henhaobio
Sign in

PLoS ONEMolecular BiologyPeer-reviewed journal article

eDNA metabarcoding as a biomonitoring tool for marine protected areas

Zachary J. GoldiD, Joshua L. Sprague, David J. KushneriD, Erick Zerecero Marin, Paul Henry BarberiD

PDFDOIPreprint
In-depth full-text digest available · Jump to digest ↓

Abstract

Monitoring of marine protected areas (MPAs) is critical for marine ecosystem management, yet current protocols rely on SCUBA-based visual surveys that are costly and time consuming, limiting their scope and effectiveness. Environmental DNA (eDNA) metabarcoding is a promising alternative for marine ecosystem monitoring, but more direct comparisons to visual surveys are needed to understand the strengths and limitations of each approach. This study compares fish communities inside and outside the Scorpion State Marine Reserve off Santa Cruz Island, CA using eDNA metabarcoding and underwater visual census surveys. Results from eDNA captured 76% (19/25) of fish species and 95% (19/20) of fish genera observed during pairwise underwater visual census. Species missed by eDNA were due to the inability of MiFish 12S barcodes to differentiate species of rockfishes (Sebastes, n=4) or low site occupancy rates of crevice-dwelling Lythrypnus gobies. However, eDNA detected an additional 30 fish species not recorded in paired visual surveys, but previously reported from prior visual surveys, highlighting the sensitivity of eDNA. Significant variation in eDNA signatures by location (50m) and site (~1000m) demonstrates the sensitivity of eDNA to address key questions such as community composition inside and outside MPAs. Interestingly, eDNA results recorded higher species richness outside the MPA while visual surveys observed the opposite pattern. This result is likely caused by swamping effects of high fish abundance in MPAs that reduce detection probabilities of pelagic and intertidal taxa. Results demonstrate the utility of eDNA metabarcoding for monitoring marine ecosystems, providing an important complementary tool to visual methods.

In-depth summary

Claude Opus 5.5 · max effort · 8 Nov 2026 · from the full text of version 1 (20 Aug 2020) · CC0 1.0

eDNA metabarcoding at a California marine reserve recovered 76% of diver-recorded fish species, distinguished communities 50 m apart, yet reversed the visual inside-versus-outside richness pattern.

Marine protected area (MPA) monitoring relies on costly, depth-limited SCUBA censuses of selected taxa, so only 94 of over 1000 Channel Islands reefs are surveyed, once a year. Because eDNA metabarcoding's biases are less characterized than those of visual surveys, this side-by-side field test measures its concordance with diver censuses, spatial resolution, replication needs, and cost (~$600 per site vs ~$1,200 for roving diver counts).

Key results

  1. ShowneDNA recovered most fish recorded by the paired diver censuses. Every miss was either a rockfish that the MiFish 12S marker cannot split to species (four Sebastes) or a rarely detected crevice-dwelling Lythrypnus goby.

    19 of 25 visually recorded species (76%) and 19 of 20 genera (95%) detected; missed: Sebastes atrovirens, S. auriculatus, S. chrysomelas, S. serranoides, Lythrypnus dalli, L. zebra; Lythrypnus appeared before occupancy filtering in only one replicate at two sites. · Abstract; Comparison of eDNA and Visual Census Surveys; Limitations and Caveats of eDNA

  2. ShowneDNA gave a much broader fish inventory than the paired diver surveys, adding 31 species the divers did not record. Of these, 30 were already known from earlier monitoring-program surveys; the dog-faced witch eel had never been logged by that program.

    After decontamination, 931 ASVs and 2.35 million reads gave 54 fish taxa (50 species, 48 genera, 34 families) vs 25 species, 20 genera, 13 families by visual census; eDNA-only taxa included endangered giant black seabass (Stereolepis gigas) and yellowtail amberjack (Seriola lalandi). · eDNA Results; Visual Census Surveys Results; Comparison of eDNA and Visual Census Surveys; The utility of eDNA for MPA monitoring

  3. ShowneDNA community composition differed significantly among the inside, edge, and outside sites and among sampling locations roughly 50 m apart along each transect. Location explained more variance than site.

    PERMANOVA on Bray-Curtis eDNA-index distances (Site + Location): p<0.001, betadisper p>0.05; location 26.4%, site 19.0%, unexplained 54.5% of variance; NMDS stress 0.20 (weak clustering); CAP p<0.001 with seven discriminating taxa, e.g., kelp perch higher inside and yellowtail amberjack higher outside. · eDNA Results; Discussion; The utility of eDNA for MPA monitoring

  4. ShowneDNA and divers disagreed on where richness peaked: eDNA found the most species outside the reserve, whereas divers recorded the most inside. The authors favor an explanation they did not directly test: abundant reef fish inside the MPA dilute eDNA carried in from pelagic, sandy, and intertidal habitats.

    eDNA: 42 species outside, 34 edge, 35 inside (ANOVA p<0.001; Levene's test p>0.5); visual: 21 inside, 18 edge, 13 outside; non-rocky-reef taxa (46.4%, 13/28) partly drove site differences; outside samples had ~50% more read depth; unresolved Sebastes were seen inside (n=3) and at the edge (n=1). · eDNA Results; Comparison of eDNA and Visual Census Surveys; Diversity inside and outside MPAs

  5. ShownSampling 9 x 1 L bottles per site did not capture all species. For species with occupancy above 75%, three bottles at each of three transect locations gave near-complete coverage. Species accumulation switched from exponential to linear between three and four bottles.

    All-species coverage 94.0%, 88.0%, 92.9% (MPA, edge, outside) with 9 samples vs 81.0%, 80.0%, 83.6% with three; occupancy above 75%: >99.0% with 9 vs 87.1%, 90.3%, 88.9% with three; segmented regression breakpoint at 3.36-3.53 replicates. · eDNA Results; Limitations and Caveats of eDNA

Methods, data and reproducibility

On 100 m monitoring transects inside, at the edge of, and 2.3 km outside Scorpion State Marine Reserve, 9 x 1 L seawater samples per site were Sterivex-filtered, MiFish 12S metabarcoded, and classified with the Anacapa Toolkit; occupancy-filtered detections were analyzed with rarefaction, PERMANOVA and constrained ordination and compared with diver censuses.

  • Filtered 1 L seawater eDNA samples from 100 m transects inside, at the edge of (<0.5 km), and 2.3 km outside Scorpion State Marine Reserve, Santa Cruz Island, CA · marine fish community (multi-species eDNA) · n = 27 samples (9 per site) plus 8 controls
  • Negative controls (field blanks of 1 L distilled water; molecular-grade water PCR blanks) and positive-control fish DNA extracts · Ctenopharyngodon idella and Salmo salar (positive controls) · n = 3 field blanks
  • Kelp Forest Monitoring Program SCUBA visual censuses (fish transects, roving diver counts, 1 m^2 quadrats) on the same transects within 2 weeks of eDNA sampling · bony fishes · n = 3 sites; 25 species recorded
  • Historical Kelp Forest Monitoring Program fish records used to vet species detected only by eDNA (Table S7) · California fishes
  • Per-site cost tallies for roving diver fish counts (Table S8) and the eDNA sampling design (Table S9)
Methods & tools
4 L Niskin bottle sampling with Sterivex 0.22 μm gravity filtration, DNAeasy Tissue and Blood Kit (Qiagen), modified per Spens et al. (2017), MiFish Universal Teleost 12S primers with Nextera modifications, MiSeq PE 2×300 bp sequencing (Reagent Kit V3), Anacapa Toolkit, FishCARD California fish reference database, CRUX-generated 12S reference database, Decontamination per Kelly et al. (2018) and McKnight et al. (2019); eDNA index, Site occupancy modeling, iNext rarefaction; R package segmented (piecewise regression), ANOVA, adonis PERMANOVA, BETADISPER, NMDS and CAP on Bray-Curtis distances, Kelp Forest Monitoring Program visual census protocols (Kushner et al. 2013)
Reagents & environment
Anacapa Toolkit (default parameters), iNext, and R package segmented, with no versions stated; reagents: DNAeasy Tissue and Blood Kit (Qiagen), Nextera-modified MiFish 12S primers, grass carp and Atlantic salmon DNA as positive controls.
Assay conditions & compute
4 L Niskin casts (10 m below surface; 1 m above benthos) from UCLA RV Kodiak; 1 L gravity-filtered through 0.22 μm Sterivex, filters held at −20°C; MiSeq PE 2×300 bp, Reagent Kit V3, 20% PhiX at a UCLA genomics core; no computing hardware reported.
Reproducibility gaps
Data only promised for Dryad, with a link after acceptance; no accession, repository URL, or analysis code is given, and species tables, cost tables, and PCR/library protocols are deferred to supplementary materials.
Limitations (6)
  • The MiFish 12S marker cannot separate Sebastes rockfishes, a recent radiation with conserved 12S sequence, which made up four of the six missed species; this matters for California MPA monitoring because rockfishes are key fishery and ecological taxa. · stated by the authors
  • Small crevice-dwelling Lythrypnus gobies went undetected at species level, possibly from primer mismatch (two insertions in L. dalli) or eDNA retained in the reef boundary layer; boundary-layer sampling may be needed for cryptic benthic fishes. · stated by the authors
  • Replication was limited: rarefaction indicated more samples would add species, three bottles from one location covered only 81.5% of species with at least 75% occupancy, and occupancy filtering discarded kelp forest species such as swell shark and California angel shark. · stated by the authors
  • eDNA metrics need cautious interpretation: outside samples carried ~50% more read depth despite standardized pooling, abundance could not be compared with only 3 sites, and eDNA biomass estimation remains unresolved, so the authors frame eDNA as complementing diver surveys. · stated by the authors
  • Inside, edge, and outside contrasts each rest on one transect sampled on a different day (up to 72 hours apart), with diver surveys up to 2 weeks away, so protection status is confounded with site and date and no replicate reserves test MPA effects. · noted in this summary
  • Coverage benchmarks conflict within the text: three-bottle coverage for species above 75% occupancy appears as 87.1%, 90.3%, 88.9% (Results), 81.5% (Discussion), and 89.3-91.1% (Figure 3 legend), clouding recommended replication levels. · noted in this summary

Read this if you plan eDNA surveys of reefs or marine reserves and want field benchmarks on concordance with diver censuses, replication and occupancy filtering, and pitfalls in comparing species richness.

Citations
245
FWCI
17.68
Field percentile
top 1%

Reproducibility ledger

Nobody has reported trying to reproduce this paper yet. Did you replicate an experiment or run the code? Your report saves the next lab a week.

Sign in to file a reproducibility report.

Related papers

Similarity Map
SelectedSame fieldCross-field
THISValdivia‐Ca '19Waters '26Fraija-Fern '19Marshall '26Curto '25Oka '20
81% semantic matchGenomics
Tania Valdivia‐Carrillo, Axayácatl Rocha‐Olivares, Héctor Reyes‐Bonilla+2 more·bioRxiv·Oct 16, 2019· 6 citations