Seasonal patterns of environmental DNA detection for freshwater unionid mussels

This study demonstrates that while environmental DNA (eDNA) is an effective tool for monitoring freshwater unionid mussels and generally aligns with visual survey results, its detection efficiency is significantly influenced by seasonal factors such as reproductive behaviors, burial patterns, and hydrological conditions, necessitating that sampling timing be adjusted based on local environmental variables.

The Gist

Imagine you are trying to find a specific type of shy, buried treasure (freshwater mussels) hidden deep in the mud of a river. Traditionally, you'd have to send a team of divers down to dig through the mud, looking for shells. It's hard work, expensive, and you might miss the treasure if it's buried too deep or if the water is too murky.

This paper is about testing a new, high-tech "magic nose" called eDNA (environmental DNA) to see if it can find these mussels just as well, but without the heavy digging.

Here is the story of their experiment, broken down into simple concepts:

1. The "Scent Trail" Analogy

Think of mussels like dogs. Even when they are sleeping or hiding in the mud, they leave behind a "scent trail" in the water. This trail is made of tiny bits of their DNA shed through their waste, mucus, or skin cells.

• The Old Way: You have to physically dig up the mud to find the dog.
• The New Way (eDNA): You just take a cup of water, smell the "scent," and a computer tells you, "Yes, a dog was here!"

2. The Experiment: Two River Neighborhoods

The researchers went to two different river neighborhoods in Ohio:

• Killbuck Creek: A bit wilder, with faster water sometimes.
• Walhonding River: A calmer, more stable neighborhood.

They wanted to see if the "magic nose" (eDNA) worked all year long, from early spring (April) to late autumn (October). They compared the eDNA results against the old-school "diver" method (visual surveys) to see who was right.

3. The Weather Factor: The "Dilution Effect"

One of the biggest discoveries was about the river's mood (flow and temperature).

• The High Water Flood: Imagine a giant hose spraying water into a cup of coffee. The coffee gets diluted and you can barely taste it.
  • In the spring, when the river was swollen with rain and snowmelt (high flow), the "scent trail" of the mussels got washed away and diluted. The eDNA test struggled to find them, especially the shy ones hiding deep in the mud.
  • Lesson: Don't try to use the "magic nose" when the river is raging; wait for the water to calm down.
• The Buried Treasure: In early spring, some mussels act like hibernating bears. They bury themselves deep in the mud to stay warm. Because they are so deep, they don't leave much "scent" in the water above them. The eDNA test missed them during these times.

4. The "Shy vs. Loud" Mussel

The study found that some mussels are easier to detect than others, depending on their personality and shell:

• The Loud Shells: Mussels with bumpy, sculptured shells (like a textured rock) tend to sit on top of the mud. They are "loud" and leave a strong scent trail. The eDNA found them easily.
• The Smooth Shells: Mussels with smooth shells often like to hide deep in the mud (endobenthic). They are "quiet." The eDNA had a harder time finding them, especially if they were rare.
• The Rare Gems: The researchers were looking for very rare, endangered mussels (like the Purple Cat's Paw). Even though there were only a few of them, the eDNA found them! However, it took a lot of water samples to be sure. It's like finding a needle in a haystack; you need to check a lot of hay to be 100% sure the needle is there.

5. Timing is Everything

The researchers found that the "magic nose" works best when the water is calm and the survey happens close in time to the "diver" survey.

• If you take a water sample and then send divers down the next day, the results match perfectly.
• If you wait a month, the river changes, the mussels move, and the results get fuzzy.

The Bottom Line

Does the "magic nose" work?
Yes! It is a powerful tool that can find mussels, including rare and endangered ones, without needing to dig up the riverbed.

But there are rules:

1. Wait for calm waters: Don't test when the river is flooding; the scent gets washed away.
2. Wait for spring: Don't test too early in the year when mussels are still hibernating deep in the mud.
3. Take more samples: If you are looking for rare, shy mussels, you need to take more water samples to be sure you didn't miss them.

In short: eDNA is like a high-tech metal detector for the river. It's faster and less invasive than digging, but you still need to know when and where to use it to get the best results. It's a great new tool for conservation, but it works best when used alongside the old-school methods, not as a total replacement.

Technical Summary

1. Problem Statement

Freshwater unionid mussels are a highly imperiled group, with nearly one-third of U.S. species federally listed as endangered or threatened. Traditional monitoring relies on visual-tactile surveys, which are labor-intensive, require expert taxonomists, and suffer from imperfect detection due to:

• Behavioral factors: Vertical migration (endobenthic burial), reproductive cycles, and shell morphology.
• Environmental factors: Substrate type, water clarity, flow discharge, and observer fatigue.
• Seasonal limitations: Current protocols in Ohio restrict surveys to May–October based on assumptions that mussels bury deeply and reduce metabolic activity in colder months.

While Environmental DNA (eDNA) metabarcoding offers a non-invasive alternative, its implementation in regulatory permitting is limited by a lack of understanding regarding how seasonal dynamics, hydrological conditions, and species-specific reproductive behaviors (tachytictic vs. bradytictic) influence detection probabilities and abundance estimates compared to traditional visual surveys.

2. Methodology

Study Sites and Design:

• Locations: Two diverse mussel beds in central Ohio: Killbuck Creek and the Walhonding River.
• Target Species: Included federally endangered species such as Epioblasma obliquata (Purple Cat's Paw), E. triquetra (Snuffbox), and Plethobasus cyphyus (Sheepnose).
• Sampling Period: April 25 to October 22, 2024 (10 events at Killbuck Creek; 9 events at Walhonding River).
• Comparison: At Walhonding River, eDNA results were compared against a quantitative visual-tactile quadrat survey conducted on September 5–6, 2024.

Sample Collection:

• Water Sampling: Three replicate water samples (1,000 mL each) were collected per site per event using peristaltic pumps and GF/C filters. Sampling occurred across the river width, moving upstream to avoid contamination.
• Controls: Field blanks (distilled water) and laboratory negative controls were processed alongside samples.

Laboratory and Bioinformatics:

• Extraction: DNA was extracted using a modified Qiagen DNeasy protocol.
• Amplification: A ~175 bp fragment of the mitochondrial 16S gene was amplified (validated for unionids).
• Sequencing: Illumina MiSeq (2x300 bp) metabarcoding.
• Processing:
  • Sequences were processed via QIIME 2 (DADA2 for denoising).
  • MOTU Clustering: 97.5% similarity threshold.
  • Mitotype Filtering: Only female mitotypes were retained to avoid confusion from Doubly Uniparental Inheritance (DUI) in males.
  • Error Correction: Mistag/index-hopping rates were estimated using negative controls and removed from the dataset.

Statistical Analysis:

• Detection Index: Categorized repeatability as Low (≤33%), Moderate (33–67%), or High (>67%) based on replicate detection.
• Occupancy Modeling: Bayesian integrated occupancy models (using spOccupancy in R) estimated detection probability ($p$) and occurrence, testing covariates like Julian day (seasonality) and flow exceedance (discharge).
• Comparisons: Linear regression and Bray-Curtis dissimilarity were used to compare eDNA abundance/communities against visual survey data.

3. Key Contributions

1. Seasonal Validation: The study validates that eDNA can effectively detect freshwater mussel assemblages throughout the currently prescribed regulatory window (May–October) and extends slightly into early spring (April) and late autumn (October).
2. Reproductive Strategy Analysis: It demonstrates that both short-term brooders (tachytictic) and long-term brooders (bradytictic) are consistently detectable via eDNA, despite their differing spawning and glochidia release schedules.
3. Hydrological Influence: It quantifies the negative impact of high discharge on detection, showing that high flow events dilute eDNA signals and reduce species richness detection, particularly in early spring.
4. Methodological Benchmarking: It provides a direct, quantitative comparison between eDNA and visual tactile surveys, establishing that eDNA often detects higher species richness and correlates well with visual abundance when temporal and hydrological conditions are controlled.

4. Key Results

• Detection Success: Mussel DNA was consistently detected across all seasons. Walhonding River showed higher detection repeatability (74% high repeatability) compared to Killbuck Creek (56%).
• Species Richness:
  • Walhonding River: Detected 29 species (26 identified to species level).
  • Killbuck Creek: Detected 26 species.
  • Seasonal Trend: Species richness was lowest in early spring (April/May) coinciding with high flow exceedance (>60%) and low temperatures, suggesting reduced detection efficiency due to burial behavior and DNA dilution.
• Rare Species Detection:
  • Federally listed species (E. obliquata, E. triquetra, etc.) were detected.
  • Epioblasma obliquata (Purple Cat's Paw) was visually observed (16 individuals) but detected by eDNA in only 4 of 9 events, likely due to its endobenthic (buried) behavior.
  • Obliquaria reflexa (single visual individual) was never detected by eDNA, suggesting a minimum abundance threshold for detection with the current sampling effort (3 replicates).
• Abundance Correlation:
  • There was a significant positive correlation ($R^2 = 0.54, p < 0.001$) between mean eDNA sequence reads and visual mussel counts.
  • Congruence: The correlation between eDNA and visual surveys was strongest when surveys were conducted within ±20 days of each other and during low discharge periods.
• Covariate Effects:
  • Discharge: High flow negatively affected detection probability for 15 species in Killbuck Creek.
  • Shell Morphology: Species with sculptured shells had higher detection probabilities than smooth-shelled species, likely due to a higher likelihood of being at the sediment surface.
  • Brooding Strategy: No clear influence of brooding strategy (tachytictic vs. bradytictic) on detection probability was found, though tachytictic species were generally more abundant.

5. Significance and Implications

• Regulatory Utility: The study supports the use of eDNA as a viable tool for regulatory surveys within the standard May–October window, potentially offering a less intrusive and more efficient alternative to visual searches.
• Survey Optimization:
  • Timing: Surveys should avoid early spring high-flow events to maximize detection of rare and buried species.
  • Effort: For rare species (e.g., Epioblasma), increasing the number of water replicates beyond three is recommended to improve detection probability.
  • Interpretation: High repeatability in eDNA detection strongly correlates with visual presence. Low repeatability may indicate regional presence but local absence, or simply low abundance/burial depth.
• Limitations: The study highlights that endobenthic species (like Epioblasma) and extremely rare individuals (singletons) remain challenging for eDNA, requiring careful interpretation of "false negatives" in the context of flow and burial behavior.

In conclusion, this research provides a critical framework for integrating eDNA into freshwater mussel management, emphasizing that while eDNA is a powerful tool, its efficacy is modulated by hydrological conditions, species-specific behaviors, and survey timing.