)
#todo Ahn et al. 2010 — lowering NH4+/NO2-/NO3- concentration may lower N2O emission potential; cited as the mechanistic basis for why higher-P/F-ratio phases (which had lower TAN/nitrite) showed lower N2O (p.2, p.5)
#todo Ni et al. 2013 — evaluated mathematical models for N2O production by ammonia-oxidizing bacteria; cited alongside the amoA functional-gene interpretation (p.5, p.6)
#todo Smart and Bloom 2001 — documented that wheat leaves themselves can emit N2O during nitrate assimilation; cited as a caveat that plant-side (not just water-side) N2O emission from the hydroponic bed was not measured in this study and should be in future work (p.5)

Fish and plant performance (FCR, yield)

This paper: FCR improved modestly and monotonically with P/F ratio (1.73 to 1.85) and was better than the fish-only control’s 1.96 at every phase, which the author attributes to improved water quality (lower TAN/nitrite/nitrate) provided by the plants. No fish mortality or plant disease was observed over the full 125-day period. Plant (water celery) harvested biomass rose from 18.0 g to 310 g (whole 4-bed-system totals, not per-plant or per-m2) as initial planted biomass was increased across phases to hit each target P/F ratio, while the harvested-to-initial “plant yield” percentage slipped slightly, from 24.0% to 21.4%, as the ratio increased — i.e., proportionally slightly less of the larger initial biomass was gained back by harvest at the highest ratio, even though absolute harvested mass was far higher.

Compared with:

#todo Naylor et al. 2000 — cited for a “conventional aquaponic system” FCR range that this paper’s 1.73-1.85 results are said to fall within, contrasted with the higher control FCR of 1.96 (p.3)
#todo Molleda et al. 2007 — water quality in recirculating aquaculture systems for Arctic charr; cited generically to support the claim that good water quality (via nitrogen removal) underlies good fish growth performance (p.4)

Citations to chase

Extraction notes

Type classification judgment call (quasi-experiment, not experiment): The paper reports a defined set of manipulated treatments (four P/F biomass ratio phases) and a formal statistical test (one-way/two-way ANOVA + Tukey HSD, p<=0.05, Section 3), which would suggest experiment under SCHEMA.md’s simple test. However, SCHEMA.md’s Part 2 decision rule 2 specifically requires “randomised treatments with replication” for experiment; here there is exactly one physical bench-scale system, run through the four phases sequentially in a fixed, increasing order (never randomized), and the paper’s own stated “n=21” (Table 1) is 21 water samples taken every 2nd day within each phase from that same single system — i.e., repeated/temporal pseudo-replication of one experimental unit, not independent replicate units of each treatment. This is a materially different (weaker) design than the vault’s other ANOVA-tested papers (e.g. mourantianBasilFunctionalGrowth2023’s 6 independent channel-replicates/treatment, pantanellaAquaponicsHydroponicsProduction2012’s 3 independent tank-systems/treatment). Per SCHEMA.md’s own rule (“without randomisation or true replication -> quasi-experiment”), quasi-experiment was judged the better fit. This is a design-quality observation, not a numeric contradiction, so it carries no ⚠️ severity tag, but it is the single most important caveat for interpreting this paper’s statistics and is called out in the Opinion callout and in every trial row’s Experimental Remarks under “Replicates (n)”.

Trial structure: Four trials.csv rows, one per P/F biomass ratio phase (T1=0.06, T2=0.30, T3=0.50, T4=0.95). All four share ONE physical system and ONE continuous Koi cohort (16 fish, 84.8 +/- 20.2 g at initial stocking) run sequentially; each row’s TRIAL DEFINITION states this explicitly. This paper has no hydroponic-only (plant-only) control — the only comparator is a separately-run fish-only aquaculture control (no plants) — so every HYD-labelled cell (HYD, Tissue nitrate HYD) is NA rather than NR across all four rows, per SCHEMA.md’s “papers missing half the schema” convention. The fish-only control’s own values (FCR 1.96; N2O emission rate/factor/conversion-% per Table 4) have no dedicated column in the schema (there is no “fish-only control” field distinct from HYD) and are recorded as NO COLUMN items in each row’s remarks instead of being force-fit into the HYD columns, since a fish-only control is a conceptually different comparator than a plant-only hydroponic control.

⚠️MATERIAL — P/F biomass ratio label for Phase II (T2), p.3-6. Methods 2.2 (“0.06, 0.30, 0.50, and 0.95”), Table 1, Table 2, and Table 3 all state Phase II’s nominal ratio as 0.30. Table 4 and the Fig. 2 caption instead give 0.29 for the same phase (“Phase I, II, III, & IV represents different plant fish biomass ratio of 0.06, 0.29, 0.49, and 0.95 respectively”). Four locations (Methods text + 3 tables) support 0.30 against two (1 table + 1 figure caption) supporting 0.29; Methods 2.2 is also the paper’s primary definitional statement of the intended treatment levels. Recorded 0.30 as T2’s defining ratio; 0.29 preserved in T2’s Experimental Remarks as the minority reading. Affects only which nominal ratio label the row is filed under — every measured value for Phase II (FCR, feed, biomass, water quality) is internally consistent regardless of which ratio label is used, so no other cell is affected.

⚠️MATERIAL — P/F biomass ratio label for Phase III (T3), p.3-6. Same pattern: Methods 2.2, Table 1, Table 2, and Table 3 state 0.50; Table 4 and the Fig. 2 caption state 0.49. Recorded 0.50 as T3’s defining ratio for the same majority-source reasoning; 0.49 preserved in T3’s remarks. No other cell affected.

WARN, paper-level, no cell affected — the paper’s own “17% N2O reduction” claim does not reconcile with its own table. Abstract and Results 4.3 both state a 17% reduction in N2O emission at the highest P/F ratio versus control, and Results 4.3 explicitly names the metric: “N2O emission factor (g N2O/kg fish) decreased by 17%”. Recomputing from Table 4’s Emission factor column (Control 0.68 -> Phase IV 0.50 g N2O-N/kg fish) gives 26.5%, not 17%. Recomputing instead from the adjacent Nitrous oxide emission rate column (Control 1.56 -> Phase IV 1.30 mg/day) gives 16.7%, which rounds to 17%. So the paper’s stated percentage matches a different column than the one it names. This was checked by recomputation only (per SCHEMA.md’s explicit allowance to verify a reported figure without entering the recomputed number in a cell) and is recorded in T4’s Experimental Remarks as evidence. It receives no formal ⚠️BLOCK/MATERIAL/CHECK tag and no REVIEW.md row because neither N2O metric (emission rate or emission factor) has a dedicated trials.csv column — no cell is at stake — but it is flagged prominently here and in the Opinion callout because it is the paper’s own headline result.

Duration interpretation (not a formal contradiction): Fish trial duration (days) is recorded as 125 for all four rows, the paper’s one explicitly stated figure for “the operating period” covering all four phases plus control (Section 4.1; Methods 2.2). The paper never states how many of the 125 days belong to each individual phase, so the same total is used for every row rather than attempting an unstated per-phase split. A separate ambiguity (not a numeric conflict, since no second candidate number is offered) is whether the preceding 2-week fish-only acclimation period is included inside or additional to the 125 days — Methods 2.2’s wording suggests “additional to,” while Fig. 2’s 0-125 day x-axis plots the acclimation phase inside that same window. Noted for transparency in each row’s remarks; does not change the recorded value.

Water-quality compartment convention: Per SCHEMA.md’s rule to take the plant-bed/hydroponic-unit value when a paper reports separate per-compartment figures, Aq pH, Dissolved Oxigen, and Water temperature all use Table 1’s hydroponic-bed (H-bed) column values, with the fish-tank column values noted alongside in each row’s remarks for reference. The one exception is NO3-N, where only the fish-tank compartment has an explicit trial-level summary figure (the Abstract’s “62.5 +/- 4.2 mg/L” levelling-off value at Phase IV/P/F=0.95); no H-bed NO3-N summary value is stated anywhere in text or tables (only via the Fig. 2c time-series, which is not extracted per the never-read-a-figure rule), so T4’s NO3-N cell necessarily uses the fish-tank figure with this deviation explained in the cell itself. T1-T3 have no stated NO3-N summary at all (time-series only) and are NR.

Region/Country left UNCLEAR (judgment call): The paper never states where the physical bench-scale system was located. Circumstantial evidence points in two directions: the author’s dual affiliation spans Tribhuvan University (Nepal) and Korea University’s Department of Environmental Engineering (Sejong, Republic of Korea); nearly all named equipment suppliers are Korean (Youngi Electric, Sinnong, Woo Sung Feed Co. Ltd., DS 6200 DSS gas chromatograph). This is suggestive of South Korea but is never stated as the study site (contrast with, e.g., pantanellaAquaponicsHydroponicsProduction2012’s explicit “Experimental Farm of Tuscia University, Viterbo”). Per the prime directive and per SCHEMA.md’s coordinate-resolution principle (“read only if the result is geographically consistent with a stated site”), Region and Country are both recorded UNCLEAR rather than inferred from equipment brand/affiliation alone, and no Meta/Region/ tag was applied. Lat/Long are NR (no coordinates given anywhere).

[not reported] / NR fields, grouped (identical pattern across all four rows since they share one system/population):

Fish: Fish Category, Initial Stock density, SGR, N/P/K feed composition beyond crude protein, % of body weight, Fish size final, Fish weight gain (per-fish), pHOptimal, FUE AP/HYD, WUE, EC, TAN/NH4-N, NO2-N — either never categorized/stated, or only available as an aggregate/total that would require derivation to convert to the schema’s expected per-fish or per-time basis, or (for TAN/NO2-N) presented only as an uncaptioned daily time-series figure (Fig. 2a/2b) with no table or text summary value.
Water: Water type, Water classification, Daily Water exchange rate (stated only as ”<=0.50% every 2nd day”, not a daily rate).
Plant: Plant Category, Days Plant after transplant, Plants/m2, SPAD, Plant height (only initial planting-stalk height given, not harvest height), Leaf count, Plant fresh weight (only whole-system totals given, see AP column and remarks instead), Plant dry matter, Tissue nitrate AP (no tissue nitrate/food-safety assay performed; this paper measures total plant N for a mass-balance nitrogen-uptake calculation, not edible-tissue nitrate).
If the same field pattern (whole-system totals not convertible to per-plant/per-m2 without derivation) recurs in later engineering-style aquaponics papers, that is a sign the schema’s Plant fresh weight (g/plant) column needs either an explicit whole-system-total escape hatch or a dedicated area/count field, rather than repeated per-paper judgment calls about where to park total biomass.

PDF quality: Clean, fully machine-readable text layer throughout (8 pages, standard two-column Elsevier Aquaculture typesetting, digitally native PDF with a search-highlighted term visible in the extracted text — “water celery” and a few other terms carry a highlight artifact from a prior reader, not an OCR issue). No pages required re-reading; no scanned/garbled text encountered.

New tags introduced: Meta/Fish/Koi reused exactly as spelled in andersonGrowthTissueElemental2017 and leePhotosynthesisGrowthYield2022. Meta/Plant/Water-Celery is new (no existing tag for Oenanthe javanica in the vault; follows the existing hyphenated Meta/Plant/Water-Spinach convention from endutaNutrientRemovalAquaculture2011, a different species, Ipomoea aquatica, not to be confused with this one). Meta/Type/Quasi-experiment reused exactly as spelled in nozziNutrientManagementAquaponics2018, delaideLettuceLactucaSativa2016, goddekComparisonLactucaSativa2018, graberAquaponicSystemsNutrient2009, and lennardComparisonPlantGrowth2019. No Meta/Region/ tag applied — see judgment-call note above.

New wikilink targets introduced: [[S. Paudel]] (no existing author note found in the vault), [[Water celery (Oenanthe javanica)]], [[Koi carp (Cyprinus carpio)]] (no existing organism notes found). Reused [[Feed Conversion Rate (FCR)]] per the canonical form named in CLAUDE.md’s own worked example. [[Nitrogen Use Efficiency (NUE)]] is new — no existing note found under this or a fragment spelling (NUE, Nitrogen use efficiency); flagged here per CLAUDE.md’s instruction to list rather than guess the canonical form if this later turns out to duplicate an existing concept note.

quality: ok — 0 ⚠️BLOCK, 2 ⚠️MATERIAL (both P/F ratio labeling, neither affecting a measured-value cell), 0 ⚠️CHECK (so no REVIEW.md rows from this paper). Per SCHEMA.md’s thresholds (ok = 0 BLOCK and <=2 MATERIAL), this lands exactly at the top of the ok band.

Source: Paudel - 2020 - Nitrogen transformation in engineered aquaponics w.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

paudelNitrogenTransformationEngineered2020-T1

Fish

Field	Value
Fish	Koi carp (Cyprinus carpio)
FCR	1.73
Protein	32
Fish size initial	84.8 +/- 20.2
Feed routine	Once daily; pellets consumed within 10 min of feeding (Methods 2.2)
Feed regime	Commercial feed pellets (Woo Sung Feed Co. Ltd., South Korea), fed at a constant rate of 4.75 g/day; composition per manufacturer specification (registration CCBDN0055): 32% protein, 2% lipids, 1% calcium, 5% fibre, 1.8% phosphate, dry-weight basis (Methods 2.2)
Total Feed (kg)	0.0998
Fish biomass created (kg)	0.0577
Fish survival rate	100% (no fish mortality observed during the 125-day operating period, Section 4.1)
Fish trial duration (days)	125

Water

Field	Value
Water recycle	0.0069
Water volume in the system	200 (fish tank only, effective volume, 80x50x50 cm; system also includes a 4 L settler and 4 x 4 L floating-raft hydroponic beds = 16 L total bed volume; total combined system volume not stated as one figure, Methods 2.1)
Aq pH	6.9 +/- 0.1
Dissolved Oxigen	6.5 +/- 0.2
Water temperature	24.2 +/- 0.3

Plant

Field	Value
Plant	Water celery (Oenanthe javanica)
Details	Cut-stalks with 2-3 nodes (initial height 22 +/- 4.2 cm) planted into floating-raft hydroponic beds at differing masses to achieve the target P/F biomass ratio; replanted at each phase change (Methods 2.2)

System & Setup

Field	Value
System type	Floating raft (hydroponic bed with gravel + sponge root support, connected in series, Methods 2.1)
Media Details	Half of each hydroponic pot (180 mL, Sinnong, Korea) filled with gravel (5-15 mm diameter) and hydroponic urethane sponge (77 pieces) to hold plants upright; 4 floating-raft hydroponic beds connected in series, each 4 L effective volume (100x10x4 cm), covered with black sheet (Methods 2.1)
Biological system already in use	Y (Biofilter bag (100 sponge media, 1x1x1cm) pre-immobilized with activated sludge before system startup; biomass on media measured as 4.45 +/- 0.22 mg VSS/media, per method of Paudel et al. 2015 (Methods 2.1))
Air supplement	Y (Ceramic air diffuser connected to an air pump in the fish tank (Methods 2.1))
pH Buffers	Y (5 g sodium bicarbonate periodically applied to the aquaculture tank to maintain neutral pH (Methods 2.2, citing Boley et al. 2000))
Climate control	Y (Thermostat and chiller maintained system water temperature at 24 +/- 0.5 C; ambient room temperature was not controlled and ranged 20-24 C throughout (Methods 2.2))
Artificial Lighting	Y (Fluorescent light (ML-2, 12V, Youngi Electric, Korea) mounted 1.1 m above the hydroponic beds; intensity maintained 490-510 lux; controlled by a 24 h timer giving a 12 h light-on photoperiod (Methods 2.1))
Equipment	Submersible mixing pump (YI-50, Youngi Electric, Korea); ceramic air diffuser + air pump; biofilter bag with 100 sponge media (1x1x1cm) pre-immobilized with activated sludge; Thermo Scientific Orion 5 star meter (online DO/pH/temperature sensors, fish tank and hydroponic-bed outlet); thermostat and chiller (system water temperature control); fluorescent light (ML-2, 12V, Youngi Electric, Korea) on a 24 h timer; peristaltic pump (recirculation, 10 L/day); HACH reaction kit (TN, TAN); ion chromatography (ICS-900, Dionex, US) for NO2-/NO3-; analytical balance (Model GF-3000); gas chromatograph (DS 6200 DSS, Korea) with PDHI detector and Hayesep D packed column (N2O); Illumina Eco Real Time PCR System (qPCR, amoA/16S rDNA)
Control Parameters	DO target ~5.5 mg/L (fish tank, for optimal fish growth, Methods 2.2); pH maintained neutral via periodic 5 g sodium bicarbonate dosing; system water temperature 24 +/- 0.5 C via thermostat/chiller; room air temperature uncontrolled, 20-24 C ambient; light intensity 490-510 lux, 12 h light/12 h dark photoperiod; flow rate 10 L/day (1-day hydraulic retention time in hydroponic bed); feed rate fixed at 4.75 g/day
Combination	Koi carp (Cyprinus carpio) and water celery (Oenanthe javanica) in a single bench-scale recirculating floating-raft aquaponic system, operated sequentially through 4 increasing plant-to-fish (P/F) biomass ratio phases, compared against a fish-only (no plant) aquaculture control run in the same type of system

Site

Field	Value
Average room Temperature	20-24 (range only, no trial mean reported; ambient/uncontrolled room temperature, Methods 2.2)

Results & Statistics

Field	Value
Measured Unit	mg/L (TN, TAN, NO2-N, NO3-N in water); g (fish/plant biomass, dry wt for feed, wet wt for fish/plant biomass per Table 2 footnotes); mg N/day (nitrogen removal/uptake/denitrification rates, Table 3); mg/day and g N2O-N/kg fish (N2O emission rate and emission factor, Table 4); mg N/g plant (specific plant N uptake, text)
Statistic Details	One-way or two-way ANOVA followed by Tukey HSD (p<=0.05), SPSS version 16.0 (IBM Corporation, Somers, NY) (Section 3, Statistical analysis)
Statistically analysed	Y
Replicates (n)	1
AP	18.0 +/- 1.5

Experimental Remarks: TRIAL DEFINITION: T1 = aquaponic treatment at plant-to-fish (P/F) biomass ratio 0.06 (Phase I), run as one segment of a single continuously-operated bench-scale recirculating system (200 L Koi tank + 4 L settler + 4 x 4 L floating-raft water celery beds in series, zero water exchange except <=0.50% evapotranspiration replenishment every 2nd day, Methods 2.1-2.2). Water celery cut-stalks (2-3 nodes, initial height 22 +/- 4.2 cm) were replanted at each phase change to achieve the target P/F ratio. Paired control = a separate fish-only aquaculture run with no plants coupled (same tank type); this paper has no hydroponic-only (plant-only) control, so all HYD-labelled columns are NA. Control values: FCR 1.96; N2O emission rate see Table 4 (NO COLUMN) mg/day; N2O emission factor see Table 4 (NO COLUMN) g N2O-N/kg fish; %N2O-N conversion see Table 4 (NO COLUMN)% (Table 4) — recorded here in remarks as NO COLUMN since the schema has no fish-only-control field distinct from HYD. All four phases share ONE physical system and ONE continuous fish population/cohort (16 Koi, 84.8 +/- 20.2 g at initial stocking) run sequentially in fixed order I->II->III->IV (not randomized, not independently replicated) — see Extraction notes for the design/quasi-experiment judgment call. Table 1 water-quality figures used here take the hydroponic-bed (H-bed) compartment value per SCHEMA.md’s compartment rule: Aq pH 6.9 +/- 0.1 (fish-tank value 6.8 +/- 0.1 for reference); DO 6.5 +/- 0.2 mg/L (fish-tank 7.0 +/- 0.1 mg/L for reference); water temperature 24.2 +/- 0.3 C (fish-tank 24.5 +/- 0.4 C for reference), all Table 1, Mean +/- SD, n=21 water samples/phase (see Extraction notes on non-independence of this n). Plant: initial biomass 75.0 +/- 3.0 g (wet wt) planted at phase start; harvested biomass 18.0 +/- 1.5 g (wet wt, whole 4-bed system total, NOT per-plant or per-m2 — no plant count or bed area given for a per-plant/per-m2 conversion without derivation); plant yield 24.0 +/- 0.5% (= harvested/initial biomass x100, paper’s own stated %, Table 2, independently verified to match the underlying Table 2 masses to within rounding — recomputation used only to check the paper’s own arithmetic, not entered as a derived cell per SCHEMA.md); root biomass 2.2 +/- 0.2 g (Table 2, NO COLUMN, no dedicated field for root mass exists in the schema). AP column above holds the harvested-biomass figure with its Measured Unit noted as a whole-system total, not per-plant/per-m2, since no per-plant fresh-weight column value could be computed without derivation. | Replicates (n): recorded as 1 (one physical bench-scale system, no independent replicate units of any treatment/phase). Table 1’s caption states ‘sample size (n) = 21’, which is the count of repeated water-quality samples taken every 2nd day WITHIN each phase from the single system, not independent experimental units; ANOVA + Tukey HSD (Section 3) is applied across phases using these 21-per-phase repeated samples, which is temporal pseudo-replication rather than the schema’s ‘experimental units per treatment’ definition. This design property, not a numeric contradiction, is the basis for classifying the paper as quasi-experiment rather than experiment (see Extraction notes) and is flagged here for anyone using n=21 downstream as if it were independent replication. | Fish trial duration recorded as the paper’s one explicitly stated overall figure, 125 days (‘the operating period (125 days)’, Section 4.1; ‘the aquaponic system was continuously operated for 125 days by applying four different P/F biomass ratios together with control’, Methods 2.2) applied identically to all four rows, since all four phases are segments of one continuous run and the paper never states how many of the 125 days belong to each individual phase. A separate 2-week acclimation period is also mentioned before the hydroponic beds were first coupled; Methods 2.2’s wording ties the ‘125 days’ figure to the phases+control period specifically (after coupling), while Fig. 2’s 0-125 day x-axis appears to place the acclimation phase INSIDE the same 0-125 day window. This is a minor internal ambiguity about whether acclimation is included in or additional to the 125 days; it does not change the recorded cell value (125, the only number the paper gives) and is noted here for transparency rather than tagged as a formal contradiction, since no second numeric candidate is offered anywhere in the paper. | NOT DERIVED, left NR: Initial Stock density (16 Koi fish, 84.8 +/- 20.2 g each, in a 200 L tank are all stated, but no kg/m3 figure is given; computing 16x84.8g/0.2m3 would be derivation); % of body weight (feed fixed at 4.75 g/day, fish biomass stated, but no daily-ration-as-%-of-body-weight figure given); Fish size final (only aggregate tank biomass increase per phase given, Table 2, not a final mean individual weight); Fish weight gain per fish (only the population-level aggregate biomass increase is stated, recorded instead under Fish biomass created); N, P, K feed composition beyond crude protein (Methods 2.2 gives protein/lipid/calcium/fibre/phosphate %, not elemental N/P/K in the schema’s expected basis — ‘phosphate’ is not necessarily elemental P and converting it would assume a basis not stated, so left NR, full figures under NO COLUMN); Daily Water exchange rate (paper states ’<=0.50% on every 2nd day’, not a stated daily %; halving it to a daily-equivalent rate would assume linear averaging not stated by the authors, so left NR rather than derived); Plants/m2, Days Plant after transplant (no plant count or per-phase day-count given, only phase-change events and a 125-day total operating period); Plant height (only initial cutting height at planting, 22 +/- 4.2 cm, is given — this is not the schema’s harvest-height definition, so recorded NR with the initial value noted here instead); Water recycle is the one exception where a genuine unit conversion (not derivation) was applied — see UNIT CONVERSION ONLY note below. | UNIT CONVERSION ONLY: Water recycle 10 L/day (Methods 2.2, ‘constant flow rate of 10 L/day, equivalent to 1 day of hydraulic retention time’) -> 0.0069 L/min (10/1440); Total Feed g->kg (Table 2, dry weight); Fish biomass created g->kg (Table 2, wet weight). None of these change the value’s information content, only its unit. | NO COLUMN: Table 3 nitrogen-transformation rates by phase (mgN/day) — Nitrogen removal rate / Uptake rate by plant / Denitrification rate: Phase I (P/F 0.06) 4.50+/-0.50 / 2.55+/-0.30 / 1.55; Phase II (0.30) 22.40+/-2.58 / 13.20+/-1.52 / 8.85; Phase III (0.50) 34.50+/-4.20 / 20.80+/-2.20 / 13.00; Phase IV (0.95) 56.00+/-5.24 / 34.50+/-3.60 / 20.50. Table 4 N2O emission data by phase (mg/day emission rate / g N2O-N per kg fish emission factor / %N2O-N conversion): I 1.53+/-0.11 / 0.56+/-0.12 / 0.56; II 1.48+/-0.15 / 0.55+/-0.21 / 0.54; III 1.41+/-0.12 / 0.53+/-0.16 / 0.51; IV 1.30+/-0.14 / 0.50+/-0.10 / 0.47; Control 1.56+/-0.15 / 0.68+/-0.17 / 0.57 (Table 4). Nitrogen budget (Fig. 4, % of N input, read from stacked-bar figure with accompanying text percentages, not purely a figure-only read since Results 4.4 states the same ranges in text): fish biomass recovery 26.5-28.4%; total NUE (fish+plant assimilation) 29.3-39.1% as P/F increased 0.06->0.95; water nitrogen retention fell 50.5%->27.8%; total gaseous N loss (N2O+N2) rose 17.75%->30.16% as P/F increased. Functional gene analysis (Section 4.5, Fig. 5, sampled at end of Phase III only): ratio of amoA to 16S rDNA genes ~10% in biofilter vs ~0.001% in water celery root mat, interpreted as nitrification dominant in the biofilter and denitrification more prevalent on root surfaces — no schema column exists for functional-gene/qPCR data. Root biomass (g, Table 2): I 2.2+/-0.2; II 9.4+/-0.03; III 14.1+/-0.2; IV 23.0+/-0.2. Specific N uptake by plant tissue, 2.41-2.98 mg N/g plant (text, p.4), stated as an aggregate range across all phases (not broken out per-phase) and compared against a secondary/external figure of 7.8 mg N/g from Santamaria et al. 1999 (batch experiment, different paper) — excluded from plant_measurements.csv because (a) no per-phase/per-trial value is given, only a whole-experiment range, and (b) it reads as an uptake RATE analogous to Table 3’s uptake-rate column rather than a static tissue-composition percentage, so it does not cleanly fit the ‘mineral’ category’s intended use (see Extraction notes). Feed composition beyond crude protein: 2% lipids, 1% calcium, 5% fibre, 1.8% phosphate, dry weight basis (Methods 2.2).

paudelNitrogenTransformationEngineered2020-T2

Fish

Field	Value
Fish	Koi carp (Cyprinus carpio)
FCR	1.75
Protein	32
Fish size initial	84.8 +/- 20.2
Feed routine	Once daily; pellets consumed within 10 min of feeding (Methods 2.2)
Feed regime	Commercial feed pellets (Woo Sung Feed Co. Ltd., South Korea), fed at a constant rate of 4.75 g/day; composition per manufacturer specification (registration CCBDN0055): 32% protein, 2% lipids, 1% calcium, 5% fibre, 1.8% phosphate, dry-weight basis (Methods 2.2)
Total Feed (kg)	0.0998
Fish biomass created (kg)	0.0570
Fish survival rate	100% (no fish mortality observed during the 125-day operating period, Section 4.1)
Fish trial duration (days)	125

Water

Field	Value
Water recycle	0.0069
Water volume in the system	200 (fish tank only, effective volume, 80x50x50 cm; system also includes a 4 L settler and 4 x 4 L floating-raft hydroponic beds = 16 L total bed volume; total combined system volume not stated as one figure, Methods 2.1)
Aq pH	7.0 +/- 0.1
Dissolved Oxigen	6.0 +/- 0.3
Water temperature	24.2 +/- 0.6

Plant

Field	Value
Plant	Water celery (Oenanthe javanica)
Details	Cut-stalks with 2-3 nodes (initial height 22 +/- 4.2 cm) planted into floating-raft hydroponic beds at differing masses to achieve the target P/F biomass ratio; replanted at each phase change (Methods 2.2)

System & Setup

Field	Value
System type	Floating raft (hydroponic bed with gravel + sponge root support, connected in series, Methods 2.1)
Media Details	Half of each hydroponic pot (180 mL, Sinnong, Korea) filled with gravel (5-15 mm diameter) and hydroponic urethane sponge (77 pieces) to hold plants upright; 4 floating-raft hydroponic beds connected in series, each 4 L effective volume (100x10x4 cm), covered with black sheet (Methods 2.1)
Biological system already in use	Y (Biofilter bag (100 sponge media, 1x1x1cm) pre-immobilized with activated sludge before system startup; biomass on media measured as 4.45 +/- 0.22 mg VSS/media, per method of Paudel et al. 2015 (Methods 2.1))
Air supplement	Y (Ceramic air diffuser connected to an air pump in the fish tank (Methods 2.1))
pH Buffers	Y (5 g sodium bicarbonate periodically applied to the aquaculture tank to maintain neutral pH (Methods 2.2, citing Boley et al. 2000))
Climate control	Y (Thermostat and chiller maintained system water temperature at 24 +/- 0.5 C; ambient room temperature was not controlled and ranged 20-24 C throughout (Methods 2.2))
Artificial Lighting	Y (Fluorescent light (ML-2, 12V, Youngi Electric, Korea) mounted 1.1 m above the hydroponic beds; intensity maintained 490-510 lux; controlled by a 24 h timer giving a 12 h light-on photoperiod (Methods 2.1))
Equipment	Submersible mixing pump (YI-50, Youngi Electric, Korea); ceramic air diffuser + air pump; biofilter bag with 100 sponge media (1x1x1cm) pre-immobilized with activated sludge; Thermo Scientific Orion 5 star meter (online DO/pH/temperature sensors, fish tank and hydroponic-bed outlet); thermostat and chiller (system water temperature control); fluorescent light (ML-2, 12V, Youngi Electric, Korea) on a 24 h timer; peristaltic pump (recirculation, 10 L/day); HACH reaction kit (TN, TAN); ion chromatography (ICS-900, Dionex, US) for NO2-/NO3-; analytical balance (Model GF-3000); gas chromatograph (DS 6200 DSS, Korea) with PDHI detector and Hayesep D packed column (N2O); Illumina Eco Real Time PCR System (qPCR, amoA/16S rDNA)
Control Parameters	DO target ~5.5 mg/L (fish tank, for optimal fish growth, Methods 2.2); pH maintained neutral via periodic 5 g sodium bicarbonate dosing; system water temperature 24 +/- 0.5 C via thermostat/chiller; room air temperature uncontrolled, 20-24 C ambient; light intensity 490-510 lux, 12 h light/12 h dark photoperiod; flow rate 10 L/day (1-day hydraulic retention time in hydroponic bed); feed rate fixed at 4.75 g/day
Combination	Koi carp (Cyprinus carpio) and water celery (Oenanthe javanica) in a single bench-scale recirculating floating-raft aquaponic system, operated sequentially through 4 increasing plant-to-fish (P/F) biomass ratio phases, compared against a fish-only (no plant) aquaculture control run in the same type of system

Site

Field	Value
Average room Temperature	20-24 (range only, no trial mean reported; ambient/uncontrolled room temperature, Methods 2.2)

Results & Statistics

Field	Value
Measured Unit	mg/L (TN, TAN, NO2-N, NO3-N in water); g (fish/plant biomass, dry wt for feed, wet wt for fish/plant biomass per Table 2 footnotes); mg N/day (nitrogen removal/uptake/denitrification rates, Table 3); mg/day and g N2O-N/kg fish (N2O emission rate and emission factor, Table 4); mg N/g plant (specific plant N uptake, text)
Statistic Details	One-way or two-way ANOVA followed by Tukey HSD (p<=0.05), SPSS version 16.0 (IBM Corporation, Somers, NY) (Section 3, Statistical analysis)
Statistically analysed	Y
Replicates (n)	1
AP	95.8 +/- 3.5

Experimental Remarks: TRIAL DEFINITION: T2 = aquaponic treatment at plant-to-fish (P/F) biomass ratio 0.30 (Phase II; see WARN-MATERIAL below on a conflicting 0.29 reading), run as one segment of a single continuously-operated bench-scale recirculating system (200 L Koi tank + 4 L settler + 4 x 4 L floating-raft water celery beds in series, zero water exchange except <=0.50% evapotranspiration replenishment every 2nd day, Methods 2.1-2.2). Water celery cut-stalks (2-3 nodes, initial height 22 +/- 4.2 cm) were replanted at each phase change to achieve the target P/F ratio. Paired control = a separate fish-only aquaculture run with no plants coupled (same tank type); this paper has no hydroponic-only (plant-only) control, so all HYD-labelled columns are NA. Control values: FCR 1.96; N2O emission rate see Table 4 (NO COLUMN) mg/day; N2O emission factor see Table 4 (NO COLUMN) g N2O-N/kg fish; %N2O-N conversion see Table 4 (NO COLUMN)% (Table 4) — recorded here in remarks as NO COLUMN since the schema has no fish-only-control field distinct from HYD. All four phases share ONE physical system and ONE continuous fish population/cohort (16 Koi, 84.8 +/- 20.2 g at initial stocking) run sequentially in fixed order I->II->III->IV (not randomized, not independently replicated) — see Extraction notes for the design/quasi-experiment judgment call. Table 1 water-quality figures used here take the hydroponic-bed (H-bed) compartment value per SCHEMA.md’s compartment rule: Aq pH 7.0 +/- 0.1 (fish-tank value 6.8 +/- 0.1 for reference); DO 6.0 +/- 0.3 mg/L (fish-tank 6.6 +/- 0.1 mg/L for reference); water temperature 24.2 +/- 0.6 C (fish-tank 24.4 +/- 0.2 C for reference), all Table 1, Mean +/- SD, n=21 water samples/phase (see Extraction notes on non-independence of this n). Plant: initial biomass 404.5 +/- 15.0 g (wet wt) planted at phase start; harvested biomass 95.8 +/- 3.5 g (wet wt, whole 4-bed system total, NOT per-plant or per-m2 — no plant count or bed area given for a per-plant/per-m2 conversion without derivation); plant yield 23.7 +/- 0.8% (= harvested/initial biomass x100, paper’s own stated %, Table 2, independently verified to match the underlying Table 2 masses to within rounding — recomputation used only to check the paper’s own arithmetic, not entered as a derived cell per SCHEMA.md); root biomass 9.4 +/- 0.03 g (Table 2, NO COLUMN, no dedicated field for root mass exists in the schema). AP column above holds the harvested-biomass figure with its Measured Unit noted as a whole-system total, not per-plant/per-m2, since no per-plant fresh-weight column value could be computed without derivation. | WARN-MATERIAL Phase II P/F biomass ratio label: Methods 2.2 (‘0.06, 0.30, 0.50, and 0.95’), Table 1, Table 2, and Table 3 all state Phase II’s nominal plant-to-fish biomass ratio as 0.30. Table 4 and the Fig. 2 caption instead state 0.29 for the same phase (‘Phase I, II, III, & IV represents different plant fish biomass ratio of 0.06, 0.29, 0.49, and 0.95 respectively’). Four independent locations (Methods text + 3 tables) agree on 0.30 against two (Table 4 + one figure caption) giving 0.29; Methods 2.2 is also the paper’s primary definitional statement of the intended treatment levels. 0.30 recorded as this trial’s defining ratio; 0.29 is preserved here as the minority reading. Affects: which nominal ratio this row is filed under; does not affect any measured value (FCR, biomass, water-quality figures for Phase II are internally consistent regardless of which ratio label is used). | Replicates (n): recorded as 1 (one physical bench-scale system, no independent replicate units of any treatment/phase). Table 1’s caption states ‘sample size (n) = 21’, which is the count of repeated water-quality samples taken every 2nd day WITHIN each phase from the single system, not independent experimental units; ANOVA + Tukey HSD (Section 3) is applied across phases using these 21-per-phase repeated samples, which is temporal pseudo-replication rather than the schema’s ‘experimental units per treatment’ definition. This design property, not a numeric contradiction, is the basis for classifying the paper as quasi-experiment rather than experiment (see Extraction notes) and is flagged here for anyone using n=21 downstream as if it were independent replication. | Fish trial duration recorded as the paper’s one explicitly stated overall figure, 125 days (‘the operating period (125 days)’, Section 4.1; ‘the aquaponic system was continuously operated for 125 days by applying four different P/F biomass ratios together with control’, Methods 2.2) applied identically to all four rows, since all four phases are segments of one continuous run and the paper never states how many of the 125 days belong to each individual phase. A separate 2-week acclimation period is also mentioned before the hydroponic beds were first coupled; Methods 2.2’s wording ties the ‘125 days’ figure to the phases+control period specifically (after coupling), while Fig. 2’s 0-125 day x-axis appears to place the acclimation phase INSIDE the same 0-125 day window. This is a minor internal ambiguity about whether acclimation is included in or additional to the 125 days; it does not change the recorded cell value (125, the only number the paper gives) and is noted here for transparency rather than tagged as a formal contradiction, since no second numeric candidate is offered anywhere in the paper. | NOT DERIVED, left NR: Initial Stock density (16 Koi fish, 84.8 +/- 20.2 g each, in a 200 L tank are all stated, but no kg/m3 figure is given; computing 16x84.8g/0.2m3 would be derivation); % of body weight (feed fixed at 4.75 g/day, fish biomass stated, but no daily-ration-as-%-of-body-weight figure given); Fish size final (only aggregate tank biomass increase per phase given, Table 2, not a final mean individual weight); Fish weight gain per fish (only the population-level aggregate biomass increase is stated, recorded instead under Fish biomass created); N, P, K feed composition beyond crude protein (Methods 2.2 gives protein/lipid/calcium/fibre/phosphate %, not elemental N/P/K in the schema’s expected basis — ‘phosphate’ is not necessarily elemental P and converting it would assume a basis not stated, so left NR, full figures under NO COLUMN); Daily Water exchange rate (paper states ’<=0.50% on every 2nd day’, not a stated daily %; halving it to a daily-equivalent rate would assume linear averaging not stated by the authors, so left NR rather than derived); Plants/m2, Days Plant after transplant (no plant count or per-phase day-count given, only phase-change events and a 125-day total operating period); Plant height (only initial cutting height at planting, 22 +/- 4.2 cm, is given — this is not the schema’s harvest-height definition, so recorded NR with the initial value noted here instead); Water recycle is the one exception where a genuine unit conversion (not derivation) was applied — see UNIT CONVERSION ONLY note below. | UNIT CONVERSION ONLY: Water recycle 10 L/day (Methods 2.2, ‘constant flow rate of 10 L/day, equivalent to 1 day of hydraulic retention time’) -> 0.0069 L/min (10/1440); Total Feed g->kg (Table 2, dry weight); Fish biomass created g->kg (Table 2, wet weight). None of these change the value’s information content, only its unit. | NO COLUMN: Table 3 nitrogen-transformation rates by phase (mgN/day) — Nitrogen removal rate / Uptake rate by plant / Denitrification rate: Phase I (P/F 0.06) 4.50+/-0.50 / 2.55+/-0.30 / 1.55; Phase II (0.30) 22.40+/-2.58 / 13.20+/-1.52 / 8.85; Phase III (0.50) 34.50+/-4.20 / 20.80+/-2.20 / 13.00; Phase IV (0.95) 56.00+/-5.24 / 34.50+/-3.60 / 20.50. Table 4 N2O emission data by phase (mg/day emission rate / g N2O-N per kg fish emission factor / %N2O-N conversion): I 1.53+/-0.11 / 0.56+/-0.12 / 0.56; II 1.48+/-0.15 / 0.55+/-0.21 / 0.54; III 1.41+/-0.12 / 0.53+/-0.16 / 0.51; IV 1.30+/-0.14 / 0.50+/-0.10 / 0.47; Control 1.56+/-0.15 / 0.68+/-0.17 / 0.57 (Table 4). Nitrogen budget (Fig. 4, % of N input, read from stacked-bar figure with accompanying text percentages, not purely a figure-only read since Results 4.4 states the same ranges in text): fish biomass recovery 26.5-28.4%; total NUE (fish+plant assimilation) 29.3-39.1% as P/F increased 0.06->0.95; water nitrogen retention fell 50.5%->27.8%; total gaseous N loss (N2O+N2) rose 17.75%->30.16% as P/F increased. Functional gene analysis (Section 4.5, Fig. 5, sampled at end of Phase III only): ratio of amoA to 16S rDNA genes ~10% in biofilter vs ~0.001% in water celery root mat, interpreted as nitrification dominant in the biofilter and denitrification more prevalent on root surfaces — no schema column exists for functional-gene/qPCR data. Root biomass (g, Table 2): I 2.2+/-0.2; II 9.4+/-0.03; III 14.1+/-0.2; IV 23.0+/-0.2. Specific N uptake by plant tissue, 2.41-2.98 mg N/g plant (text, p.4), stated as an aggregate range across all phases (not broken out per-phase) and compared against a secondary/external figure of 7.8 mg N/g from Santamaria et al. 1999 (batch experiment, different paper) — excluded from plant_measurements.csv because (a) no per-phase/per-trial value is given, only a whole-experiment range, and (b) it reads as an uptake RATE analogous to Table 3’s uptake-rate column rather than a static tissue-composition percentage, so it does not cleanly fit the ‘mineral’ category’s intended use (see Extraction notes). Feed composition beyond crude protein: 2% lipids, 1% calcium, 5% fibre, 1.8% phosphate, dry weight basis (Methods 2.2).

paudelNitrogenTransformationEngineered2020-T3

Fish

Field	Value
Fish	Koi carp (Cyprinus carpio)
FCR	1.80
Protein	32
Fish size initial	84.8 +/- 20.2
Feed routine	Once daily; pellets consumed within 10 min of feeding (Methods 2.2)
Feed regime	Commercial feed pellets (Woo Sung Feed Co. Ltd., South Korea), fed at a constant rate of 4.75 g/day; composition per manufacturer specification (registration CCBDN0055): 32% protein, 2% lipids, 1% calcium, 5% fibre, 1.8% phosphate, dry-weight basis (Methods 2.2)
Total Feed (kg)	0.0998
Fish biomass created (kg)	0.0554
Fish survival rate	100% (no fish mortality observed during the 125-day operating period, Section 4.1)
Fish trial duration (days)	125

Water

Field	Value
Water recycle	0.0069
Water volume in the system	200 (fish tank only, effective volume, 80x50x50 cm; system also includes a 4 L settler and 4 x 4 L floating-raft hydroponic beds = 16 L total bed volume; total combined system volume not stated as one figure, Methods 2.1)
Aq pH	7.1 +/- 0.1
Dissolved Oxigen	5.8 +/- 0.2
Water temperature	23.8 +/- 0.7

Plant

Field	Value
Plant	Water celery (Oenanthe javanica)
Details	Cut-stalks with 2-3 nodes (initial height 22 +/- 4.2 cm) planted into floating-raft hydroponic beds at differing masses to achieve the target P/F biomass ratio; replanted at each phase change (Methods 2.2)

System & Setup

Field	Value
System type	Floating raft (hydroponic bed with gravel + sponge root support, connected in series, Methods 2.1)
Media Details	Half of each hydroponic pot (180 mL, Sinnong, Korea) filled with gravel (5-15 mm diameter) and hydroponic urethane sponge (77 pieces) to hold plants upright; 4 floating-raft hydroponic beds connected in series, each 4 L effective volume (100x10x4 cm), covered with black sheet (Methods 2.1)
Biological system already in use	Y (Biofilter bag (100 sponge media, 1x1x1cm) pre-immobilized with activated sludge before system startup; biomass on media measured as 4.45 +/- 0.22 mg VSS/media, per method of Paudel et al. 2015 (Methods 2.1))
Air supplement	Y (Ceramic air diffuser connected to an air pump in the fish tank (Methods 2.1))
pH Buffers	Y (5 g sodium bicarbonate periodically applied to the aquaculture tank to maintain neutral pH (Methods 2.2, citing Boley et al. 2000))
Climate control	Y (Thermostat and chiller maintained system water temperature at 24 +/- 0.5 C; ambient room temperature was not controlled and ranged 20-24 C throughout (Methods 2.2))
Artificial Lighting	Y (Fluorescent light (ML-2, 12V, Youngi Electric, Korea) mounted 1.1 m above the hydroponic beds; intensity maintained 490-510 lux; controlled by a 24 h timer giving a 12 h light-on photoperiod (Methods 2.1))
Equipment	Submersible mixing pump (YI-50, Youngi Electric, Korea); ceramic air diffuser + air pump; biofilter bag with 100 sponge media (1x1x1cm) pre-immobilized with activated sludge; Thermo Scientific Orion 5 star meter (online DO/pH/temperature sensors, fish tank and hydroponic-bed outlet); thermostat and chiller (system water temperature control); fluorescent light (ML-2, 12V, Youngi Electric, Korea) on a 24 h timer; peristaltic pump (recirculation, 10 L/day); HACH reaction kit (TN, TAN); ion chromatography (ICS-900, Dionex, US) for NO2-/NO3-; analytical balance (Model GF-3000); gas chromatograph (DS 6200 DSS, Korea) with PDHI detector and Hayesep D packed column (N2O); Illumina Eco Real Time PCR System (qPCR, amoA/16S rDNA)
Control Parameters	DO target ~5.5 mg/L (fish tank, for optimal fish growth, Methods 2.2); pH maintained neutral via periodic 5 g sodium bicarbonate dosing; system water temperature 24 +/- 0.5 C via thermostat/chiller; room air temperature uncontrolled, 20-24 C ambient; light intensity 490-510 lux, 12 h light/12 h dark photoperiod; flow rate 10 L/day (1-day hydraulic retention time in hydroponic bed); feed rate fixed at 4.75 g/day
Combination	Koi carp (Cyprinus carpio) and water celery (Oenanthe javanica) in a single bench-scale recirculating floating-raft aquaponic system, operated sequentially through 4 increasing plant-to-fish (P/F) biomass ratio phases, compared against a fish-only (no plant) aquaculture control run in the same type of system

Site

Field	Value
Average room Temperature	20-24 (range only, no trial mean reported; ambient/uncontrolled room temperature, Methods 2.2)

Results & Statistics

Field	Value
Measured Unit	mg/L (TN, TAN, NO2-N, NO3-N in water); g (fish/plant biomass, dry wt for feed, wet wt for fish/plant biomass per Table 2 footnotes); mg N/day (nitrogen removal/uptake/denitrification rates, Table 3); mg/day and g N2O-N/kg fish (N2O emission rate and emission factor, Table 4); mg N/g plant (specific plant N uptake, text)
Statistic Details	One-way or two-way ANOVA followed by Tukey HSD (p<=0.05), SPSS version 16.0 (IBM Corporation, Somers, NY) (Section 3, Statistical analysis)
Statistically analysed	Y
Replicates (n)	1
AP	162.5 +/- 4.3

Experimental Remarks: TRIAL DEFINITION: T3 = aquaponic treatment at plant-to-fish (P/F) biomass ratio 0.50 (Phase III; see WARN-MATERIAL below on a conflicting 0.49 reading), run as one segment of a single continuously-operated bench-scale recirculating system (200 L Koi tank + 4 L settler + 4 x 4 L floating-raft water celery beds in series, zero water exchange except <=0.50% evapotranspiration replenishment every 2nd day, Methods 2.1-2.2). Water celery cut-stalks (2-3 nodes, initial height 22 +/- 4.2 cm) were replanted at each phase change to achieve the target P/F ratio. Paired control = a separate fish-only aquaculture run with no plants coupled (same tank type); this paper has no hydroponic-only (plant-only) control, so all HYD-labelled columns are NA. Control values: FCR 1.96; N2O emission rate see Table 4 (NO COLUMN) mg/day; N2O emission factor see Table 4 (NO COLUMN) g N2O-N/kg fish; %N2O-N conversion see Table 4 (NO COLUMN)% (Table 4) — recorded here in remarks as NO COLUMN since the schema has no fish-only-control field distinct from HYD. All four phases share ONE physical system and ONE continuous fish population/cohort (16 Koi, 84.8 +/- 20.2 g at initial stocking) run sequentially in fixed order I->II->III->IV (not randomized, not independently replicated) — see Extraction notes for the design/quasi-experiment judgment call. Table 1 water-quality figures used here take the hydroponic-bed (H-bed) compartment value per SCHEMA.md’s compartment rule: Aq pH 7.1 +/- 0.1 (fish-tank value 6.7 +/- 0.1 for reference); DO 5.8 +/- 0.2 mg/L (fish-tank 6.3 +/- 0.1 mg/L for reference); water temperature 23.8 +/- 0.7 C (fish-tank 24.2 +/- 0.2 C for reference), all Table 1, Mean +/- SD, n=21 water samples/phase (see Extraction notes on non-independence of this n). Plant: initial biomass 724.0 +/- 25.0 g (wet wt) planted at phase start; harvested biomass 162.5 +/- 4.3 g (wet wt, whole 4-bed system total, NOT per-plant or per-m2 — no plant count or bed area given for a per-plant/per-m2 conversion without derivation); plant yield 22.4 +/- 1.4% (= harvested/initial biomass x100, paper’s own stated %, Table 2, independently verified to match the underlying Table 2 masses to within rounding — recomputation used only to check the paper’s own arithmetic, not entered as a derived cell per SCHEMA.md); root biomass 14.1 +/- 0.2 g (Table 2, NO COLUMN, no dedicated field for root mass exists in the schema). AP column above holds the harvested-biomass figure with its Measured Unit noted as a whole-system total, not per-plant/per-m2, since no per-plant fresh-weight column value could be computed without derivation. | WARN-MATERIAL Phase III P/F biomass ratio label: Methods 2.2, Table 1, Table 2, and Table 3 all state Phase III’s nominal plant-to-fish biomass ratio as 0.50. Table 4 and the Fig. 2 caption instead state 0.49 for the same phase. Same majority-source logic as Phase II (see T2 remarks): 0.50 recorded as this trial’s defining ratio, 0.49 preserved as the minority reading. Does not affect any measured value for Phase III. | Replicates (n): recorded as 1 (one physical bench-scale system, no independent replicate units of any treatment/phase). Table 1’s caption states ‘sample size (n) = 21’, which is the count of repeated water-quality samples taken every 2nd day WITHIN each phase from the single system, not independent experimental units; ANOVA + Tukey HSD (Section 3) is applied across phases using these 21-per-phase repeated samples, which is temporal pseudo-replication rather than the schema’s ‘experimental units per treatment’ definition. This design property, not a numeric contradiction, is the basis for classifying the paper as quasi-experiment rather than experiment (see Extraction notes) and is flagged here for anyone using n=21 downstream as if it were independent replication. | Fish trial duration recorded as the paper’s one explicitly stated overall figure, 125 days (‘the operating period (125 days)’, Section 4.1; ‘the aquaponic system was continuously operated for 125 days by applying four different P/F biomass ratios together with control’, Methods 2.2) applied identically to all four rows, since all four phases are segments of one continuous run and the paper never states how many of the 125 days belong to each individual phase. A separate 2-week acclimation period is also mentioned before the hydroponic beds were first coupled; Methods 2.2’s wording ties the ‘125 days’ figure to the phases+control period specifically (after coupling), while Fig. 2’s 0-125 day x-axis appears to place the acclimation phase INSIDE the same 0-125 day window. This is a minor internal ambiguity about whether acclimation is included in or additional to the 125 days; it does not change the recorded cell value (125, the only number the paper gives) and is noted here for transparency rather than tagged as a formal contradiction, since no second numeric candidate is offered anywhere in the paper. | NOT DERIVED, left NR: Initial Stock density (16 Koi fish, 84.8 +/- 20.2 g each, in a 200 L tank are all stated, but no kg/m3 figure is given; computing 16x84.8g/0.2m3 would be derivation); % of body weight (feed fixed at 4.75 g/day, fish biomass stated, but no daily-ration-as-%-of-body-weight figure given); Fish size final (only aggregate tank biomass increase per phase given, Table 2, not a final mean individual weight); Fish weight gain per fish (only the population-level aggregate biomass increase is stated, recorded instead under Fish biomass created); N, P, K feed composition beyond crude protein (Methods 2.2 gives protein/lipid/calcium/fibre/phosphate %, not elemental N/P/K in the schema’s expected basis — ‘phosphate’ is not necessarily elemental P and converting it would assume a basis not stated, so left NR, full figures under NO COLUMN); Daily Water exchange rate (paper states ’<=0.50% on every 2nd day’, not a stated daily %; halving it to a daily-equivalent rate would assume linear averaging not stated by the authors, so left NR rather than derived); Plants/m2, Days Plant after transplant (no plant count or per-phase day-count given, only phase-change events and a 125-day total operating period); Plant height (only initial cutting height at planting, 22 +/- 4.2 cm, is given — this is not the schema’s harvest-height definition, so recorded NR with the initial value noted here instead); Water recycle is the one exception where a genuine unit conversion (not derivation) was applied — see UNIT CONVERSION ONLY note below. | UNIT CONVERSION ONLY: Water recycle 10 L/day (Methods 2.2, ‘constant flow rate of 10 L/day, equivalent to 1 day of hydraulic retention time’) -> 0.0069 L/min (10/1440); Total Feed g->kg (Table 2, dry weight); Fish biomass created g->kg (Table 2, wet weight). None of these change the value’s information content, only its unit. | NO COLUMN: Table 3 nitrogen-transformation rates by phase (mgN/day) — Nitrogen removal rate / Uptake rate by plant / Denitrification rate: Phase I (P/F 0.06) 4.50+/-0.50 / 2.55+/-0.30 / 1.55; Phase II (0.30) 22.40+/-2.58 / 13.20+/-1.52 / 8.85; Phase III (0.50) 34.50+/-4.20 / 20.80+/-2.20 / 13.00; Phase IV (0.95) 56.00+/-5.24 / 34.50+/-3.60 / 20.50. Table 4 N2O emission data by phase (mg/day emission rate / g N2O-N per kg fish emission factor / %N2O-N conversion): I 1.53+/-0.11 / 0.56+/-0.12 / 0.56; II 1.48+/-0.15 / 0.55+/-0.21 / 0.54; III 1.41+/-0.12 / 0.53+/-0.16 / 0.51; IV 1.30+/-0.14 / 0.50+/-0.10 / 0.47; Control 1.56+/-0.15 / 0.68+/-0.17 / 0.57 (Table 4). Nitrogen budget (Fig. 4, % of N input, read from stacked-bar figure with accompanying text percentages, not purely a figure-only read since Results 4.4 states the same ranges in text): fish biomass recovery 26.5-28.4%; total NUE (fish+plant assimilation) 29.3-39.1% as P/F increased 0.06->0.95; water nitrogen retention fell 50.5%->27.8%; total gaseous N loss (N2O+N2) rose 17.75%->30.16% as P/F increased. Functional gene analysis (Section 4.5, Fig. 5, sampled at end of Phase III only): ratio of amoA to 16S rDNA genes ~10% in biofilter vs ~0.001% in water celery root mat, interpreted as nitrification dominant in the biofilter and denitrification more prevalent on root surfaces — no schema column exists for functional-gene/qPCR data. Root biomass (g, Table 2): I 2.2+/-0.2; II 9.4+/-0.03; III 14.1+/-0.2; IV 23.0+/-0.2. Specific N uptake by plant tissue, 2.41-2.98 mg N/g plant (text, p.4), stated as an aggregate range across all phases (not broken out per-phase) and compared against a secondary/external figure of 7.8 mg N/g from Santamaria et al. 1999 (batch experiment, different paper) — excluded from plant_measurements.csv because (a) no per-phase/per-trial value is given, only a whole-experiment range, and (b) it reads as an uptake RATE analogous to Table 3’s uptake-rate column rather than a static tissue-composition percentage, so it does not cleanly fit the ‘mineral’ category’s intended use (see Extraction notes). Feed composition beyond crude protein: 2% lipids, 1% calcium, 5% fibre, 1.8% phosphate, dry weight basis (Methods 2.2).

paudelNitrogenTransformationEngineered2020-T4

Fish

Field	Value
Fish	Koi carp (Cyprinus carpio)
FCR	1.85
Protein	32
Fish size initial	84.8 +/- 20.2
Feed routine	Once daily; pellets consumed within 10 min of feeding (Methods 2.2)
Feed regime	Commercial feed pellets (Woo Sung Feed Co. Ltd., South Korea), fed at a constant rate of 4.75 g/day; composition per manufacturer specification (registration CCBDN0055): 32% protein, 2% lipids, 1% calcium, 5% fibre, 1.8% phosphate, dry-weight basis (Methods 2.2)
Total Feed (kg)	0.1045
Fish biomass created (kg)	0.0565
Fish survival rate	100% (no fish mortality observed during the 125-day operating period, Section 4.1)
Fish trial duration (days)	125

Water

Field	Value
Water recycle	0.0069
Water volume in the system	200 (fish tank only, effective volume, 80x50x50 cm; system also includes a 4 L settler and 4 x 4 L floating-raft hydroponic beds = 16 L total bed volume; total combined system volume not stated as one figure, Methods 2.1)
Aq pH	7.2 +/- 0.1
Dissolved Oxigen	5.7 +/- 0.3
Water temperature	24.0 +/- 0.3
NO3-N	62.5 +/- 4.2 (fish-tank compartment; abstract-stated leveling-off value at Phase IV; H-bed value only available via Fig. 2c time-series and not extracted per never-read-a-figure rule; compartment differs from SCHEMA.md’s plant-bed preference by necessity, since no H-bed NO3-N summary value is stated anywhere)

Plant

Field	Value
Plant	Water celery (Oenanthe javanica)
Details	Cut-stalks with 2-3 nodes (initial height 22 +/- 4.2 cm) planted into floating-raft hydroponic beds at differing masses to achieve the target P/F biomass ratio; replanted at each phase change (Methods 2.2)

System & Setup

Field	Value
System type	Floating raft (hydroponic bed with gravel + sponge root support, connected in series, Methods 2.1)
Media Details	Half of each hydroponic pot (180 mL, Sinnong, Korea) filled with gravel (5-15 mm diameter) and hydroponic urethane sponge (77 pieces) to hold plants upright; 4 floating-raft hydroponic beds connected in series, each 4 L effective volume (100x10x4 cm), covered with black sheet (Methods 2.1)
Biological system already in use	Y (Biofilter bag (100 sponge media, 1x1x1cm) pre-immobilized with activated sludge before system startup; biomass on media measured as 4.45 +/- 0.22 mg VSS/media, per method of Paudel et al. 2015 (Methods 2.1))
Air supplement	Y (Ceramic air diffuser connected to an air pump in the fish tank (Methods 2.1))
pH Buffers	Y (5 g sodium bicarbonate periodically applied to the aquaculture tank to maintain neutral pH (Methods 2.2, citing Boley et al. 2000))
Climate control	Y (Thermostat and chiller maintained system water temperature at 24 +/- 0.5 C; ambient room temperature was not controlled and ranged 20-24 C throughout (Methods 2.2))
Artificial Lighting	Y (Fluorescent light (ML-2, 12V, Youngi Electric, Korea) mounted 1.1 m above the hydroponic beds; intensity maintained 490-510 lux; controlled by a 24 h timer giving a 12 h light-on photoperiod (Methods 2.1))
Equipment	Submersible mixing pump (YI-50, Youngi Electric, Korea); ceramic air diffuser + air pump; biofilter bag with 100 sponge media (1x1x1cm) pre-immobilized with activated sludge; Thermo Scientific Orion 5 star meter (online DO/pH/temperature sensors, fish tank and hydroponic-bed outlet); thermostat and chiller (system water temperature control); fluorescent light (ML-2, 12V, Youngi Electric, Korea) on a 24 h timer; peristaltic pump (recirculation, 10 L/day); HACH reaction kit (TN, TAN); ion chromatography (ICS-900, Dionex, US) for NO2-/NO3-; analytical balance (Model GF-3000); gas chromatograph (DS 6200 DSS, Korea) with PDHI detector and Hayesep D packed column (N2O); Illumina Eco Real Time PCR System (qPCR, amoA/16S rDNA)
Control Parameters	DO target ~5.5 mg/L (fish tank, for optimal fish growth, Methods 2.2); pH maintained neutral via periodic 5 g sodium bicarbonate dosing; system water temperature 24 +/- 0.5 C via thermostat/chiller; room air temperature uncontrolled, 20-24 C ambient; light intensity 490-510 lux, 12 h light/12 h dark photoperiod; flow rate 10 L/day (1-day hydraulic retention time in hydroponic bed); feed rate fixed at 4.75 g/day
Combination	Koi carp (Cyprinus carpio) and water celery (Oenanthe javanica) in a single bench-scale recirculating floating-raft aquaponic system, operated sequentially through 4 increasing plant-to-fish (P/F) biomass ratio phases, compared against a fish-only (no plant) aquaculture control run in the same type of system

Site

Field	Value
Average room Temperature	20-24 (range only, no trial mean reported; ambient/uncontrolled room temperature, Methods 2.2)

Results & Statistics

Field	Value
Measured Unit	mg/L (TN, TAN, NO2-N, NO3-N in water); g (fish/plant biomass, dry wt for feed, wet wt for fish/plant biomass per Table 2 footnotes); mg N/day (nitrogen removal/uptake/denitrification rates, Table 3); mg/day and g N2O-N/kg fish (N2O emission rate and emission factor, Table 4); mg N/g plant (specific plant N uptake, text)
Statistic Details	One-way or two-way ANOVA followed by Tukey HSD (p<=0.05), SPSS version 16.0 (IBM Corporation, Somers, NY) (Section 3, Statistical analysis)
Statistically analysed	Y
Replicates (n)	1
AP	310 +/- 6.3

Experimental Remarks: TRIAL DEFINITION: T4 = aquaponic treatment at plant-to-fish (P/F) biomass ratio 0.95 (Phase IV), run as one segment of a single continuously-operated bench-scale recirculating system (200 L Koi tank + 4 L settler + 4 x 4 L floating-raft water celery beds in series, zero water exchange except <=0.50% evapotranspiration replenishment every 2nd day, Methods 2.1-2.2). Water celery cut-stalks (2-3 nodes, initial height 22 +/- 4.2 cm) were replanted at each phase change to achieve the target P/F ratio. Paired control = a separate fish-only aquaculture run with no plants coupled (same tank type); this paper has no hydroponic-only (plant-only) control, so all HYD-labelled columns are NA. Control values: FCR 1.96; N2O emission rate see Table 4 (NO COLUMN) mg/day; N2O emission factor see Table 4 (NO COLUMN) g N2O-N/kg fish; %N2O-N conversion see Table 4 (NO COLUMN)% (Table 4) — recorded here in remarks as NO COLUMN since the schema has no fish-only-control field distinct from HYD. All four phases share ONE physical system and ONE continuous fish population/cohort (16 Koi, 84.8 +/- 20.2 g at initial stocking) run sequentially in fixed order I->II->III->IV (not randomized, not independently replicated) — see Extraction notes for the design/quasi-experiment judgment call. Table 1 water-quality figures used here take the hydroponic-bed (H-bed) compartment value per SCHEMA.md’s compartment rule: Aq pH 7.2 +/- 0.1 (fish-tank value 6.6 +/- 0.1 for reference); DO 5.7 +/- 0.3 mg/L (fish-tank 6.2 +/- 0.2 mg/L for reference); water temperature 24.0 +/- 0.3 C (fish-tank 24.3 +/- 0.3 C for reference), all Table 1, Mean +/- SD, n=21 water samples/phase (see Extraction notes on non-independence of this n). Plant: initial biomass 1450.0 +/- 35.0 g (wet wt) planted at phase start; harvested biomass 310 +/- 6.3 g (wet wt, whole 4-bed system total, NOT per-plant or per-m2 — no plant count or bed area given for a per-plant/per-m2 conversion without derivation); plant yield 21.4 +/- 1.8% (= harvested/initial biomass x100, paper’s own stated %, Table 2, independently verified to match the underlying Table 2 masses to within rounding — recomputation used only to check the paper’s own arithmetic, not entered as a derived cell per SCHEMA.md); root biomass 23.0 +/- 0.2 g (Table 2, NO COLUMN, no dedicated field for root mass exists in the schema). AP column above holds the harvested-biomass figure with its Measured Unit noted as a whole-system total, not per-plant/per-m2, since no per-plant fresh-weight column value could be computed without derivation. | WARN paper-level (no trials.csv cell affected, no dedicated column exists for this metric): Abstract and Results 4.3 both state the N2O reduction at the highest P/F ratio as ‘17%’, and Results 4.3 explicitly names the metric: ‘N2O emission factor (g N2O/kg fish) decreased by 17% when the highest biomass ratio employed’. Recomputing from Table 4’s own Emission factor column (Control 0.68 -> Phase IV 0.50 g N2O-N/kg fish) gives a 26.5% reduction, not 17%. The Nitrous oxide emission RATE column (Control 1.56 -> Phase IV 1.30 mg/day) gives 16.7%, which rounds to the stated 17%. The text’s named metric (emission factor) and its own stated percentage (17%) are therefore inconsistent with each other; the 17% figure appears to actually describe the emission-RATE column while being verbally attached to the emission-FACTOR column. Recorded here as evidence only, per SCHEMA.md’s allowance to recompute a value to check it without entering the recomputed number in a cell; no trials.csv column exists for either N2O metric (see NO COLUMN) so no cell is affected, but this bears directly on how the paper’s own headline ‘17% reduction’ claim should be read if cited. | Replicates (n): recorded as 1 (one physical bench-scale system, no independent replicate units of any treatment/phase). Table 1’s caption states ‘sample size (n) = 21’, which is the count of repeated water-quality samples taken every 2nd day WITHIN each phase from the single system, not independent experimental units; ANOVA + Tukey HSD (Section 3) is applied across phases using these 21-per-phase repeated samples, which is temporal pseudo-replication rather than the schema’s ‘experimental units per treatment’ definition. This design property, not a numeric contradiction, is the basis for classifying the paper as quasi-experiment rather than experiment (see Extraction notes) and is flagged here for anyone using n=21 downstream as if it were independent replication. | Fish trial duration recorded as the paper’s one explicitly stated overall figure, 125 days (‘the operating period (125 days)’, Section 4.1; ‘the aquaponic system was continuously operated for 125 days by applying four different P/F biomass ratios together with control’, Methods 2.2) applied identically to all four rows, since all four phases are segments of one continuous run and the paper never states how many of the 125 days belong to each individual phase. A separate 2-week acclimation period is also mentioned before the hydroponic beds were first coupled; Methods 2.2’s wording ties the ‘125 days’ figure to the phases+control period specifically (after coupling), while Fig. 2’s 0-125 day x-axis appears to place the acclimation phase INSIDE the same 0-125 day window. This is a minor internal ambiguity about whether acclimation is included in or additional to the 125 days; it does not change the recorded cell value (125, the only number the paper gives) and is noted here for transparency rather than tagged as a formal contradiction, since no second numeric candidate is offered anywhere in the paper. | NOT DERIVED, left NR: Initial Stock density (16 Koi fish, 84.8 +/- 20.2 g each, in a 200 L tank are all stated, but no kg/m3 figure is given; computing 16x84.8g/0.2m3 would be derivation); % of body weight (feed fixed at 4.75 g/day, fish biomass stated, but no daily-ration-as-%-of-body-weight figure given); Fish size final (only aggregate tank biomass increase per phase given, Table 2, not a final mean individual weight); Fish weight gain per fish (only the population-level aggregate biomass increase is stated, recorded instead under Fish biomass created); N, P, K feed composition beyond crude protein (Methods 2.2 gives protein/lipid/calcium/fibre/phosphate %, not elemental N/P/K in the schema’s expected basis — ‘phosphate’ is not necessarily elemental P and converting it would assume a basis not stated, so left NR, full figures under NO COLUMN); Daily Water exchange rate (paper states ’<=0.50% on every 2nd day’, not a stated daily %; halving it to a daily-equivalent rate would assume linear averaging not stated by the authors, so left NR rather than derived); Plants/m2, Days Plant after transplant (no plant count or per-phase day-count given, only phase-change events and a 125-day total operating period); Plant height (only initial cutting height at planting, 22 +/- 4.2 cm, is given — this is not the schema’s harvest-height definition, so recorded NR with the initial value noted here instead); Water recycle is the one exception where a genuine unit conversion (not derivation) was applied — see UNIT CONVERSION ONLY note below. | UNIT CONVERSION ONLY: Water recycle 10 L/day (Methods 2.2, ‘constant flow rate of 10 L/day, equivalent to 1 day of hydraulic retention time’) -> 0.0069 L/min (10/1440); Total Feed g->kg (Table 2, dry weight); Fish biomass created g->kg (Table 2, wet weight). None of these change the value’s information content, only its unit. | NO COLUMN: Table 3 nitrogen-transformation rates by phase (mgN/day) — Nitrogen removal rate / Uptake rate by plant / Denitrification rate: Phase I (P/F 0.06) 4.50+/-0.50 / 2.55+/-0.30 / 1.55; Phase II (0.30) 22.40+/-2.58 / 13.20+/-1.52 / 8.85; Phase III (0.50) 34.50+/-4.20 / 20.80+/-2.20 / 13.00; Phase IV (0.95) 56.00+/-5.24 / 34.50+/-3.60 / 20.50. Table 4 N2O emission data by phase (mg/day emission rate / g N2O-N per kg fish emission factor / %N2O-N conversion): I 1.53+/-0.11 / 0.56+/-0.12 / 0.56; II 1.48+/-0.15 / 0.55+/-0.21 / 0.54; III 1.41+/-0.12 / 0.53+/-0.16 / 0.51; IV 1.30+/-0.14 / 0.50+/-0.10 / 0.47; Control 1.56+/-0.15 / 0.68+/-0.17 / 0.57 (Table 4). Nitrogen budget (Fig. 4, % of N input, read from stacked-bar figure with accompanying text percentages, not purely a figure-only read since Results 4.4 states the same ranges in text): fish biomass recovery 26.5-28.4%; total NUE (fish+plant assimilation) 29.3-39.1% as P/F increased 0.06->0.95; water nitrogen retention fell 50.5%->27.8%; total gaseous N loss (N2O+N2) rose 17.75%->30.16% as P/F increased. Functional gene analysis (Section 4.5, Fig. 5, sampled at end of Phase III only): ratio of amoA to 16S rDNA genes ~10% in biofilter vs ~0.001% in water celery root mat, interpreted as nitrification dominant in the biofilter and denitrification more prevalent on root surfaces — no schema column exists for functional-gene/qPCR data. Root biomass (g, Table 2): I 2.2+/-0.2; II 9.4+/-0.03; III 14.1+/-0.2; IV 23.0+/-0.2. Specific N uptake by plant tissue, 2.41-2.98 mg N/g plant (text, p.4), stated as an aggregate range across all phases (not broken out per-phase) and compared against a secondary/external figure of 7.8 mg N/g from Santamaria et al. 1999 (batch experiment, different paper) — excluded from plant_measurements.csv because (a) no per-phase/per-trial value is given, only a whole-experiment range, and (b) it reads as an uptake RATE analogous to Table 3’s uptake-rate column rather than a static tissue-composition percentage, so it does not cleanly fit the ‘mineral’ category’s intended use (see Extraction notes). Feed composition beyond crude protein: 2% lipids, 1% calcium, 5% fibre, 1.8% phosphate, dry weight basis (Methods 2.2).

Studies references

Explorer

Nitrogen transformation in engineered aquaponics with water celery (Oenanthe javanica) and koi carp (Cyprinus carpio): Effects of plant to fish biomass ratio

Nitrogen transformation in engineered aquaponics with water celery (Oenanthe javanica) and koi carp (Cyprinus carpio): Effects of plant to fish biomass ratio

Metadata

Abstract

Summary

Nitrogen transformation and the plant-to-fish biomass ratio

Nitrous oxide (N2O) emissions

Fish and plant performance (FCR, yield)

Citations to chase

Extraction notes

Data Tables

Trial Parameters

paudelNitrogenTransformationEngineered2020-T1

paudelNitrogenTransformationEngineered2020-T2

paudelNitrogenTransformationEngineered2020-T3

paudelNitrogenTransformationEngineered2020-T4

Graph View

Table of Contents

Backlinks