Quartz 5

Nutrient load estimation in the waste of Nile tilapia Oreochromis niloticus (L.) reared in cages in tropical climate conditions

Properties1
tagsmeta/type/modelling, meta/region/south-america, meta/fish/tilapia

20 min read

Nutrient load estimation in the waste of Nile tilapia Oreochromis niloticus (L.) reared in cages in tropical climate conditions

Metadata

  • Cite key: montanhiniNutrientLoadEstimation2015
  • Item type: Journal Article
  • Authors: R. Montanhini Neto, A. Ostrensky
  • Affiliation: Integrated Group for Aquaculture and Environmental Studies, Department of Animal Sciences, Federal University of Parana, Curitiba, Parana, Brazil
  • Journal: Aquaculture Research 46(6) (2015) 1309-1322
  • Date: 06/2015 (print issue); online first 09/2013
  • Date added: [not reported]
  • DOI: 10.1111/are.12280
  • Funding: [not reported] — Acknowledgments thank the Service of Inspection of Livestock Input, Ministry of Agriculture/Livestock/Supply (Parana state), for access to feed-registration records, and the authors’ own research group for field-data access; no funding body or grant number is named anywhere in the paper.
  • URL: https://doi.org/10.1111/are.12280
  • PDF: Montanhini Neto and Ostrensky - 2015 - Nutrient load estimation in the waste of Nile tila.pdf

Opinion

A carefully constructed nutrient mass-balance estimation study, not a live aquaponics or even a live fish trial — there is no plant component anywhere in the paper, and the fish side is not a designed experiment either. It stitches together three independent data sources (a registry-wide survey of the chemical composition/digestibility of 130 commercially registered tilapia feeds; a meta-analysis of production-performance data from 73 published studies merged with field data from 17 commercial cage farms; and the authors’ own laboratory proximate/mineral analysis of 10 harvest-weight fish) and runs them through explicit mass-balance equations (Eqs. 1-4) to estimate how much organic matter, nitrogen and phosphorus end up in the waste stream per tonne of tilapia produced. The arithmetic is unusually easy to audit and it all checks out: Tables 5-7’s three waste fractions (feeding losses, faeces, soluble excretion) sum exactly to Table 8’s totals for every nutrient, and every stated deposition-efficiency percentage recomputes cleanly from Table 8’s own rows. The one rough edge is Table 4, where the printed meta-analysis mean for ash appears truncated in the source text (“40.” with no decimal digits, unlike every other value in that table) — flagged below rather than guessed at. Because there is no aquaponic or hydroponic treatment, no plant, and no manipulated fish treatment, this paper produces a note only; it earns its place in this vault as a parametrization source for fish-waste nutrient-partitioning fractions (feed-loss %, digestibility %, N/P retention vs. excretion split), which the vault already cites elsewhere for exactly that purpose (see Extraction notes).

Abstract

We sought to estimate the nutrient load in the waste released into aquatic environments based on the feeding of Nile tilapia (Oreochromis niloticus, L.) reared in cages that were installed in artificial reservoirs. For the calculation, an analysis of the chemical composition of commercial feeds intended for this species in their various stages of production was conducted (N = 130). We combined this information with a meta-analysis of published data from commercial producers in Brazil about expected feed intake, feed conversion and other animal production indices, and body composition. With these data, it was possible to estimate the load. We estimated that 18% of the feed given to the animals is not consumed and is lost in the aquatic environment. The calculated average digestibility was 71.97% for the organic matter in the diet, 84.06% for protein and 54.40% for phosphorus. The estimated nutrient deposition efficiency, with respect to what was actually consumed by the tilapia, was 26.39% for organic matter, 43.25% for protein and 34.07% for phosphorus. The total nutrient load in the waste per tonne of biomass of produced tilapias was estimated to be 1040.63 kg of organic matter, 44.95 kg of nitrogen and 14.26 kg of phosphorus, representing 78%, 65% and 72% of the respective nutrient amounts supplied by the feed. The information obtained in this study serves as a reference for predicting the potential impact of tilapia farming in reservoirs and to establish scientific parameters for the planning of this activity.

Summary

The authors combined three data streams to build a stage-by-stage (fry/juvenile/growth/termination) nutrient mass-balance model for caged Nile tilapia farmed in Brazilian artificial reservoirs: a chemical-composition and digestibility survey of all 130 commercial tilapia feeds registered with Brazil’s Ministry of Agriculture in Parana state; a meta-analysis of production-performance indices from 73 published studies (2002-2012, restricted to 600-900 g harvest weight) merged with field data from 17 commercial cage farms on the Paranapanema River; and the authors’ own laboratory proximate/mineral analysis (AOAC methods) of 10 harvest-weight fish, compared against a separate 26-study literature meta-analysis of tilapia body composition. Using explicit mass-balance equations, they partitioned feed nutrients into three waste fractions — uneaten feed, the indigestible (faecal) fraction, and soluble/endogenous excretion — and estimated that 18% of feed is never consumed, digestibility averages 72% (organic matter), 84% (protein) and 54% (phosphorus), and that per tonne of harvested tilapia biomass roughly 1041 kg organic matter, 45 kg nitrogen and 14 kg phosphorus reach the environment as waste (78%, 65% and 72% of what was supplied via feed, respectively). Because the laboratory body-composition values differed significantly (Student’s t-test, p<0.05) from the literature meta-analysis values for ash, protein/nitrogen, calcium and phosphorus, with lower variability in the lab data, the authors used their own laboratory measurements rather than the meta-analysis values as the body-composition input to the final load calculations. There is no aquaponic or hydroponic component, no plant of any kind, and no manipulated fish treatment (density, diet, or system design) anywhere in the study — it is a literature-and-registry-parametrized computational estimate, not an empirical trial, framed for pond/reservoir cage aquaculture rather than recirculating or aquaponic systems.

Experiment data

  • Location: Feed-composition data: commercial tilapia feeds registered with MAPA (Ministry of Agriculture, Livestock and Supply) in Parana state, Brazil. Field validation data: commercial cage farms in the Paranapanema River channel, on the Parana/Sao Paulo state border, southern Brazil. Laboratory body-composition analysis: Animal Nutrition Laboratory, Federal University of Parana (UFPR). No physical experimental system (no tank, cage, or aquaponic unit) was built or run by the authors for this study.
  • Design: Not a manipulated experiment. Three parallel data-synthesis streams feed into an explicit nutrient mass-balance model (Eqs. 1-4: feeding-loss, faecal/indigestible, and soluble-excretion fractions, summed to a total waste load per nutrient): (1) chemical-composition/digestibility calculation for N=130 registered commercial feeds across 4 production stages (fry/juvenile/growth/termination), using Optimix v4.1 software per Furuya (2010), NRC (2011) and Rostagno (2011); (2) a systematic-review-based meta-analysis of production-performance indices from 73 published articles (keywords “production,” “Oreochromis niloticus,” “cages”; harvest weight 600-900 g; 2002-2012) merged with field data from 17 commercial Paranapanema River cage farms; (3) the authors’ own laboratory proximate/mineral analysis (AOAC 2012 methods) of 10 sexually-reversed tilapia (600-900 g) vs. a separate 26-article literature meta-analysis of body composition, compared by Student’s t-test.
  • Replicates / n: N=130 commercial feeds (chemical composition/digestibility); n=73 published articles + n=17 commercial farms (production-performance meta-analysis/field data); n=26 published articles (body-composition meta-analysis) vs. n=10 fish (own laboratory assay).
  • Duration: [not reported] as a single figure — the model spans the full fry-to-termination production cycle (initial weight 8.65 g to final 791.36 g) but no total cycle length in days is stated anywhere in the paper, only per-stage weight trajectories (Table 3).
  • Organisms: Nile tilapia (Oreochromis niloticus, L.), sexually reversed, harvest weight 600-900 g. No plant organism — this paper has no hydroponic or aquaponic component at all.
  • Statistics: Confidence intervals (alpha=90%) for the waste-load estimates (Tables 5-8); Student’s t-test (p<0.05) comparing laboratory vs. meta-analysis body-composition means (Table 4). Software: Statistica v8.0 (StatSoft Inc.) for descriptive statistics.
  • Feed digestibility: organic matter 71.97%, protein 84.06% (highest), phosphorus 54.40%/54.41% (lowest) — see Extraction notes for the minor abstract/results rounding difference.
  • Feeding losses: 18.0% weighted average (fry 30.8%, juveniles 17.0%, growth 17.6%, termination 16.7%).
  • Total nutrient load per tonne tilapia biomass: 1040.63 kg organic matter (78% of feed input), 44.95 kg nitrogen (65%), 14.26 kg phosphorus (72%); OM:N:P ratio 72.98:3.15:1.00.

Feed composition and digestibility

This paper: Across the 130 registered commercial feeds (fry n=32, juvenile n=30, growth n=38, termination n=30), crude protein declined sharply across stages (399.6 g/kg fry to 240.8 g/kg termination), as did total/digestible calcium and phosphorus, while crude fibre and non-nitrogen (carbohydrate) content rose with stage. Digestible fractions were calculated per stage from digestibility coefficients (Furuya 2010; NRC 2011; Rostagno 2011); phosphorus was the least digestible nutrient (54.40-54.41%, see Extraction notes) and protein the most digestible (84.06%). Indigestible-fraction values (Table 2) were cross-checked here against Table 1 (total minus digestible) and matched exactly for organic matter and phosphorus in every stage, confirming internal arithmetic consistency (recomputation only, not a new cell).

Compared with:

  • todo Bechara et al. 2005 — pacu (Piaractus mesopotamicus); found high-quality feed reduces pond pollution potential and can raise per-area production disproportionately to cost, cited as supporting the case for feed-quality improvement as a load-reduction lever (Discussion, p.17).
  • todo Amirkolaie 2011 — review on reducing aquaculture waste through feed/feeding; phytase supplementation cited as improving mineral retention (Ca, P) and reducing excretion (Discussion, p.17).
  • todo Xie, Cui, Yang & Liu 1997b — Nile tilapia fry; found a negative correlation between nutrient digestibility and ration size, cited as underscoring the importance of feed processing/production practice for environmental outcomes (Discussion, p.17).

Tilapia production performance and feeding losses

This paper: The pooled performance meta-analysis (73 articles) plus field data (17 Paranapanema River farms) gave initial fry weight 8.65 +/- 4.01 g and final (termination) weight 791.36 +/- 95.18 g, with stage-chained weights (each stage’s final weight = next stage’s initial weight, internally consistent). Total-cycle apparent feed conversion averaged 1.639 +/- 0.305 and true feed conversion 1.345 +/- 0.272 (these total-cycle figures differ from, and are not a simple average of, the four per-stage FCR values also given in Table 3 — both are the paper’s own stated figures, not in conflict, just different levels of aggregation). Feeding losses (the gap between apparent and true feed conversion) were highest at the fry stage (30.8%) and declined to 16.7% by termination, weighting to an 18.0% overall average — the paper explicitly frames this as consistent with Pearson & Gowen’s (1990) commonly cited ~20% feed-loss benchmark for caged fish generally.

Compared with:

  • todo Pearson and Gowen 1990 — general caged-fish feed loss benchmark, ~20%, cited as consistent with this paper’s 18.0% estimate (Discussion, p.17).
  • todo Islam 2005 — notes commercial feeding practice can waste up to 30x more feed than laboratory settings, cited as context for why controlled-environment waste studies likely underestimate real losses (Discussion, p.17).
  • todo Guerrero 1980 and Jauncey 1998 — tilapia’s relatively slow feeding behaviour (vs. other farmed species) cited as a mechanistic reason for elevated feed loss in cages (Discussion, p.17).

Body composition: laboratory assay vs. literature meta-analysis

This paper: The authors’ own AOAC-method laboratory analysis of 10 harvest-weight (600-900 g) tilapia was compared against a separate 26-article literature meta-analysis of tilapia body composition (Table 4). Student’s t-tests found significant differences (p<0.05) for ash, crude protein/nitrogen, calcium and phosphorus, with the laboratory data showing markedly lower variability than the pooled literature values. The authors explicitly chose to parametrize their nutrient-load model with the laboratory values rather than the meta-analysis values on this basis (p.6). See Extraction notes for a data-quality flag on the printed ash meta-analysis value.

Compared with:

  • todo Xie, Zheng, Chen, Zhang, Zhu & Yang 2011 — Nile tilapia; nitrogen retention efficiency highest at 28C, lowest at 37C, cited as one explanation for body-composition variability across the literature meta-analysis (Discussion, p.18).

Nutrient load estimates and comparison with other cage-farmed species

This paper: Combining the three data streams via the mass-balance equations (Eqs. 1-4), the paper estimates a total waste load per tonne of tilapia biomass of 1040.63 kg organic matter (78% of feed-supplied OM), 44.95 kg nitrogen (65% of feed-supplied N) and 14.26 kg phosphorus (72% of feed-supplied P), split into feeding losses (18% of input for all three nutrients), the indigestible/faecal fraction (23% OM, 13% N, 37% P of input) and soluble excretion (37% OM, 33% N, 17% P of input). These loads are two- to ten-fold higher than most other cage-aquaculture literature the authors cite, which they attribute to Nile tilapia’s high metabolic rate and to using commercially formulated (rather than experimental/optimized) Brazilian feeds.

Compared with:

  • todo Guo and Li 2003 — cage-cultured fish generally; found nutrient deposition rates of only 15% (N) and 11% (P), versus this paper’s much higher estimated retention (35% N, 28% P of feed input into body); the paper attributes the difference to tilapia’s high metabolic/excretion rate (Discussion, p.18).
  • todo Ackefors and Enell 1994 — caged salmonids; loads of up to 10 kg P and 60 kg N per tonne produced, and ~2500 kg organic matter per tonne (wet basis) — lower P and higher OM than this paper’s tilapia estimates (Discussion, p.18).
  • todo Fernandes et al. 2007 — caged southern bluefin tuna; estimated 88-93% of feed N deposited as waste (59-64% as soluble excretion) and 75-90 kg N/tonne, both somewhat higher than this paper’s tilapia figures (65% N waste, 44.95 kg N/tonne) (Discussion, p.18).
  • todo Hakanson 2005 — general cage fish farming; estimated 10-20 kg P and 50-75 kg N per tonne produced, a range the authors note as matching their own 14.26 kg P / 44.95 kg N estimates closely (Discussion, p.18-19).
  • todo Penczak, Galicka, Molinski, Kusto and Zalewski 1982 — cited for the general estimate that only ~32% of feed phosphorus is used in fish metabolism, broadly consistent with this paper’s ~28-34% P retention figures (Discussion, p.18).
  • todo Alves and Baccarin 2005 — intensive pond feeding; found 66% of deposited phosphorus goes to sediment, 11% stays dissolved, 23% is incorporated into fish — a different three-way partition than this paper’s feeding-loss/faeces/soluble-excretion split but broadly comparable retention (23% vs. this paper’s 28%) (Discussion, p.18).
  • todo Sara 2007 — meta-analysis of ecological effects of aquaculture on dissolved water-column nutrients, cited as corroborating that fish-cage waste loads are ecologically significant and under-studied (Discussion, p.16, p.18).

Relevance to aquaponics (why this note-only fish paper is in the vault)

This paper has no plant, hydroponic, or aquaponic component whatsoever — it is a pure open-reservoir cage-aquaculture nutrient-budget study, framed entirely around environmental/eutrophication impact of tilapia farming in Brazilian reservoirs. Its relevance to this vault is as a parametrization source for fish-waste nutrient-partitioning fractions (feed-loss %, digestibility %, and the split of consumed-but-not-retained nutrients between faeces and soluble/dissolved excretion) that aquaponics system designers need to size biofilters, remineralization units, or sludge-handling components. The vault already cites this exact paper for this purpose: notes/goddekNavigatingDecoupledAquaponic2016.md uses “Neto and Ostrensky 2013” (same paper, cited there under its online-first year — see Extraction notes) to parametrize a sludge/ANRC nutrient-partitioning model (Table 5 of that paper), comparing this paper’s literature-derived N/P sludge-partitioning values against its own RAS “personal observations.” That existing citation can now be resolved to this full note.

Linked claims

Citations to chase

  • todo Pearson TH, Gowen RJ (1990) — Impact of caged farming on the marine environment, in Interaction Between Aquaculture and Environment — source of the ~20% general feed-loss benchmark cited here.
  • todo Islam MS (2005) — Nitrogen and phosphorus budget in coastal and marine cage aquaculture, Marine Pollution Bulletin 50:48-61 — commercial vs. laboratory feed-waste comparison.
  • todo Guo L, Li Z (2003) — Effects of nitrogen and phosphorous from fish cage-culture on the communities of a shallow lake, Aquaculture 226:201-212 — much lower N/P deposition rates than this paper’s estimates.
  • todo Ackefors H, Enell M (1994) — The release of nutrients and organic matter from aquaculture systems in Nordic countries, Journal of Applied Ichthyology 10:225-241 — salmonid cage nutrient loads.
  • todo Fernandes M, Lauer P, Cheshire A, Angove M (2007) — Preliminary model of nitrogen loads from southern bluefin tuna aquaculture, Marine Pollution Bulletin 54:1321-1332 — tuna cage nitrogen mass-balance model, same general method as this paper.
  • todo Hakanson L (2005) — Changes to ecosystem structure resulting from fish cage farm emissions, Lake and Reservoir Management 10:71-80 — general cage-farming N/P load range, closely matches this paper’s estimates.
  • todo Penczak T, Galicka W, Molinski M, Kusto E, Zalewski M (1982) — The enrichment of a mesotrophic lake by carbon, phosphorus and nitrogen from cage aquaculture of rainbow trout, Journal of Applied Ecology 19:371-393.
  • todo Alves RCP, Baccarin AL (2005) — Efeitos da producao de peixes em tanques-rede sobre sedimentacao de material em suspensao e de nutrientes, in Ecologia de Reservatorios — phosphorus partitioning in intensive pond feeding.
  • todo Sara G (2007) — A meta-analysis on the ecological effects of aquaculture on the water column: dissolved nutrients, Marine and Environmental Research 63:390-408.
  • todo Bechara JA, Roux JP, Diaz FJR, Quintana CIF, De Meabe CAL (2005) — Effect of dietary protein level on pond water quality and feed utilization efficiency of pacu, Aquaculture Research 36:546-553.
  • todo Amirkolaie AK (2011) — Reduction in the environmental impact of waste discharged by fish farms through feed and feeding, Reviews in Aquaculture 3:19-26.
  • todo Xie S, Zheng K, Chen J, Zhang Z, Zhu X, Yang Y (2011) — Effect of water temperature on energy budget of Nile tilapia, Aquaculture Nutrition 17:683-690.

Extraction notes

Type classification (modelling), and why it is not experiment, meta-analysis, or review: No fish were subjected to a designed/manipulated treatment (no stocking-density, diet, or system comparison), so this is not experiment, quasi-experiment, field-trial, or exploratory. The paper is not purely secondary either — the authors collected genuinely original data themselves: a registry-wide chemical-composition/digestibility calculation for N=130 commercially registered feeds (their own compilation and Optimix-software calculation, not a value taken from another paper), and a laboratory proximate/mineral assay (AOAC 2012 methods) of n=10 tilapia performed at their own institution (UFPR), explicitly preferred over the literature meta-analysis values on statistical grounds (Table 4, p.6). This original data is combined with two separate literature meta-analyses (73 articles on production performance, 26 articles on body composition) and field data from 17 commercial farms, all fed into explicit mass-balance equations (Eqs. 1-4, p.4-5: FLn, FCn, SEn, TLWn) that compute the paper’s headline outputs (Tables 5-8) — outputs that were never directly measured as a whole by any single method. This satisfies the modelling test in SCHEMA.md (“simulation or computational model”) better than any other category, with the caveat (per SCHEMA.md’s mixed-paper rule) that it also contains a genuine meta-analysis component and a small amount of first-hand laboratory data; both are noted here rather than up-classifying the paper to experiment (no experimental design exists) or down-classifying it to meta-analysis/review (original data disqualifies both, per the prime-directive test “did the authors collect data themselves?”).

No CSV rows produced. Per SCHEMA.md, only experiment, quasi-experiment, field-trial, and exploratory papers produce trials.csv/plant_measurements.csv rows; modelling does not (CLAUDE.md: “Reviews and meta-analyses get a note only,” and SCHEMA.md’s “Which types produce CSV rows” list excludes modelling). out/montanhiniNutrientLoadEstimation2015.trials.csv and .plant.csv were written with header rows only, zero data rows — verified with a Python csv-module field count (87 and 11 columns respectively). This is also consistent with the paper having no aquaponic/hydroponic system, no AP/HYD comparison, and no plant of any kind to extract into either file.

⚠️MINOR — Phosphorus average digestibility, Abstract vs. Results, p.1 vs p.5: Abstract: “54.40% for phosphorus.” Results (p.5, following Table 2): “the one that showed the lowest average digestibility was phosphorus, at 54.41%.” Trivial rounding-level discrepancy (0.01 percentage points); no interpretive consequence, and no cell is affected since this paper produces no CSV rows.

⚠️MINOR — Table 4 (p.6) ash meta-analysis mean rendered as truncated/garbled “40.”: unlike every other value in Table 4 (which are consistently printed to 2 decimal places, e.g. “693.61,” “192.06,” “30.73”), the ash meta-analysis column reads “40. (+/-13.18)” with no digits after the decimal point in the extracted text — anomalous formatting suggesting digits were lost, likely a PDF-extraction or original-typesetting artifact rather than a value the authors actually printed as “40.” flat. As a cross-check only (not entered as a fact): the Discussion text (p.6) states ash showed a “21% difference between methods,” and recomputing (meta-analysis mean / laboratory mean) using the printed laboratory value 33.32 gives 33.32 x 1.21 = ~40.3, consistent with the family of other rows’ 2-decimal-place values (e.g. a value like 40.3x) rather than exactly 40.00. The same recomputation cleanly reproduces the paper’s other three stated ”% difference” figures (protein/nitrogen ~25%, calcium ~41%, phosphorus ~39%) using the fully legible values in the same table, which is why the ash cell specifically is flagged as suspect rather than assumed correct as printed. Recorded here as [unclear]/garbled rather than guessed; does not affect any extracted cell since this paper has no CSV rows. Anyone citing this paper’s exact laboratory-vs-literature ash values should check the original PDF table image directly.

Internal arithmetic verified (not a flag, a confidence note): Tables 5, 6 and 7 (uneaten feed, faeces, and soluble excretion, respectively) sum exactly to Table 8’s “Total waste load” row for all three tracked nutrients (organic matter: 236.12+306.29+498.22=1040.63; nitrogen: 12.62+9.08+23.25=44.95; phosphorus: 3.56+7.40+3.30=14.26). Table 8’s stated deposition-efficiency percentages also recompute cleanly from its own rows (e.g. body deposition N / feed intake N = 24.64/56.97 = 43.25%, matching the abstract’s “43.25% for protein” figure; body deposition N / feed input N = 24.64/69.59 = 35.41%, matching the “considering the supplied feed” restatement). Table 1’s indigestible fractions (Table 2) also match total-minus-digestible exactly for organic matter and phosphorus in every production stage. This is unusually clean internal consistency for a paper combining three separate data sources, and materially increases confidence in the reported final figures despite the ash-value flag above.

Calcium omitted from the final Table 8 mass balance. The Discussion text (p.6) states calcium deposition efficiency (36.96% of intake, 30.30% of feed input), and Tables 5-7 each include a calcium row, but Table 8’s final three-nutrient balance (organic matter, nitrogen, phosphorus only) does not carry calcium through to a “total waste load” figure. Not a contradiction — simply an incomplete final table relative to the fuller per-fraction tables — noted here since a reader might otherwise expect a calcium total-load estimate that the paper does not actually state as a single number.

Citation-year cross-reference: the vault already cites this exact paper (same DOI, 10.1111/are.12280) as [[Neto and Ostrensky 2013]] in notes/goddekNavigatingDecoupledAquaponic2016.md (both in its “Nutrient and sludge partitioning” section and its “Citations to chase” list), using the Crossref online-first publication year (2013-09-05). This note instead uses year 2015, matching the print journal issue (Aquaculture Research 46(6), Crossref published-print 2015-06) and the source PDF’s own header (“Aquaculture Research, 2015, 46, 1309-1322”) and filename. Both years refer to the identical article; flagged here so the existing “Neto and Ostrensky 2013” citation-to-chase entries can be resolved against this note (Paper ID montanhiniNutrientLoadEstimation2015) rather than treated as a separate uncatalogued source.

Tags judgment calls: Meta/Fish/Tilapia applied — Nile tilapia (Oreochromis niloticus) is the sole organism studied, both in the authors’ own laboratory assay and in the literature searches (both meta-analyses were explicitly restricted to this species). Meta/Region/South-America applied — all data sources (feed registry, field farms, laboratory) are located in Parana state and the Parana/Sao Paulo border region of Brazil. No Meta/Plant/ tag — the paper has no plant, hydroponic, or aquaponic component of any kind; it studies open-reservoir cage aquaculture exclusively.

PDF quality: clean, fully extractable text layer; no OCR artifacts observed apart from the isolated Table 4 ash-value truncation noted above; no pages missing or garbled; all 14 pages (main text + references) were read in full.

[not reported] fields, grouped (note-only, no CSV cells affected):

  • Total production-cycle duration in days (only per-stage initial/final weights given, Table 3; no single cycle-length figure stated)
  • Zotero “Date added” (no zotero-export.csv present for this batch)
  • Exact confidence-interval calculation method/distributional assumption behind the stated alpha=90% intervals (only “we determined the lower and upper limits” is stated, p.4)
  • Whether the meta-analysis literature search used any database names beyond “academic article search portals” (no named databases, unlike a fully specified systematic review)

Source: Montanhini Neto and Ostrensky - 2015 - Nutrient load estimation in the waste of Nile tila.pdf

Graph View

Backlinks