© F. Baudry and T. Baudry, Published by EDP Sciences 2025

1 Background

Environmental DNA (eDNA) has emerged as a transformative methodology in biodiversity monitoring of aquatic ecosystems, revolutionizing ecological studies by allowing the detection of species through genetic material shed into the environment (Ficetola et al., 2008; Thomsen and Willerslev, 2015). Its rapid expansion is mainly driven by the improvement of analysis methods (i.e. development of specialized DNA kits or ready-to-use Master Mixes) and democratization of sequencing technologies (Seymour, 2019; Baudry et al., 2024). Since its democratization in 2008, the published literature contains thousands of studies dealing with eDNA and even more when accounting the grey literature (Belle et al., 2019; Baudry et al., 2024). This trend led to a wide variety of protocols, from field sampling to in-lab eDNA handling and analysis, which hampers data comparability across studies (Belle et al., 2019; Baudry et al., 2024) and could have limited the integration of eDNA method into regulatory frameworks (in Europe especially). Efforts to standardize eDNA methods have been initiated, such as the DNAqua-Net project in Europe, which aims to harmonize DNAbased monitoring approaches (Vasselon et al., 2025). However, the availability of standardized, cost-effective, and user-friendly sampling equipment remains limited. Commercial solutions like the Smith-Root eDNA Sampler offer advanced features, including programmable sampling parameters and GPS integration, but their high cost (>10 k €) can be prohibitive for widespread use (Thomas et al., 2018). Alternative approaches, such as passive sampling using glass fiber filters, have shown promise in capturing eDNA efficiently, yet these methods also require further validation and standardization (Kirtane et al., 2020; Chen et al., 2022).

Here, we proposed “eDIY-sampler”, a compact, low-cost and easy-to-assemble filtration system, simplifying and reducing the time and effort dedicated to on-field water sampling. The following sections gives more details on how the device is composed and how it can be improved or modified, but as stated in this article, we addressed four key gaps at a price <500€:

• It is compact (H 30 cm × W 22 cm × D 16 cm) and light (∼5 kg, including 2,9 kg for the backpack without battery and 2 kg for the full-equipped pole), making it easy to transport (by car or plane, within luggage) and very convenient to carry on field, thanks to its backpack style.

• Its low energy-consuming electrical pump allows easily to process 15 sampling sites per day, for three consecutive days without re-charging it.

• The device is highly compatible: as described here, it is using Smith-Root filter holder, allowing the use of all filters 47 mm in diameter (with a pore size >0.8 μm) of every composition (Glass Fiber, Nitrocellulose, etc.) and tubing system can be changed (as well as the pump) to fit with every type of filter holder (i.e. Sterivex cartridges or Waterra).

• eDIY-sampler, as it is named, is easy-to-assemble, made with basic components which can be easily repaired, or even improved (see further our recommendations) (see diagram figure in Supplementary material).

2 Description and testings

The eDIY-sampler device is basically composed of two main components, the backpack (Fig. 1) and the telescopic pole (Fig. 2).

First, this backpack is made from an electrical case, certified IP66 (meaning waterproof, for the rain and water projections), fixed on a regular backpack (Fig. 1). This certification certainly justifies the price (217.6 €, almost half of the total price – Tab. 1) and some savings can be done here by changing it with another cheaper alternative. The battery used here is waterproof (Yuasa NP4-12, long 90 mm × 70 mm large × 106 mm high) providing 12 V and 4AH (∼3 days sampling easily) (Tab. 1 and Fig. 1) but the backpack can handle bigger one, up to 151 mm long and 98 mm large (such as Yuasa NP12-12), for longer time use. This kind of batteries used here are waterresistant and suitable for continuous operation (typically up to several tens of minutes). Users should always consult the manufacturer’s safety guidelines prior to operation and avoid prolonged use beyond the recommended limits to prevent overheating or damage. This battery is controlled by a ON/OFF button (Fig. 1C) to supply the diaphragm pump (Seaflo 12 V 1.5 GPM and 35 PSI) (Tab. 1 and Fig. 1). We recommended to use the same characteristics because our trials showed that less powerful pumps were not able to take the water through the filter and inversely, the use of more powerful ones resulted in a tearing of the filters. These diaphragm pumps are also reputed robust and particularly tolerant to suspended particles, making them suitable for remote fieldwork. The backpack (and basically the pump) is connected to the pole with 8 mm tubing (but the pump can connect to 3 mm tubing), with a quick disconnect system, making the transport and manipulation of the device independent and therefore easier (Fig. 1).

The pole is telescopic and so can be stored or transported easily (<60 cm) and is composed of: an aluminium plate carrying the distribution system of the water, made with plumbing pieces (Fig. 2), flow indicators (Fig. 1E) showing if water is still running (through 8 mm PVC transparent tubes) and allowing to turn off this pipe, individually, thanks to plumber tap (Fig. 1D) if a filter is clogged. These filters are loaded on the extremity of the pole in Smith-Root filter holders (Tab. 1, Fig. 1F), in triplicates, disposed on a thermoformed (and holed-drilled) Plexiglas plate (Fig. 1G). We chose these holders for their convenience, because they are sufficiently resistant to be reused after disinfection, using bleach (20%) or soap and overnight-UV treatment, and they handle every filter with 47 mm diameter size, with a pore size >0.8 μm. For instance, pore size of 0.45 μm, we used before (with 1 L-filtration unit (Nalgene^TM), is too small and the pump is either too weak to draw water through the filter or too strong causing the filters to tear. Tubing used here is 8 mm, which seems to be quite universal, and any other filter holders can be used instead of Smith-Root ones and tubing can also be changed to fit. The sampling pole is highly practical, as it allows water collection from the bank in small waterbodies when needed. It can also be replaced with a longer pole to extend reach, and the entire setup can be deployed from a boat to sample larger waterbodies.

In order to validate the concept, we compared our device with a Nalgene 1 L filter unit (see for instance Baudry et al., 2021), we used for regular biomonitoring, by using 1.2 μm pore size filters in both case (47 mm of diameter, Nitrocellulose Sartorius ledR). We tested how the device can impact:

• The volume of water filtered (Fig. 2A) until the filter clogged, by sampling water from same rivers (5 stations, in Martinique) at the same time.

• The eDNA concentration (Fig. 2B) isolated from filters from 16 stations in France rivers (32 filters in total when combining the two methods).

• The results on single-species detection, by qPCR for Cherax quadricarinatus (in Martinique) (see protocol in Baudry et al., 2021) (Fig. 2C).

• Then by metabarcoding (reads count) on fish and decapods (from Martinique) detection (see protocol in Baudry et al., 2025) (Fig. 2D).

For all these conditions, no difference was found between both methodologies used, what is not surprising, as the yield for all these steps relies only on filters (and their composition), whatever the filtration system used. As said above, a more powerful pump will just lead to tearing the filters. But the main reason why this device deserves to be used is its efficiency: as it filters three samples in parallel, it allows processing water more quickly. As a result, in COUL station in Martinique (see Baudry et al., 2025), we were able to process 15 L of water (through three filters – 47 mm diameter Sartorius Nitrocellulose with 1.2 μm pore size, simultaneously) until filters clogged, in ∼11 min, while 15 min were needed for only one filter with Nalgene filtration unit, with handoperated vacuum pump. Further trials showed that 54 sec were needed to process 2 L when using eDIY-sampler in clear headwaters in Martinique (COUL station in Baudry et al., 2025) and up to 2.3 mins in mangroves (containing more suspended matters), for a total of 3.5 L of water filtered. These are just indications, measured with a bucket placed after the water exit of the pump, and better precision can be reached with some improvement (see after in the following section). That said, for instance in COUL station (clear headwater in Martinique), the sampling from device preparation to final cleaning lasted for 25 min, while ∼1.25 h were needed with Nalgene filtration unit (see ready field-use protocol in Supplementary material).

3 Conclusion and perspectives

The eDIY-sampler system presented in this study offers an efficient, cost-effective, and adaptable solution for aquatic eDNA sampling. Its compact design and ease of assembly enable high-throughput field sampling, even in remote environments. The ability to process multiple samples simultaneously significantly reduces sampling time without compromising the quality or yield of eDNA. Noted that in turbid waters (i.e., mangroves or downstream areas), the filtration speed (mainly dependent of the filter pore size) may decrease substantially, as suspended particles can rapidly clog the filter surface. The device is validated and was already used for two different field campaigns in Martinique (without replacing pump), for a total of >70 stations sampled (publication in preparation) and many punctual field samplings in France. This DIY-based solution contributes to fill an important gap in the standardization and democratization of eDNA protocols.

Future improvements may include the integration of additional tools such as a digital flowmeter and a microcontroller-based board (e.g., Arduino or Raspberry Pi) that could trigger automatic GPS localization and in-situ measurements of physicochemical parameters with probes (temperature, oxygen, conductivity, etc.), with data stored directly onto a USB drive. Moreover, the system could be modified for specific applications such as single-cartridge sampling in marine environments or for deep water sampling. These adjustments would further increase the versatility and robustness of eDIY-sampler across diverse aquatic research contexts.

Fig. 1 Illustration of the eDIY-sampler filtration system used here on field, composed by a 1) back-pack (to carry the 12 V diaphragm pump) (A) and the battery (B), activated with a ON/OFF button (C)) and 2) a telescopic pole with three independent tubing channels, activated by a tap (D) and allowing the water to flow (visualized by indicators (E)) through the filters placed in the Smith-Root holders at the extremity (F) on a thermoformed (and holed-drilled) Plexiglas plate (G).

Fig. 2 Results of the trials led to investigate the effect of filtration system (eDIYsampler vs. Nalgene filtration unit) on A) the volume filtered, B) the eDNA concentration when the same volume is filtered, C) the Ct-value found by qPCR when it comes to target C. quadricarinatus in Martinique rivers and D) the reads count recovered by metabarcoding, for fish and decapods detection.

Acknowledgments

We warmly thank the Office de l’Eau de Martinique (ODE), the Direction de l’Environnement, de l’Aménagement et du Logement de Martinique (DEAL) and the Office Français de la Biodiversité (OFB) for supporting the fieldwork through TB’s post-doctoral fellowship (InCrust project). Testings were carried out at the University of Poitiers, whose lab facilities were greatly appreciated. Finally, we would like to sincerely thank TB’s grandmother for her help (tailoring) in adapting the backpack system for field sampling.

Author contribution statement

FB: Conceptualization, Realization, Validation, Writing – review & editing. TB: Conceptualization, Validation, Formal analysis, Writing – original draft, Writing – review & editing.

Supplementary Material

Field-use protocole. Access here

References

All Tables

All Figures

Fig. 1 Illustration of the eDIY-sampler filtration system used here on field, composed by a 1) back-pack (to carry the 12 V diaphragm pump) (A) and the battery (B), activated with a ON/OFF button (C)) and 2) a telescopic pole with three independent tubing channels, activated by a tap (D) and allowing the water to flow (visualized by indicators (E)) through the filters placed in the Smith-Root holders at the extremity (F) on a thermoformed (and holed-drilled) Plexiglas plate (G).

In the text

	Fig. 2 Results of the trials led to investigate the effect of filtration system (eDIYsampler vs. Nalgene filtration unit) on A) the volume filtered, B) the eDNA concentration when the same volume is filtered, C) the Ct-value found by qPCR when it comes to target C. quadricarinatus in Martinique rivers and D) the reads count recovered by metabarcoding, for fish and decapods detection.
In the text