A Comprehensive Comparison of Lettuce Yield and Quality in Hydroponic and Aquaponic Systems: A Systematic Review and Meta-Analysis

Metadata

Cite key: tadesseComprehensiveComparisonLettuce2023
Item type: Journal Article
Authors: A. Tadesse, J. Avedzi, S. Asimenu
Affiliation: Debre Berhan University, College of Natural and Computational Sciences, Department of Biology, Ethiopia (Abebe Tadesse, corresponding author); University for Development Studies, Ghana (Justice Kofi Avedzi); National Peace Council, Ghana (Semi Asimenu)
Journal: Aquaponics 3(2) (2023) 1–22
Date: 07/2023
Date added: 2026-07-13
DOI: 10.5281/zenodo.8196953
Funding: [not reported]
URL: https://doi.org/10.5281/zenodo.8196953
PDF: Abebe Tadesse et al. - 2023 - A Comprehensive Comparison of Lettuce Yield and Quality in Hydroponic and Aquaponic Systems A Syste.pdf

Opinion

Do not cite this without independently re-deriving every number. The paper states a coherent, plausible-sounding PRISMA-style methodology (Web of Science/Scopus/Google Scholar search, 526→34→14 screening funnel, RevMan 5.4, random-effects SMD, GRADE, Cochrane RoB), but nearly every quantitative table contradicts the others on the most basic facts a meta-analysis must get right: how many studies were pooled (14? 20? — stated both ways repeatedly) and what the headline hydroponic-vs-aquaponic yield effect size actually is (SMD reported as 1.04, 0.34, 0.33, and two different single-arm “1.36 vs 1.12” and “0.86 vs 1.02” pairs across five different tables, none reconciling with any other). The five named “included studies” in Table 2 (Smith et al. 2018, Lee and Kim 2019, Garcia et al. 2020, Wang et al. 2021, Jones et al. 2022) do not appear anywhere in the paper’s own 18-item reference list, meaning the specific primary studies this meta-analysis claims to have pooled cannot be traced or verified at all. The Discussion and Conclusion sections repeat near-identical paragraphs two and three times, consistent with unedited AI-generated boilerplate. Combined with the outlet (a Zenodo-archived journal called “Aquaponics” at afri.et, malformed DOI string on its own masthead), this reads as a fabricated or auto-generated paper rather than a genuine systematic review. Useful only as a pointer to its reference list for chasing real primary studies (several of which — Kloas et al. 2015, Huang et al. 2020 — are plausible real papers worth checking against the vault); none of its own numbers should be trusted or propagated.

Abstract

Hydroponic and aquaponic systems have emerged as sustainable alternatives to traditional soil-based agriculture. In this systematic review and meta-analysis, we comprehensively compared lettuce yield and quality in hydroponic and aquaponic systems. We analyzed data from 20 peer-reviewed articles and found that lettuce yield and quality varied significantly across different hydroponic and aquaponic systems. Our meta-analysis revealed that hydroponic systems had higher average lettuce yields than aquaponic systems, with a standardized mean difference (SMD) of 1.04 (95% CI: 0.59-1.50). However, lettuce quality was comparable between the two systems, with no significant difference in total phenolic content (SMD: -0.05; 95% CI: -0.32-0.21) and antioxidant activity (SMD: 0.06; 95% CI: -0.22-0.34). Our findings suggest that while hydroponic systems may be more efficient for lettuce yield, aquaponic systems can provide comparable lettuce quality while also generating fish and other aquatic species as a secondary source of income. However, the variability in system design, nutrient management, and environmental conditions highlights the need for further research and standardization of these systems for optimal yield and quality.

Summary

The authors claim to have systematically searched Web of Science, Scopus, and Google Scholar for studies comparing lettuce yield and quality between hydroponic and aquaponic systems, screening 526 records down to a final pooled set (stated inconsistently as either 14 or 20 studies) and pooling standardized mean differences with a random-effects model in RevMan 5.4. The stated top-line conclusion is that hydroponic systems produce higher lettuce yields than aquaponic systems, while lettuce quality metrics (phenolic content, antioxidant activity, vitamin C, nitrate) show no consistent significant difference between the two system types. In practice, the paper’s own tables cannot agree with each other on the number of studies pooled, the resulting effect sizes, or even which quality outcomes were assessed, and the handful of specific “included studies” it names in Table 2 do not correspond to any entry in its reference list. It is only useful as an entry point to its list of cited primary and secondary aquaponics/hydroponics lettuce literature, not as a source of any extractable summary statistic.

Experiment data

Location: Not a single site — synthesis of studies stated to be from the USA, Australia, the Netherlands (Study Characteristics, p.5), with a location subgroup table (Table 5, p.9) covering USA, South Korea, Spain, and China instead — these two lists of countries do not overlap (see Extraction notes, WARN-CHECK).

Design: Systematic review and meta-analysis; databases Web of Science, Scopus, Google Scholar; hand-search of reference lists; search conducted March 2023 (p.3). Random-effects model, SMD with 95% CI, I² heterogeneity, Egger’s regression for publication bias, Cochrane RoB / ROBINS-I / GRADE for study quality (p.3-4).

Replicates / n: UNCLEAR — see WARN-BLOCK below. Stated as “20 studies” (Abstract, p.1; Results, p.4) and separately as “14 articles … included in the meta-analysis” (Search Results, p.5).

Duration: Search conducted March 2023 (p.3). Eligible publication window stated three different ways: “published between 2010 and 2021” (Methods, p.3), “published in or after 2000” (Inclusion Criteria, p.3), and “conducted between 2006 and 2021” (Study Characteristics, p.5).

Organisms: Lettuce (Lactuca sativa) — varieties named in Table 2/5 include Buttercrunch, Romaine, Lollo Rosso, Iceberg, Cos.

Statistics: Random-effects meta-analysis, standardized mean difference (SMD), 95% CI, I² heterogeneity statistic, Egger’s regression test for publication bias, GRADE framework, Review Manager (RevMan) 5.4 software (p.3-4).

Lettuce yield: Reported as SMD 1.04 (95% CI 0.59-1.50) in one place and SMD 0.34 (95% CI 0.03-0.65, p=0.03) in another, among other values — see WARN-BLOCK, unresolved.

Lettuce quality (phenolics/antioxidant/vitamin C/nitrate): No consistent significant difference reported between systems across the various (mutually inconsistent) effect-size tables — see WARN-BLOCK.

Lettuce yield

This paper: The headline claim is that hydroponic lettuce yield exceeds aquaponic lettuce yield. The actual magnitude of that difference cannot be determined — the paper reports it at least four incompatible ways (full evidence under Extraction notes, WARN-BLOCK). No single defensible SMD for hydroponic-vs-aquaponic lettuce yield can be recorded from this paper.

Compared with:

todo Kloas et al. 2015 — advanced aquaponics vs. conventional hydroponics for tomato (not lettuce, but same comparative design); already in this paper’s reference list, not yet confirmed in vault.
todo Huang et al. 2020 — nutrient concentration comparison, lettuce, hydroponic vs. aquaponic (Sustainability 12(16), 6473); in this paper’s reference list.

Lettuce quality (phenolics, antioxidant activity, vitamin C, nitrate)

This paper: Abstract and one Results paragraph (p.4-5) state no significant difference in total phenolic content (SMD -0.05, 95% CI -0.32 to 0.21) or antioxidant activity (SMD 0.06, 95% CI -0.22 to 0.34) between hydroponic and aquaponic lettuce. A separate Results paragraph (p.5, headed “Meta-Analysis”) instead reports total phenolic content at SMD -0.15 (95% CI -0.49 to 0.19, p=0.38), and introduces nitrate concentration (SMD -0.05, 95% CI -0.43 to 0.32, p=0.79) and vitamin C concentration (SMD 0.10, 95% CI -0.25 to 0.44, p=0.58) as the quality outcomes instead of antioxidant activity, which is dropped entirely from this second account. See WARN-BLOCK in Extraction notes — which set of quality outcomes and effect sizes belongs to which study subset cannot be determined.

Compared with:

todo Yavuzcan Yildiz et al. 2021 — comparative growth/yield/quality study, lettuce, hydroponic vs aquaponic (Scientia Horticulturae 279, 109881); cited in reference list with a truncated URL (“https” with nothing after it, p.21) — worth verifying independently.

Linked claims

Citations to chase

Extraction notes

Classification: meta-analysis. The title says “Systematic Review and Meta-Analysis,” and Methods (p.3) states an explicit search strategy, databases (Web of Science, Scopus, Google Scholar), search terms, and inclusion/exclusion criteria (Table 1, p.6) — satisfying the systematic-review test. However, the paper goes further and performs quantitative pooling of effect sizes (SMD, random-effects model, RevMan 5.4, I² heterogeneity, sensitivity analyses, Table 4-11) — the defining feature of meta-analysis, which takes precedence per SCHEMA.md’s decision rules (never “upgrade,” but this genuinely does pool). The authors collected no original data of their own (no Materials/Methods for growing lettuce, no original measurements) — confirmed secondary research. Per SCHEMA.md, meta-analysis gets a note only; trials.csv and plant.csv for this paper contain header rows only, deliberately, to avoid double-counting whatever primary studies it claims to summarize (which, per the finding below, cannot even be identified).

WARN-BLOCK Number of studies pooled in the meta-analysis. Abstract (p.1): “We analyzed data from 20 peer-reviewed articles.” Results, first paragraph (p.4): “Our search identified 20 studies that met the inclusion criteria, including a total of 53 hydroponic and 31 aquaponic systems.” Results, “Search Results” subsection (p.5): “Our initial search identified 526 articles. After removing duplicates and screening the titles and abstracts, we reviewed the full text of 34 articles. Of these, 14 articles met our inclusion criteria and were included in the meta-analysis.” Table 8 (p.12) and Table 9 (p.14) both use “20” studies / “5,000” lettuce-plant participants as their baseline for the yield outcome; Table 11 (p.16) uses “20” studies / “4,500” participants. There is no basis to prefer 14 over 20 or vice versa — the “14” figure comes with an internally plausible PRISMA-style screening funnel (526→34→14) while “20” appears in the abstract, the first Results paragraph, and three separate summary tables with three different (mutually inconsistent) participant counts (5,000 in Tables 8/9, 4,500 in Table 11). Recorded: UNCLEAR / not extractable. Affects: every effect size in the paper, since none states unambiguously which study set it was computed from.

WARN-BLOCK Hydroponic vs. aquaponic lettuce yield effect size. Abstract (p.1) and Results paragraph 1 (p.4): SMD = 1.04 (95% CI: 0.59-1.50), I²=99%, tied to the “20 studies, 53 hydroponic + 31 aquaponic systems” figure. Results, “Meta-Analysis” subsection (p.5): SMD = 0.34 (95% CI: 0.03 to 0.65; p=0.03), I²=70%, tied to the “14 included studies” figure (“Sensitivity analysis showed that excluding low-quality studies did not substantially change the results”). Table 4 (p.8): effect size 0.33 (95% CI 0.05, 0.61), 10 studies, hydroponic mean 142.5±54.8 g/plant vs aquaponic mean 129.4±44.7 g/plant, I²=49% (close to the 0.34 value but computed from a stated “10 studies,” not 14 or 20). Table 8 (p.12): reports two separate single-arm “effect sizes” instead of one comparative SMD — “Yield in hydroponic system” 1.36 (1.18, 1.56), I²=34%, GRADE High; “Yield in aquaponic system” 1.12 (0.97, 1.29), I²=45%, GRADE Moderate — both listed against “20 studies, 5,000 participants.” Table 11 (p.16): yet another pair of single-arm figures — hydroponic 0.86 (0.76, 0.96), aquaponic 1.02 (0.92, 1.12) — against “20 studies, 4,500 participants.” These five values/pairs (1.04; 0.34; 0.33; 1.36-vs-1.12; 0.86-vs-1.02) cannot be reconciled by any consistent definition (they aren’t, e.g., raw vs. bias-adjusted, or different subgroups stated as such) — the paper presents each as the answer to the same question in different places. Recorded: UNCLEAR / not extractable — no single value can be treated as this paper’s stated yield effect size. Affects: the paper’s own headline conclusion (“hydroponic systems had higher average lettuce yields”) is directionally consistent across all five readings (all show hydroponic > aquaponic, i.e., all positive/>1), so the direction of the finding is not in dispute — only its magnitude, which varies roughly 4-fold (0.33-1.36) depending which table is read.

WARN-BLOCK Lettuce quality effect sizes / which outcomes were even assessed. Abstract (p.1) and Results paragraph 2 (p.4-5, verbatim match): total phenolic content SMD -0.05 (95% CI -0.32 to 0.21); antioxidant activity SMD 0.06 (95% CI -0.22 to 0.34) — these two are internally consistent with each other. Results, “Meta-Analysis” subsection (p.5): a different set of three outcomes — nitrate concentration SMD -0.05 (95% CI -0.43 to 0.32, p=0.79), vitamin C concentration SMD 0.10 (95% CI -0.25 to 0.44, p=0.58), total phenolic content SMD -0.15 (95% CI -0.49 to 0.19, p=0.38) — antioxidant activity is not mentioned at all in this account, and the total phenolic content value itself differs (-0.05 vs -0.15, non-overlapping point estimates though CIs overlap). Table 4 (p.8) adds a third, again different, set: leaf area, chlorophyll content (SPAD), and nitrate content (mg/kg, not SMD-scaled) as the quality-adjacent outcomes, with nitrate content given as SMD 0.76 (95% CI 0.16, 1.35) against 5 studies — a fourth value for “nitrate,” incompatible with the -0.05 figure in the “Meta-Analysis” paragraph. No single, internally consistent list of “what quality outcomes this meta-analysis measured, and what each SMD was” can be recovered. Recorded: UNCLEAR for all quality SMDs. Affects: the paper’s secondary conclusion (“no significant difference in lettuce quality”) — every version of the quality numbers happens to be non-significant (CIs cross zero) except Table 4’s nitrate content (0.76, 95% CI 0.16-1.35, which DOES cross into significance and contradicts the “no significant difference” narrative) — so even the qualitative conclusion is not fully consistent across the paper’s own tables.

WARN-MATERIAL Table 2 “included studies” absent from the paper’s own reference list. Table 2 (p.7) names five specific “included studies” with full characteristics: Smith et al. (2018, USA, RCT, Hydroponic, Buttercrunch, n=30); Lee and Kim (2019, South Korea, Observational, Aquaponic, Romaine, n=40); Garcia et al. (2020, Spain, Quasi-experimental, Hydroponic, Lollo Rosso, n=24); Wang et al. (2021, China, RCT, Aquaponic, Iceberg, n=50); Jones et al. (2022, Australia, Observational, Hydroponic, Cos, n=12). None of “Smith,” “Lee and Kim,” “Garcia,” “Wang et al. (2021)” (a different “Wang, Liu, Chen (2018)” exists in the reference list, but not a 2021 Wang et al.), or “Jones” appear anywhere in the 18-entry References section (p.20-22). The reference list’s actual entries (Akrivos 2018, Atkinson 2014, Goddek 2018, Graamans 2019, Huang 2020, Kloas 2015, Rakocy 2004, Somerville 2014, Turan & Şen 2020, Yavuzcan Yildiz 2021, Swiader & Beaulieu 2016, Torregrosa-Muñumer 2018, Trochoulias 2019, Vojtíšek 2017, Wang/Liu/Chen 2018, Xu/Xu/Chen 2019, Yousuf 2019, Zou & Hu 2018) share no author names with Table 2, 3, or 7’s named studies. This means the specific primary studies this meta-analysis claims to have screened and quality-assessed cannot be verified as real, cited works. Not correctable — recorded as a standing data-integrity concern, not an extractable field. Affects: overall credibility of every table in the paper (Tables 2, 3, 4, 5, 6, 7, 8, 9, 10, 11 all present numbers attributed to “the included studies,” none traceable to a real citation).

WARN-CHECK Eligible publication-year window stated three ways. Methods (p.3): “peer-reviewed articles published between 2010 and 2021.” Inclusion Criteria (p.3), criterion 5: “be published in or after 2000.” Study Characteristics (p.5): “The 14 included studies were conducted between 2006 and 2021.” These could in principle be different things (a priori search filter vs. actual inclusion criterion vs. observed range of the final set), but the paper never states which definition applies where, and 2000/2010/2006 are not nested in a way that makes obvious sense (the a priori Methods filter of 2010+ is stricter than the Inclusion Criteria’s 2000+, yet the “actual” included studies are then reported as starting in 2006 — before the Methods filter’s own 2010 cutoff). Recorded: none of the three values entered any cell (no CSV row produced for this paper); all three given here for the reader’s judgment.

WARN-CHECK Location/country subgroup does not sum to either study-count figure. Study Characteristics (p.5): “conducted in various countries including the United States, Australia, and the Netherlands.” Table 5 (p.9), location subgroup for the yield analysis: USA (3 studies), South Korea (2), Spain (1), China (2) — totaling 8 studies, none of which are Australia or the Netherlands, and 8 does not equal the 10 “studies compared lettuce yield” stated in Study Characteristics, nor 14, nor 20. Recorded: none of these location/count figures entered any cell.

WARN-MINOR Duplicated Discussion/Conclusion text. The Conclusion section (p.19) contains the sentence “Ultimately, the choice of system may depend on the specific goals and resources of the grower” and its accompanying two-sentence elaboration twice in immediate succession, and a near-identical “Our systematic review and meta-analysis provides evidence that hydroponic systems may have higher yields…” paragraph three times across pp.18-19 with only minor wording changes. Does not affect any extracted value; noted only as evidence of poor editorial quality / likely unedited AI-generated boilerplate (feeds into the Opinion assessment above).

WARN-MINOR Author byline order vs. citation-line/Zotero order. The paper’s own author byline (p.1, affiliations block) lists “Justice Kofi Avedzi¹, Abebe Tadesse², Semi Asimenu³” (Avedzi first). The paper’s own “How to cite” line (p.1 footer) and the Zotero export both list the order as “Abebe Tadesse, Justice Kofi Avedzi, & Semi Asimenu” (Tadesse first). Per CLAUDE.md, Zotero is authoritative for author order, so Tadesse is recorded as first author and used for the citekey. Does not affect any data field, noted for completeness. Author name segmentation into “F. Surname” wikilinks is a judgment call: “Abebe Tadesse,” “Justice Kofi Avedzi,” and “Semi Asimenu” all follow given-name-plus-patronymic conventions rather than Western first/last-name structure; A. Tadesse, J. Avedzi, and S. Asimenu are used here (last word of each name treated as the wikilink surname) but may not match how these authors are indexed elsewhere in the vault or literature — flagging so a consistent form can be chosen if these authors recur.

[not reported] fields: Funding (no funding statement or acknowledgments section anywhere in the PDF). Specific fish species used in any of the reviewed aquaponic systems (never named — the paper only ever says “aquaponic systems,” never a species). Page-level location of the individual primary-study data extraction form used by the two reviewers (described narratively, p.3, but the completed form/dataset is not included; “Data Availability” states it is “available upon request”).

New tags introduced: Meta/Type/Meta-analysis (title-cased per existing Meta/Type/Experiment convention seen elsewhere in the vault); Meta/Region/Global (new — no existing note covers a multi-country secondary-research synthesis; the paper itself cannot even state a consistent country list, see WARN-CHECK above); Meta/Plant/Lettuce (new — the entire paper’s analytic subject is lettuce, not a passing mention, so this passes the “studied it” bar even though the data is secondary).

Source: Abebe Tadesse et al. - 2023 - A Comprehensive Comparison of Lettuce Yield and Quality in Hydroponic and Aquaponic Systems A Syste.pdf

Quartz 5

Explorer

A Comprehensive Comparison of Lettuce Yield and Quality in Hydroponic and Aquaponic Systems: A Systematic Review and Meta-Analysis

A Comprehensive Comparison of Lettuce Yield and Quality in Hydroponic and Aquaponic Systems: A Systematic Review and Meta-Analysis

Metadata

Abstract

Summary

Lettuce yield

Lettuce quality (phenolics, antioxidant activity, vitamin C, nitrate)

Linked claims

Citations to chase

Extraction notes

Graph View

Table of Contents

Backlinks