Quartz 5

Decoupled systems on trial: Eliminating bottlenecks to improve aquaponic processes

Properties1
tagsmeta/type/exploratory, meta/region/europe, meta/fish/tilapia, meta/plant/tomato

55 min read

Decoupled systems on trial: Eliminating bottlenecks to improve aquaponic processes

Metadata

  • Cite key: monseesDecoupledSystemsTrial2017
  • Item type: Journal Article
  • Authors: H. Monsees, W. Kloas, S. Wuertz
  • Affiliation: Leibniz-Institute of Freshwater Ecology and Inland Fisheries, Berlin, Germany; Albrecht Daniel Thaer-Institute of Agricultural and Horticultural Sciences, Humboldt University Berlin, Berlin, Germany (both authors’ institutes listed jointly for all three authors in the byline)
  • Journal: PLOS ONE 12(9) (2017) e0183056
  • Date: 09/2017 (Received 3 March 2017; Accepted 30 July 2017; Published 28 September 2017)
  • Date added: 2026-08-09
  • DOI: 10.1371/journal.pone.0183056
  • Funding: Elsa-Neumann Scholarship (funded HM’s PhD stipend); “the funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript”
  • URL: https://doi.org/10.1371/journal.pone.0183056
  • PDF: Monsees et al. - 2017 - Decoupled systems on trial Eliminating bottleneck.pdf

Opinion

An honestly-framed pilot study, not a hypothesis-tested experiment, and it says so itself repeatedly (“pilot study” appears in the Abstract, Discussion, and Conclusions). Each of the three system types (conventional RAS, coupled aquaponics, decoupled aquaponics) is a single physical unit — there is no system-level replication at all, and no inferential statistic (ANOVA, t-test, regression p-value) appears anywhere for this paper’s own data, only means +/- SD and running-text percentage differences. That is a real limitation for citing this as evidence of a significant effect, but the underlying comparison is still useful: it is one of the few papers that runs coupled and decoupled aquaponics side by side under otherwise identical conditions (same fish cohort size/species, same fertilizer dosing schedule, same 154-day period), rather than comparing a single aquaponic design against a hydroponic control. The core finding — 36% higher tomato fruit yield in decoupled vs coupled, attributed to fertilizer dilution in the coupled system’s larger shared water volume plus its pH compromise between fish and plant optima — is mechanistically plausible and consistent with the fact fertilizer inputs were identical between the two systems (Table 1). Fish performance itself is the paper’s most solid claim: FCR, SGR and mortality were all “in the same range” across all three RAS types including the fish-only reference, suggesting aquaponic water quality at these stocking densities (~40 kg/m3) did not measurably compromise tilapia growth. Note there is no hydroponic-only arm anywhere in this paper — it compares two flavors of aquaponics against each other and against conventional aquaculture, not against hydroponics, so AP vs HP-style downstream comparisons across the vault will need NA HYD columns and cannot be pooled the way most other notes’ trials can.

Abstract

In classical aquaponics (coupled aquaponic systems, 1-loop systems) the production of fish in recirculating aquaculture systems (RAS) and plants in hydroponics are combined in a single loop, entailing systemic compromises on the optimal production parameters (e.g. pH). Recently presented decoupled aquaponics (2-loop systems) have been awarded for eliminating major bottlenecks. In a pilot study, production in an innovative decoupled aquaponic system was compared with a coupled system and, as a control, a conventional RAS, assessing growth parameters of fish (FCR, SGR) and plants over an experimental period of 5 months. Soluble nutrients (NO3—N, NO2—N, NH4+-N, PO43-, K+, Ca2+, Mg2+, SO42-, Cl2- and Fe2+), elemental composition of plants, fish and sludge (N, P, K, Ca, Mg, Na, C), abiotic factors (temperature, pH, oxygen, and conductivity), fertilizer and water consumption were determined. Fruit yield was 36% higher in decoupled aquaponics and pH and fertilizer management was more effective, whereas fish production was comparable in both systems. The results of this pilot study clearly illustrate the main advantages of decoupled, two-loop aquaponics and demonstrate how bottlenecks commonly encountered in coupled aquaponics can be managed to promote application in aquaculture.

Summary

The authors ran three parallel, single-unit recirculating systems for 154 days at a Berlin research facility: a conventional fish-only RAS (RAS A, reference), a classical single-loop coupled aquaponic system (RAS C / HydroC), and a two-loop decoupled aquaponic system (RAS D / HydroD, water supplied to the hydroponic unit on-demand via a one-way valve and not returned to the fish tanks). All three units were stocked with Nile tilapia (~68 g at first stocking, ~40 kg/m3 initial density) and fed an identical, fixed per-interval feeding schedule; the two aquaponic units’ hydroponic sides (5 NFT trays each) grew tomato (var. Pannovy, 15 plants/unit) and received identical fertilizer top-ups on top of whatever nutrients the fish water already supplied. Fish growth, FCR (1.2-1.3), SGR (~1.0), and mortality (<1%) were similar across all three systems, supporting the paper’s claim that fish welfare/performance was not compromised by either aquaponic design relative to a conventional RAS. Tomato fruit yield, however, differed substantially: the decoupled system produced 123.5 kg total (8.2 kg/plant) versus 90.9 kg (6.1 kg/plant) in the coupled system, a 36% advantage the authors attribute to the coupled system’s larger shared water volume diluting the identical fertilizer dose, compounded by a pH compromise (coupled-loop pH ~7.1, good for nitrification but suboptimal for nutrient uptake; decoupled hydroponic-side pH ~6.4, closer to the plant optimum) — interestingly, the coupled system produced more leaf/root/stem biomass than the decoupled system, which the authors read as a nutrient-foraging root response to its poorer nutrient supply. The paper explicitly self-identifies as an unreplicated pilot study throughout (Abstract, Discussion, Conclusions) and reports no formal significance testing anywhere; its contribution is a same-facility, same-cohort side-by-side demonstration of coupled vs decoupled aquaponics rather than a hypothesis-tested comparison.

Experiment data

  • Location: Aquaponic research facility of the Leibniz-Institute of Freshwater Ecology and Inland Fisheries, Berlin, Germany
  • Design: Three parallel, single-unit systems (RAS A = fish-only conventional-RAS reference; RAS C = coupled aquaponics; RAS D = decoupled aquaponics), each with total volume 16.5 m3 (culture volume 6.8 m3, four 1.7 m3 rearing tanks). No system-level replication — one physical unit per treatment type; explicitly self-described as a “pilot study.” No randomization or blinding described.
  • Replicates / n: 1 physical system per treatment (unreplicated at the system level); sub-sample replication within specific assays: n=5 leaf/tomato tissue pairs per timepoint (Table 5), n=3 total-suspended-solids samples (Fig 3), n=5 clarifier-solids subsamples/week (Table 6), n=4 fish/sludge composition samples (Table 5)
  • Duration: 154 days (07 April - 07 September 2015)
  • Organisms: Nile tilapia (Oreochromis niloticus) (~68 g at first stocking) / Tomato (Solanum lycopersicum) (var. Pannovy)
  • Statistics: None. Descriptive means +/- SD only (Tables 2-6); no ANOVA/t-test/regression-p reported for this paper’s own data anywhere in the text — comparisons are made via stated percentage differences.
  • Fruit yield: HydroC (coupled) 90.9 kg total / 6.1 kg per plant vs HydroD (decoupled) 123.5 kg total / 8.2 kg per plant over 154 d — 36% higher in the decoupled system
  • Feed Conversion Rate (FCR) / Specific Growth Rate (SGR): FCR 1.2 (C) / 1.3 (D) / 1.3 (A, reference); SGR ~1.0 in all three systems; mortality <1% in all three
  • Aquaponic-loop water chemistry: coupled loop (RASC/HydroC, single body of water) pH 7.1+/-0.3, DO 6.5+/-1.1 mg/L, EC 1.5+/-0.3 dS/m, 26.8+/-1.0 degC; decoupled hydroponic side (HydroD) pH 6.4+/-0.7, DO 8.2+/-0.4 mg/L, EC 3.2+/-1.0 dS/m, 24.3+/-1.5 degC — markedly different from its own fish loop (RASD: pH 7.2, DO 6.3, EC 1.5, 27.3 degC)

Coupled vs decoupled aquaponic yield

This paper: Tomato fruit yield was 36% higher in the decoupled system (123.5 kg / 8.2 kg per plant) than the coupled system (90.9 kg / 6.1 kg per plant) over the same 154-day period, despite both systems receiving an identical fertilizer dosing schedule (Table 1: 694 g KristaK+, 270 g CalciNit, 175 g MannaLin M Spezial, 300 g KHCO3 per system). The authors attribute this to two compounding mechanisms: (1) the coupled system’s larger shared water volume (fish tanks + hydroponic trays as one body of water) dilutes the same absolute fertilizer dose to a lower concentration than the decoupled system’s smaller, separately-managed hydroponic loop; (2) the coupled loop’s pH (~7.1, needed for nitrification) sits well above the plant-optimal 5.5-6.5 range cited from the literature, while the decoupled hydroponic loop could be run at pH 6.4, closer to that optimum, without compromising the fish loop’s own pH. Counter-intuitively, the coupled system produced more leaf (63.7 vs 48.0 kg), root (5.8 vs 2.3 kg) and stem (25.7 vs 17.1 kg) biomass than the decoupled system — the authors read this as a root-foraging/shoot:root acclimation response to the coupled system’s comparatively poorer nutrient supply, a pattern they note has been reported before in nutrient-limited plants generally (ref. 28).

Compared with:

  • todo Kloas et al. 2015 — first prototype decoupled aquaponic system, reported 8.89 kg/plant tomato yield over 9 months; this paper’s own decoupled yield (8.2 kg/plant over only 6 months) is described as “comparable,” a faster per-month rate [this paper’s own primary comparison, not itself secondary] (p.12-13)

Fish performance across conventional, coupled, and decoupled systems

This paper: FCR (1.2-1.3), SGR (~1.0, declining from ~1.4-1.6 to ~0.5-0.6 as fish grew), and mortality (<1% in all three systems, 0.4-0.8%) were reported as “in the same range” across the fish-only conventional-RAS reference (RAS A) and both aquaponic systems (RAS C, RAS D); the paper states “initial, final weight and subsequently overall weight gain revealed no difference (<2%) between fish units” (p.5). Water consumption was higher in the conventional reference (5.3% RAS volume/day) than either aquaponic system (2.8-2.9%), which the authors attribute to backwashing of the reference unit’s automatic drum filter rather than to any difference in water quality (clarifiers were used instead of a drum filter in both aquaponic units).

Compared with:

Dissolved nutrient dynamics: coupled vs decoupled loops

This paper: Nitrate-N accumulated steadily over the 154-d period in both aquaponic hydroponic loops but reached much higher levels in the decoupled system (HydroD: 98.8 to 207.5 mg/L across the 5 sampling intervals) than the coupled system (HydroC: 19.8 to 72.8 mg/L) — consistent with the decoupled loop’s smaller water volume concentrating the same fertilizer input. The coupled loop never reached the paper’s cited literature threshold for optimal tomato nutrition (>140 mg/L NO3-N, ref. 13) even at its peak (99.5 mg/L, cited in Discussion for a slightly different date range than Table 4’s own peak), which the authors present as the proximate cause of the coupled system’s lower yield; P and K similarly failed to meet cited minimum requirements in the coupled loop. Ammonium-N spiked briefly in HydroD only, described in Results as reaching “a maximum of 6.4 mg L-1” near the start of the experiment before declining to levels comparable with the other units (Table 4’s own interval-mean figure for the same period is 3.60+/-0.28 mg/L — see Extraction notes on why this is a maximum-vs-mean distinction, not a contradiction). Potassium and phosphate were described by the authors as generally scarce relative to tomato requirements in both aquaponic loops, a pattern they say is commonly reported for RAS-derived water generally and link to their own related work on aquacultural sludge nutrient mobilization (ref. 30).

Compared with:

  • todo Rakocy, Masser and Losordo 2006 — integrating fish and plant culture (SRAC), cited for the general aquaponic pH-compromise problem (~pH 7 commonly reported) that motivates the coupled-vs-decoupled comparison (p.2)
  • todo Lattauschke 2004 — greenhouse tomato cultivation management guide, source of the >140 mg/L NO3-N and P/K minimum nutrient requirement thresholds cited against this paper’s own measured concentrations (p.13-14)
  • todo Monsees et al. 2017 (Aquacult Environ Interact, ref. 30; a companion paper by an overlapping author subset, not the paper under review here) — potential of aquacultural sludge treatment for aquaponics, cited regarding P/K recycling from fish sludge as an alternative to fertilizer supplementation (p.13)

Linked claims

Citations to chase

  • todo Kloas W, Gross R, Baganz D, Graupner J, Monsees H, Schmidt U, et al. (2015) — A new concept for aquaponic systems to improve sustainability, increase productivity, and reduce environmental impacts, Aquaculture Environment Interactions 7(2):179-92 — first prototype decoupled aquaponic system, 8.89 kg/plant tomato yield benchmark (already in vault as goddekNavigatingDecoupledAquaponic2016’s companion reference set; check whether Kloas et al. 2015 itself has a separate vault note before adding one)
  • todo Rakocy J, Masser M, Losordo T (2006) — Recirculating Aquaculture Tank Production Systems: Aquaponics — Integrating Fish and Plant Culture, SRAC 454:1-16
  • todo Lattauschke G (2004) — Gewachshaustomaten: Hinweise zum umweltgerechten Anbau, Sachsische Landesanstalt fur Landwirtschaft, Pillnitz
  • todo Ridha MT (2006) — Comparative study of growth performance of three strains of Nile tilapia at two stocking densities, Aquaculture Research 37(2):172-9
  • todo Monsees H, Keitel J, Paul M, Kloas W, Wuertz S (2017) — Potential of aquacultural sludge treatment for aquaponics: evaluation of nutrient mobilization under aerobic and anaerobic conditions, Aquaculture Environment Interactions 9:9-18

Extraction notes

Type classification, in detail: Recorded as exploratory, not experiment. This was a genuine judgment call and is worth spelling out. Against experiment: (1) there is no system-level replication anywhere — RAS A, RAS C and RAS D are each a single physical unit, not one of several replicate coupled or decoupled systems; (2) no inferential statistical test is reported anywhere in the paper for its own data (searched the full text for “signific”, “ANOVA”, “t-test”, “Tukey”, “Duncan”, “statistical” — the only two hits for “signific-” both cite OTHER papers’ findings, never this paper’s own results); (3) the paper self-identifies as a “pilot study” in the Abstract, three times in the Discussion, and again in the Conclusions. Per SCHEMA.md’s own decision rule 3 (“if there are replicates and a significance test, it is experiment even if the authors call it a pilot”), neither condition is met here, so the self-description is not being taken at face value uncritically — it is corroborated by the absence of both replication and a test. This is a controlled, defined-treatment comparison (coupled vs decoupled vs reference), which is why case-study or observational were rejected; exploratory’s own test (“pilot/proof-of-concept, descriptive, often unreplicated, no formal hypothesis test”) is the closest match.

Trial structure: Two trials.csv rows. T1 = coupled aquaponics (RAS C / HydroC) vs T2 = decoupled aquaponics (RAS D / HydroD). There is NO hydroponic-only arm anywhere in this paper — both trials’ HYD-labelled columns (HYD, Tissue nitrate HYD) are recorded NA, not NR, per SCHEMA.md’s convention for a schema block that genuinely does not apply to this paper’s design. The paper’s own “control” is a THIRD unit, RAS A, a fish-only conventional RAS with no hydroponic component at all — this does not fit either the AP or HYD role (it is neither an aquaponic treatment nor a hydroponic control) and so gets no trials.csv row of its own; its fish-performance and water-consumption data are preserved narratively in both trial rows’ Experimental Remarks since the paper itself treats it as the comparison baseline for the fish side.

Compartment choice for coupled vs decoupled water chemistry (SCHEMA.md rule, not a contradiction): In the coupled system, RAS C and HydroC are literally one body of water (Table 3 reports them as a single combined row), so there is no compartment choice to make for T1. In the decoupled system, RAS D (fish loop) and HydroD (plant loop) are physically separate and materially different in pH (7.2 vs 6.4), DO (6.3 vs 8.2 mg/L), EC (1.5 vs 3.2 dS/m) and temperature (27.3 vs 24.3 degC) — per SCHEMA.md’s instruction to “take the plant bed/hydroponic unit value and note which compartment was used” when compartments diverge, T2’s Aq pH/DO/EC/Water temperature cells all use the HydroD (plant-bed) values, with the RASD (fish-loop) values preserved in the row’s remarks for reference.

WARN-MINOR HydroD water temperature SD, Table 3 p.7 vs Results running text p.7. Table 3’s 154-d average row states HydroD temperature as “24.3(+/-1.5)degC.” Results running text restates the identical mean with a different SD: “a lower average temperature (24.3degC+/-1.7) was observed.” Same mean, SD differs by 0.2 degC — does not change any interpretation (HydroD is unambiguously the coolest unit either way). Table 3’s value (1.5) was used in T2’s Water temperature cell as the dedicated summary-statistics source; both values are preserved in that row’s Experimental Remarks.

Not a contradiction (different statistics, correctly distinguished): HydroD ammonium. Results text (p.7) states “only in HydroD a maximum of 6.4 mg L-1 NH4+-N was observed at the beginning of the experimental period, which constantly decreased to low levels comparable [to] the other systems.” Table 4’s HydroD first-interval NH4-N MEAN is 3.60+/-0.28 mg/L. Rather than treat these as conflicting values for the same quantity, they were read as two different statistics of the same underlying data — a stated instantaneous maximum (plausible as a single peak reading shortly after system startup/fish stocking) versus the table’s interval mean — and no severity flag was raised. T2’s TAN/NH4-N cell uses the interval-mean-based range (0.02-3.60 mg/L, from Table 4’s five HydroD interval means) for consistency with how the other water-ion columns are handled, not the text’s single instantaneous maximum.

Water-quality trial-mean vs range, applied per SCHEMA.md. Table 3 (dissolved oxygen, pH, temperature, conductivity) gives a genuine single 154-day trial-mean +/- SD for each system, so those four parameters are recorded as ordinary trial means. Table 4 (NO3-N, NO2-N, TAN/NH4-N, plus PO4/K/Ca/Mg/SO4/Cl/Fe, none of which have dedicated columns) gives only five discrete per-interval means with no overall summary row anywhere in the paper. Per SCHEMA.md’s instruction (“if a paper reports only a time series with no summary, record the range… do not average it yourself”), the min-max range across the five HydroC or HydroD interval means is recorded in the NO3-N/NO2-N/TAN cells for each trial, rather than a trial mean computed here.

Feed-total identity across RAS A/C/D (a design observation, not a contradiction). Table 2’s per-interval feed[kg] figures are IDENTICAL across all three RAS units for every single sampling interval (e.g. 37.2 kg in the first interval for A, C, and D alike; 325.6 kg total for all three over 154 d), while fish growth differed between units. This means all three RAS were fed a common, pre-set feeding table by calendar interval rather than a ration scaled to each system’s own measured biomass — which is exactly why FCR differs between systems (1.0/1.2/1.3 for A/C/D) despite identical feed inputs. Recorded as background in both trials’ Experimental Remarks rather than flagged as an error.

[not reported] / NOT DERIVED fields, grouped (see each trial row’s Experimental Remarks for the full paper-specific wording):

  • Fish: Fish Category, Fish size final, Fish weight gain (per-fish), feed Protein/N/P/K composition beyond a partially-garbled product-name code (“AllerFloat 37/10,” which conventionally denotes 37% crude protein/10% crude fat in Aller Aqua’s own naming scheme, but the paper never states this as a measured/labelled composition figure — left NR and flagged [unclear] rather than assumed from the naming convention), Feed routine/frequency, Total Feed’s %-of-body-weight basis (the paper uses a fixed kg-per-interval schedule instead, identical across all three systems)
  • Water: Water recycle (T2 only — T1’s RAS C by-pass pump flow rate, 10 L/min, IS stated), Water type, Water classification, pHOptimal, FUE AP, FUE HYD, WUE
  • Plant: Plants/m2 (plant count and tray dimensions given separately, never combined into a stated density), Plant height at harvest (only the pre-trial transplant height, 42.1+/-4.3cm, is given), Leaf count, Plant dry matter (%), Tissue nitrate AP/HYD (Table 5’s elemental panel covers Ca/K/Mg/Na/P/N/C, not nitrate — tissue nitrate was simply never measured in this paper)
  • Site: Lat/Long (facility named only as “Berlin, Germany,” no coordinates stated in the paper itself — not looked up externally per CLAUDE.md’s prime directive), Average room Temperature, Climate control, Artificial Lighting, Air supplement

NO COLUMN items (full figures preserved in each trial row’s Experimental Remarks): Table 1’s leaf/root/stem fresh-weight harvest totals (more non-fruit biomass in the coupled system despite less fruit — a genuinely interesting reversal with no dedicated column); Table 1’s per-interval fertilizer-gram and HydroD water-consumption-litre schedule; Table 4’s full water-mineral panel beyond NO3-N/NO2-N/TAN (PO4, K+, Ca2+, Mg2+, SO4(2-), Cl-, Fe2+); Table 5’s carbon content (%) and C/N ratio for leaf and tomato tissue (does not fit any of plant_measurements.csv’s four categories — excluded from plant.csv for the same reason physiological data was excluded in mourantianBasilFunctionalGrowth2023); Table 5’s fish and sludge elemental composition (neither a plant analyte nor water chemistry, so it fits neither trials.csv nor plant_measurements.csv’s scope); Table 6’s clarifier solids-removal data and Fig 3’s total-suspended-solids time series; Fig 4’s estimated-nitrogen-fate schematic (a mixed primary/secondary construct built from this study’s own results plus literature values, not read into any single cell).

plant_measurements.csv scope decision: Table 5’s ICP-OES elemental panel (Ca, K, Mg, Na, P, all g/kg DW) and Vario EL nitrogen content (N, % DW) were extracted for both leaf and fruit (tomato) tissue, both systems (CAP = HydroC, DCAP = HydroD), across all four sampled timepoints (days 63, 94, 122, 154) — 96 rows total, all Category mineral, with tissue type folded into the Analyte name (e.g. “Calcium (Ca) (leaf)” vs “Calcium (Ca) (fruit)”) per the vault’s existing convention for multi-tissue analytes. Carbon content and C/N ratio from the same table were NOT extracted (see NO COLUMN above). No inferential significance test is reported for any of these values (consistent with the whole-paper absence of formal statistics), so every row’s Significance field states this explicitly rather than leaving it silently blank or guessing ns.

Tags judgment call: Tagged Meta/Fish/Tilapia (Nile tilapia, Oreochromis niloticus, the sole aquaculture species, reared identically across all three system types) and Meta/Plant/Tomato (Solanum lycopersicum, the sole crop, grown in both aquaponic units). Meta/Region/Europe per Germany. No new tag facets introduced (reused Meta/Type/Exploratory, already established in buzbyScalingAquaponicSystems2014).

New wikilink targets introduced: H. Monsees, W. Kloas, S. Wuertz (no existing author notes found in the vault for any of the three). Reused existing canonical forms: Nile tilapia (Oreochromis niloticus), Tomato (Solanum lycopersicum), Feed Conversion Rate (FCR), Specific Growth Rate (SGR), Decoupled aquaponics systems outperform single-loop coupled systems, Decoupled aquaponics allows independent optimization of fish and plant pH (this last claim note already exists, created by blanchardEffectPHCucumber2020, and is an unusually direct match to this paper’s own stated mechanism).

Quality score rationale: caution. Zero ⚠️BLOCK and zero ⚠️MATERIAL contradictions were found (the two flags raised, the water-temperature SD restatement and the ammonium maximum-vs-mean distinction, are both ⚠️MINOR or explicitly “not a contradiction” per SCHEMA.md’s own examples). Per the stated scoring rule (“ok: 0 BLOCK and <=2 MATERIAL”), this paper’s contradiction profile alone would score ok. caution is applied instead because of the unreplicated, no-significance-test pilot design itself (n=1 system per treatment throughout) — a structural limitation on how much weight the yield comparison can bear, distinct from and not scored by the BLOCK/MATERIAL contradiction-count rule, but worth flagging for anyone pooling this paper’s yield numbers with replicated studies elsewhere in the vault.

PDF quality: Full text layer present throughout (18 pages), cleanly extractable for all narrative text. Two tables (Table 2, stocking/FCR/SGR/mortality/water-consumption/limestone detail; Table 4, dissolved nutrient panel) are typeset as landscape-rotated tables within the portrait page layout; pdfplumber’s character-level text extraction returned these two tables’ cell contents in a scrambled/reversed reading order (rotation metadata reported 0 at the page level despite the visible content being rotated), so both tables were instead read directly from rendered page images (PyMuPDF pixmap render, 3x-6x zoom) rather than from the garbled text layer, per CLAUDE.md’s instruction not to extract from a garbled layer. All other tables (1, 3, 5, 6) and all narrative text extracted cleanly from the text layer with no rotation issues.

Source: Monsees et al. - 2017 - Decoupled systems on trial Eliminating bottleneck.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

monseesDecoupledSystemsTrial2017-T1

Fish

FieldValue
FishNile tilapia (Oreochromis niloticus)
Initial Stock density~39 (kg/m3, start; explicitly stated for the coupled system specifically, Discussion p.13-14: ‘at fish densities between 39 (start) and 65 kg/m3 (end)’; a general, rounder ‘~40 kg/m3’ figure is also given for the pilot study as a whole, Discussion p.13)
FCR1.2
SGR1.0
Fish size initial68
Feed regimeCommercial pelleted feed (AllerFloat, Emsland-Aller Aqua, Germany; product code partially garbled in the PDF text layer, read as ‘37/10’, pellet sizes 2mm and 3mm) fed on a fixed feeding table identical across RAS A/C/D by sampling interval (Table 2), not expressed as a %-of-body-weight ration
Total Feed (kg)325.6
Fish biomass created (kg)263.4 (kg, Table 2 ‘fish growth’ total/average row, RAS C, over 154 d)
Fish survival rate0.7% mortality (of stocking, cumulative to day 154, Table 2 total/average row, RAS C)
Fish trial duration (days)154

Water

FieldValue
Water recycle10 (L/min, RAS C by-pass pump feeding the 5 NFT trays, Methods p.3)
Water volume in the system16500 (total RAS system volume incl. biofilter/sump, stated identically for RAS A/C/D, UNIT CONVERSION ONLY: 16.5 m3 -> 16500 L, Methods p.2-3; culture/rearing-tank volume alone = 6800 L, four tanks of 1700 L each; the hydroponic unit’s own water volume is not stated — 5 NFT trays of 45x30x28cm are given as dimensions only, not a water depth/volume)
Daily Water exchange rate2.9 (% RAS volume/day, Table 2 total/average row, RAS C; Results text: ‘ranging between 2-3.6%’ for both aquaponic systems)
Aq pH7.1 +/- 0.3 (RASC/HydroC combined loop — single-loop coupled system, Table 3)
Dissolved Oxigen6.5 +/- 1.1 (mg/L, RASC/HydroC combined loop, Table 3)
EC1.5 +/- 0.3 (dS/m, RASC/HydroC combined loop, Table 3; UNIT CONVERSION ONLY: mS/cm numerically = dS/m)
Water temperature26.8 +/- 1.0 (degC, RASC/HydroC combined loop, Table 3)
TAN / NH4-N0.04-0.15 (mg/L NH4-N/TAN, range only, Table 4, HydroC)
NO2-N0.05-0.08 (mg/L NO2-N, range only, Table 4, HydroC)
NO3-N19.8-72.8 (mg/L NO3-N, range only across 5 sampling intervals, no single trial mean reported, Table 4, HydroC)

Plant

FieldValue
PlantTomato (Solanum lycopersicum), variety Pannovy
DetailsTomato transplants (var. Pannovy) sourced from a commercial hydroponic-vegetable supplier (Schwantelend GmbH, Germany), grown in rockwool cubes (10x10 cm), mean height 42.1 +/- 4.3 cm at stocking (Methods p.3-4); 15 plants/RAS randomly distributed across 5 NFT trays per hydroponic unit; leaf and fruit tissue destructively sampled at 4 timepoints (interval end days 63, 94, 122, 154): 5 replicate pairs of leaves (always the 5th fully developed leaf) and 5 replicate pairs of fully ripe tomatoes per timepoint/system, freeze-dried prior to elemental analysis (Methods p.4)
Days Plant after transplant154
Plant fresh weight6100 (g/plant fruit fresh weight, HydroC; UNIT CONVERSION ONLY: authors’ own stated ‘6.1 kg plant-1’, Discussion p.12, cross-checked against Table 1: 90.9 kg total / 15 plants/RAS = 6.06, matches)

System & Setup

FieldValue
System typeNFT (Nutrient Film Technique) trays (Fig 1 caption; Methods p.3)
Media Details5 NFT trays per hydroponic unit, 45 x 30 x 28 cm (l x w x h) each; tomato in rockwool cubes (10x10 cm) prior to transplant (Methods p.3-4)
Biological system already in useY (Each RAS equipped with a moving bed biofilter (2 m3, ~1350 m2 substrate surface) for nitrification (Methods p.2-3); RAS A additionally used a drum filter (100 um mesh) for solids removal, RAS C and RAS D each used a clarifier (1.5 m3) instead)
Iron supplementedY (MannaLin M Spezial NPK fertilizer includes trace elements Fe, Mn, Zn, B, Cu, Mo (Methods p.4), dosed identically to HydroC and HydroD per Table 1; Fe2+ in the water was also monitored weekly (Table 4) but stayed at or near the 0.01-0.02 mg/L detection floor in HydroC and only slightly higher (0.01-0.12 mg/L) in HydroD)
pH BuffersY (Ca(OH)2 (limestone) added to RAS units to counteract nitrification-driven pH decline (Methods p.2); total limestone addition over 154 d: RAS C 15.6 kg, RAS D 15.3 kg (Table 2))
Nutrient supplementedY (KristaK Plus (13.7% total N as NO3-N, 46.3% K2O), CalciNit (15.5% total N: 14.4% NO3-N + 1.1% NH4-N, 26.3% CaO), MannaLin M Spezial NPK (18% total N: 11% NO3-N + 7% NH4-N, 12% P2O5, 18% K2O, 2% MgO + Fe/Mn/Zn/B/Cu/Mo trace elements), and KHCO3 (added to raise K further) dosed on an IDENTICAL schedule to HydroC and HydroD (Table 1, Methods p.3-4); total per system over 154 d: KristaK+ 694 g, CalciNit 270 g, MannaLin M Spezial 175 g, KHCO3 300 g)
EquipmentMoving bed biofilters (2 m3/RAS, ~1350 m2 substrate surface); RAS A drum filter (100 um mesh); RAS C/D clarifiers (1.5 m3 each); small mechanical filter (Eheim, Germany) ahead of the NFT trays in RAS C; HQ40d multi-probe (Hach Lange, Germany) for daily temperature/pH/oxygen; DR3900 spectrophotometer (Hach Lange, Berlin, Germany) for weekly dissolved-nutrient determination (NO3-N, NO2-N, TAN, PO4, K, Mg, Ca, SO4, Cl, Fe); iCAP 6000 ICP-OES (Thermo Fisher Scientific Inc., USA) for plant/fish/sludge Ca/K/Mg/Na/P; Vario EL system (Elementar Analysensysteme GmbH, Germany) for C/N analysis; Gigatherm high-pressure microwave oven (Switzerland) for wet digestion (HCl 37%/HNO3 65%, 1:3) prior to ICP-OES
Control ParameterspH regulated with Ca(OH)2 addition in RAS units to counteract nitrification-driven pH drop (literature-cited optimum ~pH 7 for nitrification, not a stated operational setpoint); temperature, pH and dissolved oxygen measured daily in all units; dissolved nutrients (NO3-N, NO2-N, TAN, PO4, K, Mg, Ca, SO4, Cl, Fe) measured weekly (Methods p.2-4)
CombinationNile tilapia (Oreochromis niloticus) and tomato (Solanum lycopersicum, var. Pannovy) in a pilot-scale comparison of coupled (T1) vs decoupled (T2) aquaponics; both also benchmarked in the paper’s own text against a third, fish-only conventional-RAS reference unit (RAS A, no hydroponic component), which is not itself an aquaponic treatment and so receives no trials.csv row of its own — see Experimental Remarks

Site

FieldValue
RegionEurope
CountryGermany

Results & Statistics

FieldValue
Measured Unitkg (total-system fruit/leaf/root/stem fresh-weight harvest per interval and cumulative, Table 1); g/plant (author-stated per-plant fruit yield, Discussion p.12, cross-checked against Table 1 totals divided by 15 plants per RAS); g/kg DW and % DW (leaf and fruit elemental composition, Table 5); mg/L (dissolved water nutrients, Tables 3-4); kg/RAS (cumulative fish growth, Table 2)
Statistic DetailsDescriptive statistics only: means +/- SD reported throughout (Tables 2-6); no inferential/significance test (ANOVA, t-test, regression p-value, etc.) is reported anywhere in the paper for any comparison — the word ‘significant’ appears only twice in the whole text, both times citing OTHER authors’ findings (a denitrification N-loss study; the authors’ own separate nitrate-toxicity paper), never this paper’s own results. System-to-system comparisons throughout Results/Discussion are stated only as percentage differences or qualitative range overlaps.
Statistically analysedN
Replicates (n)1 (system-level: one physical RAS per treatment type A/C/D — explicitly an unreplicated pilot study, self-described as such in Abstract/Discussion/Conclusions; sub-sample replication exists within some assays: n=5 leaf/tomato tissue pairs per timepoint (Table 5), n=3 TSS samples per RAS (Fig 3), n=5 clarifier-solids subsamples/week (Table 6), n=4 fish/sludge composition samples, Table 5)
AP6100 (g/plant fruit fresh weight, HydroC; UNIT CONVERSION ONLY: authors’ own stated ‘6.1 kg plant-1’, Discussion p.12, cross-checked against Table 1: 90.9 kg total / 15 plants/RAS = 6.06, matches)

Experimental Remarks: TRIAL DEFINITION: T1 = coupled aquaponics (RAS C / HydroC) — single-loop system, five NFT hydroponic trays connected directly to the RAS via a by-pass pump (10 L/min) installed in the pump sump, water constantly circulated RAS<->hydroponic and back (Methods p.3, Fig 1a). There is NO hydroponic-only control in this paper at all (HYD columns recorded NA throughout); the paper’s comparison structure is coupled-vs-decoupled aquaponics, benchmarked against a third, fish-only conventional-RAS reference (RAS A, no hydroponic unit — not an aquaponic treatment, so it gets no row of its own, but its fish-performance data is given below for context since the paper uses it as its explicit ‘control’). Paired trial = T2 (decoupled), sharing the same fish species, feed regime, fertilizer dosing schedule, and 154-d experimental period, but a different water-loop architecture. | RAS A (fish-only conventional-RAS reference, NOT a trial row): FCR 1.3, SGR 1.0, mortality 0.8%, cumulative fish growth 257.7 kg over 154 d, water consumption 5.3% RAS/d (all Table 2 total/average row) — the paper’s Results (p.5) states fish growth/FCR/SGR were ‘in the same range among all three RAS (A, C, D)’ and ‘initial, final weight and subsequently overall weight gain revealed no difference (<2%) between fish units’, i.e. the paper’s own headline finding is that aquaponic water quality does NOT measurably impair tilapia performance relative to a purpose-built aquaculture-only system. RAS A’s higher water-exchange rate (5.3% vs 2.8-2.9% in C/D) is attributed by the authors to backwashing of its automatic drum filter vs the clarifiers used in C/D (Discussion p.13), not to any water-quality difference. | WARN-MINOR HydroD water temperature SD, restated with a different value (does not affect T1’s own cells, included here for completeness since it concerns Table 3/Results consistency generally): Table 3 states HydroD temperature as ‘24.3(+/-1.5)degC’ (154-d average row); Results running text (p.7) restates the same mean with a different SD: ‘24.3 degC +/-1.7’. Table 3’s value is used in T2’s Water temperature cell as the more structured, single-purpose summary-statistics source; the text’s 1.7 is preserved in T2’s remarks. Does not change any interpretation (both describe the same 24.3 degC mean). | NOT DERIVED, left NR: Fish size final (only per-RAS aggregate biomass given — e.g. RAS C/D stocking-end totals in Table 2 — with no stated fish count per tank/RAS to back out a mean individual weight); Fish weight gain per fish (same reason; population-level aggregate recorded instead as Fish biomass created); Feed routine/frequency (not stated, only the fixed per-interval kg feed totals in Table 2); feed Protein/N/P/K composition (the feed product name reads ‘AllerFloat 37/10’ in the extracted text, which by Aller Aqua’s own naming convention conventionally denotes 37% crude protein/10% crude fat, but the paper itself never states this as a composition figure — recorded NR per the no-inference rule rather than assumed from the product-naming convention; flagged [unclear] rather than confidently extracted); Plants/m2 (paper gives plant count, 15/RAS, and tray dimensions, 45x30cm x5 trays, separately but never combines them into a stated density — computing one would be derivation); Plant height at harvest (only the pre-trial TRANSPLANT height, 42.1+/-4.3cm, is given — see Details — no harvest-time height is ever stated); Plant dry matter% (elemental-analysis samples were freeze-dried, but no dry-matter percentage of fresh weight is stated anywhere); Water type/Water classification (RAS water source and category never named); Lat/Long (facility named only as ‘Leibniz-Institute of Freshwater Ecology and Inland Fisheries, Berlin, Germany’ — no coordinates stated in the paper itself, and per CLAUDE.md’s prime directive coordinates are not looked up externally when the paper is silent); Average room Temperature (no ambient/room temperature distinct from the RAS/Hydro water temperatures in Table 3 is given, and the facility’s indoor/greenhouse status is not stated either); pHOptimal, WUE, FUE AP, FUE HYD (no stated optimum-pH target or efficiency metric matching these columns; the Discussion cites LITERATURE pH optima of 5.5-6.5 for plants and ~7 for nitrification, but these are secondary figures from cited sources [16,18], not this paper’s own targets, so left NR rather than repurposed here). | NO COLUMN: Table 1’s leaf/root/stem fresh-weight harvest totals over 154 d (leaf: HydroC 63.7 kg vs HydroD 48.0 kg; root: HydroC 5.8 kg vs HydroD 2.3 kg; stem: HydroC 25.7 kg vs HydroD 17.1 kg — i.e. MORE non-fruit biomass in the coupled system despite LESS fruit, Results p.5, attributed by the authors to suboptimal coupled-system nutrient supply driving a root-foraging/shoot:root acclimation response, Discussion p.11-12) — no dedicated leaf/root/stem yield columns exist alongside the single ‘Plant fresh weight’ (fruit) column. Table 1’s per-interval fertilizer-gram and water-consumption-litre schedule (both systems dosed identically per Table 1; HydroD water consumption ranged 1.4-2.4 L/plant/day, Results p.5; HydroC water consumption is not separately tracked since it is coupled to the RAS’s own volume, Table 1 caption). Table 2’s per-interval (rather than whole-trial) FCR/SGR/mortality/water-consumption/limestone breakdown (used here only as the source for the recorded total/average-row values). Table 4’s full water-mineral panel beyond NO3-N/NO2-N/TAN: PO4, K+, Ca2+, Mg2+, SO4(2-), Cl-, Fe2+ (mg/L) for both HydroC and HydroD across all 5 sampling intervals — no dedicated trials.csv columns exist for these ions. Table 5’s carbon content (%) and C/N ratio for leaf and tomato tissue in both systems (does not fit any of plant_measurements.csv’s four categories — biochemistry/mineral/microbiology/proximate — carbon content is neither a standard proximate component nor a mineral element; excluded from plant.csv per the same principle used for physiological data in other vault notes, e.g. mourantianBasilFunctionalGrowth2023). Table 5’s fish (RAS A-C-D pooled, n=4) and sludge (RAS C-D pooled, n=4) elemental composition (Ca/K/Mg/Na/P/N/C, C/N) — neither a plant analyte nor water chemistry, so it fits neither trials.csv nor plant_measurements.csv’s scope; preserved here only: fish (g/kg DW unless %): Ca 31.7(1.0), K 1.5(0.1), Mg 2.1(0.1), Na 0.7(0.0), P 17.7(0.5) g/kg; N 7.4(0.2)%, C 56.5(3.3)%, C/N 7.6(0.5); sludge: Ca 11.9(5.8), K 8.3(0.1), Mg 0.6(0.1), Na 3.5(0.1), P 8.9(2.8) g/kg; N 4.1(0.2)%, C 36.6(1.0)%, C/N 9.0(0.6). Table 6’s clarifier solids-removal data (RAS C 1.8+/-0.07 g DW/L vs RAS D 2.0+/-0.09 g DW/L weekly mean, n=5, 3 consecutive weeks) and Fig 3’s TSS time series (RAS A/C/D at day 30/94/154, n=3) — system-performance/waste-stream data with no matching trials.csv column. Fig 4’s estimated-nitrogen-fate schematic (modelled from this study’s results plus literature values for %N of protein, fish N excretion, nitrification and denitrification rates, and tomato nitrate uptake — a mixed primary/secondary construct, not read into any cell). Discussion’s cross-paper yield comparison: Kloas et al. 2015 (ref 5, a related prior study by an overlapping author group, first decoupled-aquaponics prototype) reported 8.89 kg/plant tomato yield over 9 months, vs this paper’s 8.2 kg/plant (HydroD) and 6.1 kg/plant (HydroC) over only 6 months — secondary comparison figure, see Citations to chase in the note. | UNIT CONVERSION ONLY: RAS total volume 16.5 m3 -> 16500 L; culture volume 6.8 m3 -> 6800 L (both Methods p.2-3); conductivity reported in Table 3 as mS/cm, numerically identical to the schema’s dS/m (1 mS/cm = 1 dS/m), so no arithmetic conversion was actually needed, only a unit-name equivalence — noted for auditability. | Compartment-choice note (not a contradiction, per SCHEMA.md’s own instruction): T1’s Aq pH/Dissolved Oxygen/EC/Water temperature cells use the single RASC/HydroC value from Table 3, since the coupled design means fish-tank and hydroponic-tray water are literally the same body of water (Table 3 reports them as one combined ‘RASC/HydroC’ row, not two separate ones) — unlike T2, there is no compartment choice to make here. | Water-quality trial-mean vs range note: NO3-N/NO2-N/TAN/NH4-N have NO stated single trial-mean across the 154-d period anywhere in the paper (Table 4 gives only 5 discrete per-interval means, no summary row, unlike Table 3’s DO/pH/temp/conductivity which DO have a stated 154-d average). Per SCHEMA.md’s ‘range only, no trial mean reported’ rule, the min-max range across HydroC’s 5 intervals is recorded in each of these cells rather than an average computed here. | Feed-total identity across RAS A/C/D (not a contradiction, a design observation): Table 2’s feed[kg] column is IDENTICAL for every sampling interval across all three RAS units (e.g. 37.2 kg in interval 1 for A, C, and D alike; 325.6 kg total for all three) while fish growth differs between units — i.e. all three RAS were fed a common, pre-set feeding table by interval rather than a ration proportional to each system’s own measured biomass, which is why FCR differs between systems despite identical feed inputs. This explains rather than contradicts the reported FCR values (1.0/1.2/1.3 for A/C/D total-average row).

monseesDecoupledSystemsTrial2017-T2

Fish

FieldValue
FishNile tilapia (Oreochromis niloticus)
Initial Stock density~40 (kg/m3, general pilot-study figure, Discussion p.13: ‘fish were stocked at around 40 kg/m3’; unlike RAS C, no decoupled-specific start/end density range is separately restated in the Discussion)
FCR1.3
SGR1.0
Fish size initial68
Feed regimeCommercial pelleted feed (AllerFloat, Emsland-Aller Aqua, Germany; product code partially garbled in the PDF text layer, read as ‘37/10’, pellet sizes 2mm and 3mm) fed on a fixed feeding table identical across RAS A/C/D by sampling interval (Table 2), not expressed as a %-of-body-weight ration
Total Feed (kg)325.6
Fish biomass created (kg)256.9 (kg, Table 2 ‘fish growth’ total/average row, RAS D, over 154 d)
Fish survival rate0.4% mortality (of stocking, cumulative to day 154, Table 2 total/average row, RAS D)
Fish trial duration (days)154

Water

FieldValue
Water recycleNR (no flow-rate figure stated for RAS D’s on-demand one-way-valve supply to HydroD, unlike RAS C’s stated 10 L/min by-pass pump)
Water volume in the system16500 (total RAS system volume incl. biofilter/sump, stated identically for RAS A/C/D, UNIT CONVERSION ONLY: 16.5 m3 -> 16500 L, Methods p.2-3; culture/rearing-tank volume alone = 6800 L, four tanks of 1700 L each; the hydroponic unit’s own water volume is not stated — 5 NFT trays of 45x30x28cm are given as dimensions only, not a water depth/volume)
Daily Water exchange rate2.8 (% RAS volume/day, Table 2 total/average row, RAS D)
Aq pH6.4 +/- 0.7 (HydroD, plant-bed compartment per SCHEMA.md’s compartment rule; RASD fish-loop pH was 7.2 +/- 0.3, Table 3 — see Experimental Remarks)
Dissolved Oxigen8.2 +/- 0.4 (mg/L, HydroD, plant-bed compartment; RASD fish-loop DO was 6.3 +/- 1.1 mg/L, Table 3 — see Experimental Remarks)
EC3.2 +/- 1.0 (dS/m, HydroD, plant-bed compartment; RASD fish-loop conductivity was 1.5 +/- 0.3 dS/m, Table 3 — see Experimental Remarks; UNIT CONVERSION ONLY: mS/cm numerically = dS/m)
Water temperature24.3 +/- 1.5 (degC, HydroD, plant-bed compartment, Table 3; Results running text restates the same mean with SD +/-1.7 — WARN-MINOR, see Experimental Remarks; RASD fish-loop temperature was 27.3 +/- 1.2 degC)
TAN / NH4-N0.02-3.60 (mg/L NH4-N/TAN, range only [interval means], Table 4, HydroD; see Experimental Remarks re: a separately stated 6.4 mg/L instantaneous maximum in Results text)
NO2-N0.00-0.07 (mg/L NO2-N, range only, Table 4, HydroD)
NO3-N98.8-207.5 (mg/L NO3-N, range only across 5 sampling intervals, no single trial mean reported, Table 4, HydroD; matches Results text: ‘increased from 98.8 mg L-1 NO3-N to more than 170 mg L-1 from the third month on’)

Plant

FieldValue
PlantTomato (Solanum lycopersicum), variety Pannovy
DetailsTomato transplants (var. Pannovy) sourced from a commercial hydroponic-vegetable supplier (Schwantelend GmbH, Germany), grown in rockwool cubes (10x10 cm), mean height 42.1 +/- 4.3 cm at stocking (Methods p.3-4); 15 plants/RAS randomly distributed across 5 NFT trays per hydroponic unit; leaf and fruit tissue destructively sampled at 4 timepoints (interval end days 63, 94, 122, 154): 5 replicate pairs of leaves (always the 5th fully developed leaf) and 5 replicate pairs of fully ripe tomatoes per timepoint/system, freeze-dried prior to elemental analysis (Methods p.4)
Days Plant after transplant154
Plant fresh weight8200 (g/plant fruit fresh weight, HydroD; UNIT CONVERSION ONLY: authors’ own stated ‘8.2 kg plant-1’, Discussion p.12, cross-checked against Table 1: 123.5 kg total / 15 plants/RAS = 8.233, matches)

System & Setup

FieldValue
System typeNFT (Nutrient Film Technique) trays (Fig 1 caption; Methods p.3)
Media Details5 NFT trays per hydroponic unit, 45 x 30 x 28 cm (l x w x h) each; tomato in rockwool cubes (10x10 cm) prior to transplant (Methods p.3-4)
Biological system already in useY (Each RAS equipped with a moving bed biofilter (2 m3, ~1350 m2 substrate surface) for nitrification (Methods p.2-3); RAS A additionally used a drum filter (100 um mesh) for solids removal, RAS C and RAS D each used a clarifier (1.5 m3) instead)
Iron supplementedY (MannaLin M Spezial NPK fertilizer includes trace elements Fe, Mn, Zn, B, Cu, Mo (Methods p.4), dosed identically to HydroC and HydroD per Table 1; Fe2+ in the water was also monitored weekly (Table 4) but stayed at or near the 0.01-0.02 mg/L detection floor in HydroC and only slightly higher (0.01-0.12 mg/L) in HydroD)
pH BuffersY (Ca(OH)2 (limestone) added to RAS units to counteract nitrification-driven pH decline (Methods p.2); total limestone addition over 154 d: RAS C 15.6 kg, RAS D 15.3 kg (Table 2))
Nutrient supplementedY (KristaK Plus (13.7% total N as NO3-N, 46.3% K2O), CalciNit (15.5% total N: 14.4% NO3-N + 1.1% NH4-N, 26.3% CaO), MannaLin M Spezial NPK (18% total N: 11% NO3-N + 7% NH4-N, 12% P2O5, 18% K2O, 2% MgO + Fe/Mn/Zn/B/Cu/Mo trace elements), and KHCO3 (added to raise K further) dosed on an IDENTICAL schedule to HydroC and HydroD (Table 1, Methods p.3-4); total per system over 154 d: KristaK+ 694 g, CalciNit 270 g, MannaLin M Spezial 175 g, KHCO3 300 g)
EquipmentMoving bed biofilters (2 m3/RAS, ~1350 m2 substrate surface); RAS A drum filter (100 um mesh); RAS C/D clarifiers (1.5 m3 each); small mechanical filter (Eheim, Germany) ahead of the NFT trays in RAS C; HQ40d multi-probe (Hach Lange, Germany) for daily temperature/pH/oxygen; DR3900 spectrophotometer (Hach Lange, Berlin, Germany) for weekly dissolved-nutrient determination (NO3-N, NO2-N, TAN, PO4, K, Mg, Ca, SO4, Cl, Fe); iCAP 6000 ICP-OES (Thermo Fisher Scientific Inc., USA) for plant/fish/sludge Ca/K/Mg/Na/P; Vario EL system (Elementar Analysensysteme GmbH, Germany) for C/N analysis; Gigatherm high-pressure microwave oven (Switzerland) for wet digestion (HCl 37%/HNO3 65%, 1:3) prior to ICP-OES
Control ParameterspH regulated with Ca(OH)2 addition in RAS units to counteract nitrification-driven pH drop (literature-cited optimum ~pH 7 for nitrification, not a stated operational setpoint); temperature, pH and dissolved oxygen measured daily in all units; dissolved nutrients (NO3-N, NO2-N, TAN, PO4, K, Mg, Ca, SO4, Cl, Fe) measured weekly (Methods p.2-4)
CombinationNile tilapia (Oreochromis niloticus) and tomato (Solanum lycopersicum, var. Pannovy) in a pilot-scale comparison of coupled (T1) vs decoupled (T2) aquaponics; both also benchmarked in the paper’s own text against a third, fish-only conventional-RAS reference unit (RAS A, no hydroponic component), which is not itself an aquaponic treatment and so receives no trials.csv row of its own — see Experimental Remarks

Site

FieldValue
RegionEurope
CountryGermany

Results & Statistics

FieldValue
Measured Unitkg (total-system fruit/leaf/root/stem fresh-weight harvest per interval and cumulative, Table 1); g/plant (author-stated per-plant fruit yield, Discussion p.12, cross-checked against Table 1 totals divided by 15 plants per RAS); g/kg DW and % DW (leaf and fruit elemental composition, Table 5); mg/L (dissolved water nutrients, Tables 3-4); kg/RAS (cumulative fish growth, Table 2)
Statistic DetailsDescriptive statistics only: means +/- SD reported throughout (Tables 2-6); no inferential/significance test (ANOVA, t-test, regression p-value, etc.) is reported anywhere in the paper for any comparison — the word ‘significant’ appears only twice in the whole text, both times citing OTHER authors’ findings (a denitrification N-loss study; the authors’ own separate nitrate-toxicity paper), never this paper’s own results. System-to-system comparisons throughout Results/Discussion are stated only as percentage differences or qualitative range overlaps.
Statistically analysedN
Replicates (n)1 (system-level: one physical RAS per treatment type A/C/D — explicitly an unreplicated pilot study, self-described as such in Abstract/Discussion/Conclusions; sub-sample replication exists within some assays: n=5 leaf/tomato tissue pairs per timepoint (Table 5), n=3 TSS samples per RAS (Fig 3), n=5 clarifier-solids subsamples/week (Table 6), n=4 fish/sludge composition samples, Table 5)
AP8200 (g/plant fruit fresh weight, HydroD; UNIT CONVERSION ONLY: authors’ own stated ‘8.2 kg plant-1’, Discussion p.12, cross-checked against Table 1: 123.5 kg total / 15 plants/RAS = 8.233, matches)

Experimental Remarks: TRIAL DEFINITION: T2 = decoupled aquaponics (RAS D / HydroD) — two-loop system, water recirculated separately within RAS D and within the hydroponic unit; water is directed from RAS D to HydroD via a one-way valve only on-demand (to replace evapotranspiration losses) and is NOT redirected back to RAS D (Methods p.3, Fig 1b). There is NO hydroponic-only control in this paper (HYD columns recorded NA throughout); comparison structure is coupled-vs-decoupled aquaponics, benchmarked against a third, fish-only conventional-RAS reference (RAS A — see T1’s remarks for its data, not repeated here). Paired trial = T1 (coupled), sharing the same fish species, feed regime, fertilizer dosing schedule and 154-d experimental period, but a single-loop water architecture instead. | Compartment-choice note (SCHEMA.md instruction, not a contradiction): unlike the coupled system, RAS D (fish loop) and HydroD (plant loop) are physically separate bodies of water with materially different chemistry (Table 3): RASD pH 7.2+/-0.3 vs HydroD pH 6.4+/-0.7; RASD DO 6.3+/-1.1 mg/L vs HydroD DO 8.2+/-0.4 mg/L; RASD conductivity 1.5+/-0.3 dS/m vs HydroD 3.2+/-1.0 dS/m; RASD temperature ~27.3+/-1.2 degC vs HydroD 24.3+/-1.5 degC. Per SCHEMA.md’s compartment rule (‘take the plant bed/hydroponic unit value and note which compartment was used’), this row’s Aq pH/Dissolved Oxygen/EC/Water temperature cells all use the HydroD (plant-bed) values; the RASD (fish-loop) values are given here for reference and are the ones comparable to T1’s single RASC/HydroC figures and to RAS A. | WARN-MINOR HydroD water temperature SD, two sources, p.7 vs Table 3: Table 3 (154-d average row) states ‘24.3(+/-1.5)degC’; Results running text restates the identical mean with a different SD: ‘a lower average temperature (24.3degC+/-1.7) was observed’. Both give the same mean (24.3), only the SD differs (1.5 vs 1.7) — does not change any interpretation (HydroD is unambiguously cooler than the other units either way). Table 3’s value (1.5) is recorded in the Water temperature cell as the dedicated summary-statistics source; the text’s 1.7 is preserved here for the record. | Not a contradiction (different statistics, not conflicting values): Results text (p.7) states ‘only in HydroD a maximum of 6.4 mg L-1 NH4+-N was observed at the beginning of the experimental period, which constantly decreased to low levels comparable [to] the other systems.’ Table 4’s HydroD first-interval NH4-N MEAN is 3.60+/-0.28 mg/L. The text is explicitly describing a ‘maximum’ (a single peak reading, plausible immediately after system startup/fish stocking) while the table reports the INTERVAL MEAN — these are different statistics of the same underlying data, not a conflicting pair of means, so no severity flag is raised. The TAN/NH4-N cell for this row uses the interval-mean-based range (0.02-3.60 mg/L, from Table 4’s 5 HydroD means) rather than the text’s single instantaneous maximum, consistent with the ‘range only, no trial mean reported’ convention applied to the other water-ion columns. | NOT DERIVED, left NR: Fish size final (only per-RAS aggregate biomass given — e.g. RAS C/D stocking-end totals in Table 2 — with no stated fish count per tank/RAS to back out a mean individual weight); Fish weight gain per fish (same reason; population-level aggregate recorded instead as Fish biomass created); Feed routine/frequency (not stated, only the fixed per-interval kg feed totals in Table 2); feed Protein/N/P/K composition (the feed product name reads ‘AllerFloat 37/10’ in the extracted text, which by Aller Aqua’s own naming convention conventionally denotes 37% crude protein/10% crude fat, but the paper itself never states this as a composition figure — recorded NR per the no-inference rule rather than assumed from the product-naming convention; flagged [unclear] rather than confidently extracted); Plants/m2 (paper gives plant count, 15/RAS, and tray dimensions, 45x30cm x5 trays, separately but never combines them into a stated density — computing one would be derivation); Plant height at harvest (only the pre-trial TRANSPLANT height, 42.1+/-4.3cm, is given — see Details — no harvest-time height is ever stated); Plant dry matter% (elemental-analysis samples were freeze-dried, but no dry-matter percentage of fresh weight is stated anywhere); Water type/Water classification (RAS water source and category never named); Lat/Long (facility named only as ‘Leibniz-Institute of Freshwater Ecology and Inland Fisheries, Berlin, Germany’ — no coordinates stated in the paper itself, and per CLAUDE.md’s prime directive coordinates are not looked up externally when the paper is silent); Average room Temperature (no ambient/room temperature distinct from the RAS/Hydro water temperatures in Table 3 is given, and the facility’s indoor/greenhouse status is not stated either); pHOptimal, WUE, FUE AP, FUE HYD (no stated optimum-pH target or efficiency metric matching these columns; the Discussion cites LITERATURE pH optima of 5.5-6.5 for plants and ~7 for nitrification, but these are secondary figures from cited sources [16,18], not this paper’s own targets, so left NR rather than repurposed here). | NO COLUMN: Table 1’s leaf/root/stem fresh-weight harvest totals over 154 d (leaf: HydroC 63.7 kg vs HydroD 48.0 kg; root: HydroC 5.8 kg vs HydroD 2.3 kg; stem: HydroC 25.7 kg vs HydroD 17.1 kg — i.e. MORE non-fruit biomass in the coupled system despite LESS fruit, Results p.5, attributed by the authors to suboptimal coupled-system nutrient supply driving a root-foraging/shoot:root acclimation response, Discussion p.11-12) — no dedicated leaf/root/stem yield columns exist alongside the single ‘Plant fresh weight’ (fruit) column. Table 1’s per-interval fertilizer-gram and water-consumption-litre schedule (both systems dosed identically per Table 1; HydroD water consumption ranged 1.4-2.4 L/plant/day, Results p.5; HydroC water consumption is not separately tracked since it is coupled to the RAS’s own volume, Table 1 caption). Table 2’s per-interval (rather than whole-trial) FCR/SGR/mortality/water-consumption/limestone breakdown (used here only as the source for the recorded total/average-row values). Table 4’s full water-mineral panel beyond NO3-N/NO2-N/TAN: PO4, K+, Ca2+, Mg2+, SO4(2-), Cl-, Fe2+ (mg/L) for both HydroC and HydroD across all 5 sampling intervals — no dedicated trials.csv columns exist for these ions. Table 5’s carbon content (%) and C/N ratio for leaf and tomato tissue in both systems (does not fit any of plant_measurements.csv’s four categories — biochemistry/mineral/microbiology/proximate — carbon content is neither a standard proximate component nor a mineral element; excluded from plant.csv per the same principle used for physiological data in other vault notes, e.g. mourantianBasilFunctionalGrowth2023). Table 5’s fish (RAS A-C-D pooled, n=4) and sludge (RAS C-D pooled, n=4) elemental composition (Ca/K/Mg/Na/P/N/C, C/N) — neither a plant analyte nor water chemistry, so it fits neither trials.csv nor plant_measurements.csv’s scope; preserved here only: fish (g/kg DW unless %): Ca 31.7(1.0), K 1.5(0.1), Mg 2.1(0.1), Na 0.7(0.0), P 17.7(0.5) g/kg; N 7.4(0.2)%, C 56.5(3.3)%, C/N 7.6(0.5); sludge: Ca 11.9(5.8), K 8.3(0.1), Mg 0.6(0.1), Na 3.5(0.1), P 8.9(2.8) g/kg; N 4.1(0.2)%, C 36.6(1.0)%, C/N 9.0(0.6). Table 6’s clarifier solids-removal data (RAS C 1.8+/-0.07 g DW/L vs RAS D 2.0+/-0.09 g DW/L weekly mean, n=5, 3 consecutive weeks) and Fig 3’s TSS time series (RAS A/C/D at day 30/94/154, n=3) — system-performance/waste-stream data with no matching trials.csv column. Fig 4’s estimated-nitrogen-fate schematic (modelled from this study’s results plus literature values for %N of protein, fish N excretion, nitrification and denitrification rates, and tomato nitrate uptake — a mixed primary/secondary construct, not read into any cell). Discussion’s cross-paper yield comparison: Kloas et al. 2015 (ref 5, a related prior study by an overlapping author group, first decoupled-aquaponics prototype) reported 8.89 kg/plant tomato yield over 9 months, vs this paper’s 8.2 kg/plant (HydroD) and 6.1 kg/plant (HydroC) over only 6 months — secondary comparison figure, see Citations to chase in the note. | UNIT CONVERSION ONLY: RAS total volume 16.5 m3 -> 16500 L; culture volume 6.8 m3 -> 6800 L (Methods p.2-3, identical figures restated from T1’s row since stated once for all three RAS); conductivity mS/cm is numerically identical to dS/m, a unit-name equivalence not an arithmetic conversion. | Water-quality trial-mean vs range note: as in T1, NO3-N/NO2-N/TAN have no stated single trial-mean for HydroD across the 154-d period (Table 4 gives only 5 discrete interval means); the min-max range across HydroD’s 5 intervals is recorded per SCHEMA.md’s ‘range only’ rule. HydroD’s NO3-N range (98.8-207.5 mg/L) matches the Results-text description (‘increased from 98.8 mg L-1 NO3-N to more than 170 mg L-1 from the third month on’, p.7) — the third-interval value (175.0 mg/L) is indeed the point at which the series crosses 170, confirming rather than conflicting with the table. | Yield cross-check (paper’s own arithmetic, not derived here): Discussion (p.12) states ‘a comparable tomato production of 8.2 kg plant-1 (HydroD) compared to 6.1 kg plant-1 (HydroC)’. Recomputing from Table 1’s own totals for confirmation only (kg total / 15 plants/RAS): HydroD 123.5 kg / 15 = 8.233 kg/plant (paper states 8.2, matches); HydroC 90.9 kg / 15 = 6.06 kg/plant (paper states 6.1, matches). Both check out; the paper’s own per-plant figures are used directly in the Plant fresh weight and AP cells (UNIT CONVERSION ONLY: kg/plant -> g/plant) rather than a value computed fresh here. Abstract’s ‘36% higher’ tomato yield in decoupled vs coupled also checks out against Table 1: (123.5-90.9)/90.9 = 35.9%, rounds to 36% — confirms, does not conflict.

Plant Measurements

TrialSystemCategoryAnalyteValueUnitSig.Location
monseesDecoupledSystemsTrial2017-T1CAPmineralCalcium (Ca) (leaf)30.4 ± 1.9g/kg DWNR (no inferential test reported anywhere in this paper)Table 5, day 63 (interval 07.05.-08.06.15)
monseesDecoupledSystemsTrial2017-T1CAPmineralPotassium (K) (leaf)45.4 ± 1.3g/kg DWNR (no inferential test reported anywhere in this paper)Table 5, day 63 (interval 07.05.-08.06.15)
monseesDecoupledSystemsTrial2017-T1CAPmineralMagnesium (Mg) (leaf)4.4 ± 0.2g/kg DWNR (no inferential test reported anywhere in this paper)Table 5, day 63 (interval 07.05.-08.06.15)
monseesDecoupledSystemsTrial2017-T1CAPmineralSodium (Na) (leaf)0.3 ± 0.0g/kg DWNR (no inferential test reported anywhere in this paper)Table 5, day 63 (interval 07.05.-08.06.15)
monseesDecoupledSystemsTrial2017-T1CAPmineralPhosphorus (P) (leaf)5.1 ± 0.2g/kg DWNR (no inferential test reported anywhere in this paper)Table 5, day 63 (interval 07.05.-08.06.15)
monseesDecoupledSystemsTrial2017-T1CAPmineralNitrogen (N) (leaf)3.4 ± 0.1% DWNR (no inferential test reported anywhere in this paper)Table 5, day 63 (interval 07.05.-08.06.15)
monseesDecoupledSystemsTrial2017-T1CAPmineralCalcium (Ca) (leaf)32.4 ± 3.0g/kg DWNR (no inferential test reported anywhere in this paper)Table 5, day 94 (interval 09.06.-09.07.15)
monseesDecoupledSystemsTrial2017-T1CAPmineralPotassium (K) (leaf)40.3 ± 7.3g/kg DWNR (no inferential test reported anywhere in this paper)Table 5, day 94 (interval 09.06.-09.07.15)
monseesDecoupledSystemsTrial2017-T1CAPmineralMagnesium (Mg) (leaf)4.8 ± 0.5g/kg DWNR (no inferential test reported anywhere in this paper)Table 5, day 94 (interval 09.06.-09.07.15)
monseesDecoupledSystemsTrial2017-T1CAPmineralSodium (Na) (leaf)0.3 ± 0.0g/kg DWNR (no inferential test reported anywhere in this paper)Table 5, day 94 (interval 09.06.-09.07.15)
monseesDecoupledSystemsTrial2017-T1CAPmineralPhosphorus (P) (leaf)4.4 ± 0.3g/kg DWNR (no inferential test reported anywhere in this paper)Table 5, day 94 (interval 09.06.-09.07.15)
monseesDecoupledSystemsTrial2017-T1CAPmineralNitrogen (N) (leaf)3.0 ± 0.2% DWNR (no inferential test reported anywhere in this paper)Table 5, day 94 (interval 09.06.-09.07.15)
monseesDecoupledSystemsTrial2017-T1CAPmineralCalcium (Ca) (leaf)26.0 ± 2.3g/kg DWNR (no inferential test reported anywhere in this paper)Table 5, day 122 (interval 10.07.-06.08.15)
monseesDecoupledSystemsTrial2017-T1CAPmineralPotassium (K) (leaf)35.3 ± 2.2g/kg DWNR (no inferential test reported anywhere in this paper)Table 5, day 122 (interval 10.07.-06.08.15)
monseesDecoupledSystemsTrial2017-T1CAPmineralMagnesium (Mg) (leaf)3.9 ± 0.2g/kg DWNR (no inferential test reported anywhere in this paper)Table 5, day 122 (interval 10.07.-06.08.15)
monseesDecoupledSystemsTrial2017-T1CAPmineralSodium (Na) (leaf)0.3 ± 0.0g/kg DWNR (no inferential test reported anywhere in this paper)Table 5, day 122 (interval 10.07.-06.08.15)
monseesDecoupledSystemsTrial2017-T1CAPmineralPhosphorus (P) (leaf)4.7 ± 0.3g/kg DWNR (no inferential test reported anywhere in this paper)Table 5, day 122 (interval 10.07.-06.08.15)
monseesDecoupledSystemsTrial2017-T1CAPmineralNitrogen (N) (leaf)3.2 ± 0.2% DWNR (no inferential test reported anywhere in this paper)Table 5, day 122 (interval 10.07.-06.08.15)
monseesDecoupledSystemsTrial2017-T1CAPmineralCalcium (Ca) (leaf)34.0 ± 3.6g/kg DWNR (no inferential test reported anywhere in this paper)Table 5, day 154 (interval 07.08.-07.09.15)
monseesDecoupledSystemsTrial2017-T1CAPmineralPotassium (K) (leaf)33.2 ± 3.2g/kg DWNR (no inferential test reported anywhere in this paper)Table 5, day 154 (interval 07.08.-07.09.15)
monseesDecoupledSystemsTrial2017-T1CAPmineralMagnesium (Mg) (leaf)3.8 ± 0.4g/kg DWNR (no inferential test reported anywhere in this paper)Table 5, day 154 (interval 07.08.-07.09.15)
monseesDecoupledSystemsTrial2017-T1CAPmineralSodium (Na) (leaf)0.4 ± 0.0g/kg DWNR (no inferential test reported anywhere in this paper)Table 5, day 154 (interval 07.08.-07.09.15)
monseesDecoupledSystemsTrial2017-T1CAPmineralPhosphorus (P) (leaf)4.3 ± 0.5g/kg DWNR (no inferential test reported anywhere in this paper)Table 5, day 154 (interval 07.08.-07.09.15)
monseesDecoupledSystemsTrial2017-T1CAPmineralNitrogen (N) (leaf)2.6 ± 0.3% DWNR (no inferential test reported anywhere in this paper)Table 5, day 154 (interval 07.08.-07.09.15)
monseesDecoupledSystemsTrial2017-T1CAPmineralCalcium (Ca) (fruit)2.2 ± 1.0g/kg DWNR (no inferential test reported anywhere in this paper)Table 5, day 63 (interval 07.05.-08.06.15)
monseesDecoupledSystemsTrial2017-T1CAPmineralPotassium (K) (fruit)47.5 ± 0.2g/kg DWNR (no inferential test reported anywhere in this paper)Table 5, day 63 (interval 07.05.-08.06.15)
monseesDecoupledSystemsTrial2017-T1CAPmineralMagnesium (Mg) (fruit)1.3 ± 0.1g/kg DWNR (no inferential test reported anywhere in this paper)Table 5, day 63 (interval 07.05.-08.06.15)
monseesDecoupledSystemsTrial2017-T1CAPmineralSodium (Na) (fruit)0.3 ± 0.0g/kg DWNR (no inferential test reported anywhere in this paper)Table 5, day 63 (interval 07.05.-08.06.15)
monseesDecoupledSystemsTrial2017-T1CAPmineralPhosphorus (P) (fruit)4.6 ± 0.2g/kg DWNR (no inferential test reported anywhere in this paper)Table 5, day 63 (interval 07.05.-08.06.15)
monseesDecoupledSystemsTrial2017-T1CAPmineralNitrogen (N) (fruit)2.0 ± 0.1% DWNR (no inferential test reported anywhere in this paper)Table 5, day 63 (interval 07.05.-08.06.15)
monseesDecoupledSystemsTrial2017-T1CAPmineralCalcium (Ca) (fruit)2.1 ± 0.3g/kg DWNR (no inferential test reported anywhere in this paper)Table 5, day 94 (interval 09.06.-09.07.15)
monseesDecoupledSystemsTrial2017-T1CAPmineralPotassium (K) (fruit)41.6 ± 2.5g/kg DWNR (no inferential test reported anywhere in this paper)Table 5, day 94 (interval 09.06.-09.07.15)
monseesDecoupledSystemsTrial2017-T1CAPmineralMagnesium (Mg) (fruit)1.4 ± 0.1g/kg DWNR (no inferential test reported anywhere in this paper)Table 5, day 94 (interval 09.06.-09.07.15)
monseesDecoupledSystemsTrial2017-T1CAPmineralSodium (Na) (fruit)0.2 ± 0.0g/kg DWNR (no inferential test reported anywhere in this paper)Table 5, day 94 (interval 09.06.-09.07.15)
monseesDecoupledSystemsTrial2017-T1CAPmineralPhosphorus (P) (fruit)4.3 ± 0.2g/kg DWNR (no inferential test reported anywhere in this paper)Table 5, day 94 (interval 09.06.-09.07.15)
monseesDecoupledSystemsTrial2017-T1CAPmineralNitrogen (N) (fruit)1.7 ± 0.2% DWNR (no inferential test reported anywhere in this paper)Table 5, day 94 (interval 09.06.-09.07.15)
monseesDecoupledSystemsTrial2017-T1CAPmineralCalcium (Ca) (fruit)1.3 ± 0.3g/kg DWNR (no inferential test reported anywhere in this paper)Table 5, day 122 (interval 10.07.-06.08.15)
monseesDecoupledSystemsTrial2017-T1CAPmineralPotassium (K) (fruit)41.0 ± 1.6g/kg DWNR (no inferential test reported anywhere in this paper)Table 5, day 122 (interval 10.07.-06.08.15)
monseesDecoupledSystemsTrial2017-T1CAPmineralMagnesium (Mg) (fruit)1.5 ± 0.0g/kg DWNR (no inferential test reported anywhere in this paper)Table 5, day 122 (interval 10.07.-06.08.15)
monseesDecoupledSystemsTrial2017-T1CAPmineralSodium (Na) (fruit)0.3 ± 0.0g/kg DWNR (no inferential test reported anywhere in this paper)Table 5, day 122 (interval 10.07.-06.08.15)
monseesDecoupledSystemsTrial2017-T1CAPmineralPhosphorus (P) (fruit)4.0 ± 0.5g/kg DWNR (no inferential test reported anywhere in this paper)Table 5, day 122 (interval 10.07.-06.08.15)
monseesDecoupledSystemsTrial2017-T1CAPmineralNitrogen (N) (fruit)2.0 ± 0.2% DWNR (no inferential test reported anywhere in this paper)Table 5, day 122 (interval 10.07.-06.08.15)
monseesDecoupledSystemsTrial2017-T1CAPmineralCalcium (Ca) (fruit)1.1 ± 0.1g/kg DWNR (no inferential test reported anywhere in this paper)Table 5, day 154 (interval 07.08.-07.09.15)
monseesDecoupledSystemsTrial2017-T1CAPmineralPotassium (K) (fruit)42.0 ± 4.4g/kg DWNR (no inferential test reported anywhere in this paper)Table 5, day 154 (interval 07.08.-07.09.15)
monseesDecoupledSystemsTrial2017-T1CAPmineralMagnesium (Mg) (fruit)1.5 ± 0.3g/kg DWNR (no inferential test reported anywhere in this paper)Table 5, day 154 (interval 07.08.-07.09.15)
monseesDecoupledSystemsTrial2017-T1CAPmineralSodium (Na) (fruit)0.3 ± 0.2g/kg DWNR (no inferential test reported anywhere in this paper)Table 5, day 154 (interval 07.08.-07.09.15)
monseesDecoupledSystemsTrial2017-T1CAPmineralPhosphorus (P) (fruit)4.4 ± 0.3g/kg DWNR (no inferential test reported anywhere in this paper)Table 5, day 154 (interval 07.08.-07.09.15)
monseesDecoupledSystemsTrial2017-T1CAPmineralNitrogen (N) (fruit)2.0 ± 0.5% DWNR (no inferential test reported anywhere in this paper)Table 5, day 154 (interval 07.08.-07.09.15)
monseesDecoupledSystemsTrial2017-T2DCAPmineralCalcium (Ca) (leaf)26.7 ± 4.3g/kg DWNR (no inferential test reported anywhere in this paper)Table 5, day 63 (interval 07.05.-08.06.15)
monseesDecoupledSystemsTrial2017-T2DCAPmineralPotassium (K) (leaf)39.9 ± 2.4g/kg DWNR (no inferential test reported anywhere in this paper)Table 5, day 63 (interval 07.05.-08.06.15)
monseesDecoupledSystemsTrial2017-T2DCAPmineralMagnesium (Mg) (leaf)3.9 ± 0.2g/kg DWNR (no inferential test reported anywhere in this paper)Table 5, day 63 (interval 07.05.-08.06.15)
monseesDecoupledSystemsTrial2017-T2DCAPmineralSodium (Na) (leaf)1.1 ± 0.1g/kg DWNR (no inferential test reported anywhere in this paper)Table 5, day 63 (interval 07.05.-08.06.15)
monseesDecoupledSystemsTrial2017-T2DCAPmineralPhosphorus (P) (leaf)2.7 ± 0.1g/kg DWNR (no inferential test reported anywhere in this paper)Table 5, day 63 (interval 07.05.-08.06.15)
monseesDecoupledSystemsTrial2017-T2DCAPmineralNitrogen (N) (leaf)3.9 ± 0.1% DWNR (no inferential test reported anywhere in this paper)Table 5, day 63 (interval 07.05.-08.06.15)
monseesDecoupledSystemsTrial2017-T2DCAPmineralCalcium (Ca) (leaf)23.1 ± 3.3g/kg DWNR (no inferential test reported anywhere in this paper)Table 5, day 94 (interval 09.06.-09.07.15)
monseesDecoupledSystemsTrial2017-T2DCAPmineralPotassium (K) (leaf)46.0 ± 0.9g/kg DWNR (no inferential test reported anywhere in this paper)Table 5, day 94 (interval 09.06.-09.07.15)
monseesDecoupledSystemsTrial2017-T2DCAPmineralMagnesium (Mg) (leaf)3.2 ± 0.3g/kg DWNR (no inferential test reported anywhere in this paper)Table 5, day 94 (interval 09.06.-09.07.15)
monseesDecoupledSystemsTrial2017-T2DCAPmineralSodium (Na) (leaf)1.3 ± 0.1g/kg DWNR (no inferential test reported anywhere in this paper)Table 5, day 94 (interval 09.06.-09.07.15)
monseesDecoupledSystemsTrial2017-T2DCAPmineralPhosphorus (P) (leaf)2.6 ± 0.4g/kg DWNR (no inferential test reported anywhere in this paper)Table 5, day 94 (interval 09.06.-09.07.15)
monseesDecoupledSystemsTrial2017-T2DCAPmineralNitrogen (N) (leaf)3.2 ± 0.1% DWNR (no inferential test reported anywhere in this paper)Table 5, day 94 (interval 09.06.-09.07.15)
monseesDecoupledSystemsTrial2017-T2DCAPmineralCalcium (Ca) (leaf)25.5 ± 2.8g/kg DWNR (no inferential test reported anywhere in this paper)Table 5, day 122 (interval 10.07.-06.08.15)
monseesDecoupledSystemsTrial2017-T2DCAPmineralPotassium (K) (leaf)36.0 ± 1.6g/kg DWNR (no inferential test reported anywhere in this paper)Table 5, day 122 (interval 10.07.-06.08.15)
monseesDecoupledSystemsTrial2017-T2DCAPmineralMagnesium (Mg) (leaf)4.0 ± 0.2g/kg DWNR (no inferential test reported anywhere in this paper)Table 5, day 122 (interval 10.07.-06.08.15)
monseesDecoupledSystemsTrial2017-T2DCAPmineralSodium (Na) (leaf)0.9 ± 0.1g/kg DWNR (no inferential test reported anywhere in this paper)Table 5, day 122 (interval 10.07.-06.08.15)
monseesDecoupledSystemsTrial2017-T2DCAPmineralPhosphorus (P) (leaf)2.9 ± 0.2g/kg DWNR (no inferential test reported anywhere in this paper)Table 5, day 122 (interval 10.07.-06.08.15)
monseesDecoupledSystemsTrial2017-T2DCAPmineralNitrogen (N) (leaf)3.8 ± 0.1% DWNR (no inferential test reported anywhere in this paper)Table 5, day 122 (interval 10.07.-06.08.15)
monseesDecoupledSystemsTrial2017-T2DCAPmineralCalcium (Ca) (leaf)26.7 ± 11.1g/kg DWNR (no inferential test reported anywhere in this paper)Table 5, day 154 (interval 07.08.-07.09.15)
monseesDecoupledSystemsTrial2017-T2DCAPmineralPotassium (K) (leaf)32.8 ± 7.5g/kg DWNR (no inferential test reported anywhere in this paper)Table 5, day 154 (interval 07.08.-07.09.15)
monseesDecoupledSystemsTrial2017-T2DCAPmineralMagnesium (Mg) (leaf)3.2 ± 0.9g/kg DWNR (no inferential test reported anywhere in this paper)Table 5, day 154 (interval 07.08.-07.09.15)
monseesDecoupledSystemsTrial2017-T2DCAPmineralSodium (Na) (leaf)0.7 ± 0.2g/kg DWNR (no inferential test reported anywhere in this paper)Table 5, day 154 (interval 07.08.-07.09.15)
monseesDecoupledSystemsTrial2017-T2DCAPmineralPhosphorus (P) (leaf)2.6 ± 0.5g/kg DWNR (no inferential test reported anywhere in this paper)Table 5, day 154 (interval 07.08.-07.09.15)
monseesDecoupledSystemsTrial2017-T2DCAPmineralNitrogen (N) (leaf)3.2 ± 0.4% DWNR (no inferential test reported anywhere in this paper)Table 5, day 154 (interval 07.08.-07.09.15)
monseesDecoupledSystemsTrial2017-T2DCAPmineralCalcium (Ca) (fruit)1.7 ± 0.2g/kg DWNR (no inferential test reported anywhere in this paper)Table 5, day 63 (interval 07.05.-08.06.15)
monseesDecoupledSystemsTrial2017-T2DCAPmineralPotassium (K) (fruit)45.6 ± 5.2g/kg DWNR (no inferential test reported anywhere in this paper)Table 5, day 63 (interval 07.05.-08.06.15)
monseesDecoupledSystemsTrial2017-T2DCAPmineralMagnesium (Mg) (fruit)1.2 ± 0.2g/kg DWNR (no inferential test reported anywhere in this paper)Table 5, day 63 (interval 07.05.-08.06.15)
monseesDecoupledSystemsTrial2017-T2DCAPmineralSodium (Na) (fruit)0.4 ± 0.0g/kg DWNR (no inferential test reported anywhere in this paper)Table 5, day 63 (interval 07.05.-08.06.15)
monseesDecoupledSystemsTrial2017-T2DCAPmineralPhosphorus (P) (fruit)3.7 ± 0.6g/kg DWNR (no inferential test reported anywhere in this paper)Table 5, day 63 (interval 07.05.-08.06.15)
monseesDecoupledSystemsTrial2017-T2DCAPmineralNitrogen (N) (fruit)2.1 ± 0.4% DWNR (no inferential test reported anywhere in this paper)Table 5, day 63 (interval 07.05.-08.06.15)
monseesDecoupledSystemsTrial2017-T2DCAPmineralCalcium (Ca) (fruit)1.3 ± 0.1g/kg DWNR (no inferential test reported anywhere in this paper)Table 5, day 94 (interval 09.06.-09.07.15)
monseesDecoupledSystemsTrial2017-T2DCAPmineralPotassium (K) (fruit)36.1 ± 3.9g/kg DWNR (no inferential test reported anywhere in this paper)Table 5, day 94 (interval 09.06.-09.07.15)
monseesDecoupledSystemsTrial2017-T2DCAPmineralMagnesium (Mg) (fruit)1.3 ± 0.1g/kg DWNR (no inferential test reported anywhere in this paper)Table 5, day 94 (interval 09.06.-09.07.15)
monseesDecoupledSystemsTrial2017-T2DCAPmineralSodium (Na) (fruit)0.5 ± 0.0g/kg DWNR (no inferential test reported anywhere in this paper)Table 5, day 94 (interval 09.06.-09.07.15)
monseesDecoupledSystemsTrial2017-T2DCAPmineralPhosphorus (P) (fruit)3.1 ± 0.6g/kg DWNR (no inferential test reported anywhere in this paper)Table 5, day 94 (interval 09.06.-09.07.15)
monseesDecoupledSystemsTrial2017-T2DCAPmineralNitrogen (N) (fruit)2.0 ± 0.2% DWNR (no inferential test reported anywhere in this paper)Table 5, day 94 (interval 09.06.-09.07.15)
monseesDecoupledSystemsTrial2017-T2DCAPmineralCalcium (Ca) (fruit)1.1 ± 0.4g/kg DWNR (no inferential test reported anywhere in this paper)Table 5, day 122 (interval 10.07.-06.08.15)
monseesDecoupledSystemsTrial2017-T2DCAPmineralPotassium (K) (fruit)40.5 ± 2.9g/kg DWNR (no inferential test reported anywhere in this paper)Table 5, day 122 (interval 10.07.-06.08.15)
monseesDecoupledSystemsTrial2017-T2DCAPmineralMagnesium (Mg) (fruit)1.3 ± 0.1g/kg DWNR (no inferential test reported anywhere in this paper)Table 5, day 122 (interval 10.07.-06.08.15)
monseesDecoupledSystemsTrial2017-T2DCAPmineralSodium (Na) (fruit)0.4 ± 0.1g/kg DWNR (no inferential test reported anywhere in this paper)Table 5, day 122 (interval 10.07.-06.08.15)
monseesDecoupledSystemsTrial2017-T2DCAPmineralPhosphorus (P) (fruit)3.0 ± 0.8g/kg DWNR (no inferential test reported anywhere in this paper)Table 5, day 122 (interval 10.07.-06.08.15)
monseesDecoupledSystemsTrial2017-T2DCAPmineralNitrogen (N) (fruit)2.0 ± 0.3% DWNR (no inferential test reported anywhere in this paper)Table 5, day 122 (interval 10.07.-06.08.15)
monseesDecoupledSystemsTrial2017-T2DCAPmineralCalcium (Ca) (fruit)1.2 ± 0.5g/kg DWNR (no inferential test reported anywhere in this paper)Table 5, day 154 (interval 07.08.-07.09.15)
monseesDecoupledSystemsTrial2017-T2DCAPmineralPotassium (K) (fruit)41.5 ± 2.8g/kg DWNR (no inferential test reported anywhere in this paper)Table 5, day 154 (interval 07.08.-07.09.15)
monseesDecoupledSystemsTrial2017-T2DCAPmineralMagnesium (Mg) (fruit)1.4 ± 0.1g/kg DWNR (no inferential test reported anywhere in this paper)Table 5, day 154 (interval 07.08.-07.09.15)
monseesDecoupledSystemsTrial2017-T2DCAPmineralSodium (Na) (fruit)0.4 ± 0.2g/kg DWNR (no inferential test reported anywhere in this paper)Table 5, day 154 (interval 07.08.-07.09.15)
monseesDecoupledSystemsTrial2017-T2DCAPmineralPhosphorus (P) (fruit)3.4 ± 0.6g/kg DWNR (no inferential test reported anywhere in this paper)Table 5, day 154 (interval 07.08.-07.09.15)
monseesDecoupledSystemsTrial2017-T2DCAPmineralNitrogen (N) (fruit)2.1 ± 0.4% DWNR (no inferential test reported anywhere in this paper)Table 5, day 154 (interval 07.08.-07.09.15)

Graph View

Backlinks