Propagule and Juvenile-derived Foraminiferal eDNA across intertidal habitats and its implications for accurate sea-level reconstruction

This study demonstrates that while bulk sediment foraminiferal eDNA provides reliable sea-level reconstructions in mangrove environments, the inclusion of propagule and juvenile-derived eDNA in the <63 m fraction can lead to elevation underprediction in mangroves and overprediction in transitional zones, highlighting the need for size-fraction awareness in sea-level proxy applications.

The Gist

The Big Picture: Reading the Ocean's "Receipt"

Imagine you are a detective trying to figure out how high the ocean used to be 1,000 years ago. You can't just ask the water; you have to look at the clues left behind in the mud.

For decades, scientists have used tiny, single-celled creatures called foraminifera (or "forams" for short) as their main clue. These creatures live in specific zones of the beach: some like the high tide, some the low tide, and some the middle. By finding their empty shells in ancient mud, scientists can guess the sea level.

The New Twist: Recently, scientists started using eDNA (environmental DNA). Instead of looking for the creature's shell, they look for the creature's "genetic receipt" (DNA) floating in the mud. It's like finding a receipt in a trash can instead of the actual product. It's faster and finds more species, but it has a catch: the trash can might contain receipts from things that never actually lived there.

The Problem: The "Party Crashers"

This paper asks a critical question: Does the DNA in the mud come from the creatures living right there, or is it just "party crashers" that washed in from somewhere else?

Forams have a life stage called propagules (think of them as microscopic seeds or babies). These babies are tiny, swim around, and can travel far.

• The Analogy: Imagine a neighborhood (the beach).
  • Adult Forams are the residents who live in specific houses.
  • Propagules are the kids running around the whole neighborhood, visiting every house.
  • eDNA is the DNA left behind on the sidewalk.

If you pick up DNA from the sidewalk, is it from the person who lives in that house, or is it from a kid who just ran through from three blocks away? If you assume it's the resident, you might think the "neighborhood" is different than it really is.

The Experiment: Sorting the Mud

The researchers went to Hong Kong's Mai Po Nature Reserve, a place with mangroves (trees in the water) and mudflats (bare mud). They took mud samples and did something like a sieve test:

1. Big Sieve (500–63 microns): Caught the "adults" (the shells and big DNA).
2. Tiny Sieve (<63 microns): Caught the "babies" and "seeds" (propagules and tiny DNA).
3. No Sieve (Bulk): The whole messy mix.

They then compared the DNA in these different piles to see if the "babies" were messing up the data.

The Findings: It Depends on the Neighborhood

1. The Mangrove Forest (The Quiet Neighborhood)

• What happened: In the mangroves, the trees slow down the water and trap organic matter. This acts like a sticky net.
• The Result: The DNA found in the mud mostly belongs to the creatures actually living there. Even the "baby" DNA didn't travel far because the environment kept it local.
• The Verdict: Using eDNA in mangroves is very reliable. It's like looking at a receipt in a house where the owner actually lives. You can trust it to tell you the sea level.

2. The Mudflat (The Busy Highway)

• What happened: The mudflats are open, windy, and have fast currents.
• The Result: The "baby" DNA (propagules) is everywhere. It washes in from deep water, from other parts of the bay, and from deep underground. The DNA in the mud is a smoothie of DNA from everywhere, not just the local residents.
• The Verdict: Using eDNA here is tricky. If you just look at the tiny DNA, you might think the sea level is higher or lower than it actually is because the "party crashers" are confusing the data.

3. The Transition Zone (The Border)

• The Result: This is the messiest area. It's a mix of the quiet forest and the busy highway. The DNA here is the most unreliable for guessing sea levels because it's a chaotic mix of local and non-local DNA.

The Conclusion: How to Read the Receipt

The paper concludes that eDNA is a fantastic tool, but you have to know where you are using it.

• In Mangroves: Go ahead! The eDNA is accurate and gives a clear picture of the past sea level.
• In Mudflats/Transitional Zones: Be careful. If you don't account for the "traveling babies" (propagules), your sea-level reconstruction will be wrong. You might need to filter out the tiny DNA or use different methods to get the right answer.

In short: The ocean leaves us clues, but sometimes those clues are written by tourists rather than locals. This study teaches us how to tell the difference so we can accurately reconstruct our planet's history.

Technical Summary

1. Problem Statement

While foraminiferal environmental DNA (eDNA) has emerged as a robust proxy for reconstructing relative sea level (RSL), its application faces a critical knowledge gap: the influence of non-adult life stages on assemblage composition.

• The Issue: Unlike traditional morphological methods that rely on dead adult tests (typically 63–500 µm), eDNA captures DNA from diverse sources, including extracellular DNA, propagules (dispersal stages), and juveniles.
• The Risk: Propagules and juveniles are highly mobile and can disperse rapidly across intertidal gradients. Their DNA may broaden the apparent elevation range of eDNA assemblages, introducing bias and uncertainty into RSL reconstructions.
• The Gap: It remains unclear how much propagule/juvenile-derived DNA contributes to total eDNA pools in different environments (mangroves vs. mudflats) and how this affects the accuracy of Bayesian transfer function (BTF) models used for sea-level estimation.

2. Methodology

The study employed a comparative approach using size-fractionated sediment samples from subtropical intertidal environments in Hong Kong.

• Study Area: Mai Po Nature Reserve, Deep Bay, Pearl River Estuary. Three stations were selected along an intertidal gradient:
  1. Upper-mudflat (transitional zone).
  2. Mid-mangrove.
  3. Upper-mangrove.
• Sampling Strategy: Surface sediments were collected in November 2023 (dry season). At each station, replicate samples were collected for:
  • eDNA Analysis: Sediments were wet-sieved into three fractions:
    • Bulk: Unfractionated sediment.
    • 500–63 µm: Enriched for adult foraminifera.
    • <63 µm: Enriched for propagules, juveniles, and small species.
  • Morphological Analysis: Parallel samples were processed for traditional counting of dead and living (Rose Bengal stained) foraminifera in the 63–500 µm fraction.
• Molecular Workflow:
  • DNA extraction from all fractions.
  • Metabarcoding of the 18S rDNA fragment (135–190 bp) using foraminifera-specific primers.
  • Sequencing on the NovaSeq platform (PE150).
  • Bioinformatic processing (denoising, OTU clustering at 97%, taxonomic assignment via BLAST).
• Statistical & Modeling Approach:
  • Community Analysis: ANOVA, NMDS (Non-metric Multidimensional Scaling), and PERMANOVA to assess compositional differences between fractions and stations.
  • RSL Reconstruction: A Bayesian Transfer Function (BTF) model (based on a modern training set from Liu et al., 2025a) was applied to estimate elevation (SWLI) for all size fractions and bulk samples. Results were compared against observed elevations and morphological estimates.

3. Key Contributions

• Differentiation of eDNA Sources: The study explicitly isolates the contribution of propagule/juvenile-derived DNA (<63 µm fraction) from adult-derived DNA (500–63 µm fraction) to assess their distinct impacts on community structure.
• Environment-Specific Bias Quantification: It demonstrates that the reliability of bulk eDNA for RSL reconstruction is environment-dependent, varying significantly between mangrove forests and mudflat/transitional zones.
• Refinement of Proxy Application: The paper provides practical guidelines for when bulk sediment eDNA is sufficient versus when size-fractionation or caution is required to avoid reconstruction errors.

4. Key Results

A. Community Composition and Sources

• Dominance of Soft-Walled Taxa: Saccamminidae (soft-walled monothalamids) dominated eDNA assemblages across all fractions (59–80% in the <63 µm fraction).
• Mangrove vs. Mudflat Dynamics:
  • Mangroves: Bulk eDNA assemblages closely resembled the combined signals of both size fractions, suggesting a strong contribution from in-situ adult populations and accumulated extracellular DNA (protected by high organic carbon and low pH).
  • Mudflats: Bulk eDNA closely mirrored the <63 µm fraction, indicating a dominance of propagule- and juvenile-derived DNA. This is attributed to lower organic carbon (faster degradation), higher sedimentation rates, and the transport of allochthonous DNA from deeper estuarine environments.
• Taxonomic Shifts: The <63 µm fraction in mangroves showed a systematic underrepresentation of high-elevation indicator taxa (e.g., Miliamminidae), while mudflat fractions showed high mobility of taxa not present in local morphological assemblages (e.g., planktonic Globigerinitidae).

B. RSL Reconstruction Accuracy (BTF Modeling)

• Mangrove Environments:
  • Bulk and 500–63 µm fractions: Provided accurate elevation estimates within observed ranges.
  • <63 µm fraction: Systematically underpredicted elevation (though within 2σ uncertainty), likely due to the absence of high-elevation indicator taxa in the propagule pool during the dry season.
• Mudflat/Transitional Zones:
  • All fractions: Exhibited systematic overprediction of elevation.
  • Transitional Zone Specifics: Samples from the mudflat-mangrove transition showed significant inaccuracy (underprediction in some cases), driven by the mixing of exogenous signals and high propagule mobility.
  • Barren Mudflats: Interestingly, barren mudflat samples (without transitional vegetation) showed accurate elevation estimates, suggesting that the transitional nature of the habitat (mixing of sources) is the primary driver of error, rather than just the presence of propagules.

5. Significance and Implications

• Validation of Bulk eDNA in Mangroves: The study confirms that for mangrove environments, using bulk sediment eDNA is a robust method for RSL reconstruction. The high organic content and sheltered conditions preserve extracellular DNA, stabilizing the assemblage against the noise introduced by mobile propagules.
• Caution in Transitional Zones: The findings highlight a critical limitation: eDNA-based RSL reconstruction in mudflat-mangrove transitional zones is prone to significant bias. The dominance of mobile propagules and the influx of exogenous DNA from adjacent environments distort the elevation signal.
• Methodological Guidance:
  • Researchers should exercise caution when applying bulk eDNA transfer functions to transitional or barren mudflat settings.
  • In mangrove settings, bulk eDNA offers a reliable, high-resolution alternative to morphological methods, even where test preservation is poor.
  • Future studies should consider the temporal resolution of eDNA; mudflat eDNA reflects short-term, dynamic inputs, whereas mangrove eDNA integrates longer-term environmental signals.

Conclusion: The study advances the field by clarifying that while foraminiferal eDNA is a powerful tool, its accuracy is contingent on the specific environmental context and the life-stage composition of the DNA pool. It establishes that bulk sediment eDNA is reliable in mangroves but requires careful interpretation in dynamic, transitional intertidal zones.