Documenting species invasions and assessments of ecological changes depend on detection. Here, we present a simple design of a plankton light trap with specific wavelength LEDs and modifications. We used PVC pipe to create standardised small, rigid, low-cost traps that can be deployed in lentic habitats. With a cost of under $30 US each, including lights and rechargeable batteries, our traps are affordable without the need for disposable chemical lights. These small traps rely on a vacuum to retain contents upon retrieval, eliminating complicated closing mechanisms and allowing bottom entry. Our design includes submersible LED lights that can withstand pressures of at least 5 atm. We expect that the included instructions for underwater light construction and rubber weights using sand may be broadly applicable. However, we designed and field-tested our traps focusing on the detection and capture of the infective copepodid lifestage of a freshwater parasitic copepod, Salmincola californiensis. This lifestage had previously only been observed by rearing in a laboratory setting and is of concern due to continued spread outside of its native range and detrimental impacts on salmonids, especially in freshwater reservoirs. We used a 445–450 nm wavelength LED light for capturing Salmincola copepodids, but the light design can be modified to any readily available LED and heat sink to attract other target organisms. In our case, the overall affordability of the trap and components allowed for the extensive trapping needed to capture and map the occurrence of rarely-observed species and lifestages, such as the copepodids of S. californiensis. In general, increasing the number of traps that can be deployed within or across sites can aid in the spatial comparisons of plankton distributions needed in studies of ecology and species life histories. Light traps may aid in the detection of introduced zooplankton, such as S. californiensis, outside of their native range and associated plankton community changes.

lentic, parasitic copepod, pressure, reservoir, Salmincola californiensis, salmon, zooplankton

Zooplankton are integral to the ecology of aquatic habitats (Lehman 1988), including free-living holoplankton species as well as juvenile life-stages of invertebrates and vertebrates that transition to non-planktonic adult forms. Introductions of both free-living and parasitic zooplankton can have important ecological implications (Yan et al. 2002; Havel and Shurin 2004; Duggan and Pullan 2017). In many cases, zooplankton lifestages contain critical transitions such as host-finding (Heuch et al. 2007) or settling (Freckelton et al. 2017) that are often poorly understood outside of laboratory settings because of the difficulty of field collection. For species of concern, collecting infectious and larval stages from the wild may be important in understanding population dynamics and detecting novel invasions (Suárez-Morales and Mercado-Salas 2013). However, such collections can be challenging in environments that are difficult to sample by tows and when taxa have strong swimming abilities or are otherwise under-represented when using traditional netting collection methods (De Bernardi 1984; Porter et al. 2008).

Many plankton species, especially those with strong swimming abilities, may be phototactic in order to regulate their position in a water column (Martynova and Gordeeva 2010). This can reduce predation risk while maximising foraging opportunities for planktonic consumers that exhibit diel vertical migration, with light acting as an external cue and providing directional orientation (Forward 1988). For organisms exhibiting positive phototaxis, light traps are a logical method of collection (McLeod and Costello 2017). Light trap designs vary widely, but include a light source and an opening that narrows towards the interior of the trap (McLeod and Costello 2017). Often, traps have horizontal entrances (Meekan et al. 2001; Fisher and Bellwood 2002; Hernandez and Shaw 2003) with collection into a mesh cup or basket at the bottom of the trap. Traps are often mesh-sided; however, mesh nets may experience problems with clogging during phytoplankton blooms (Jones 1971) and by non-target animals attempting to enter the trap (McLeod and Costello 2017). Traditional light traps are often expensive and require construction by a machine shop (Floyd et al. 1984). Recent efforts towards inexpensive light traps include modifications of existing water samplers or nets and the use of disposable light sticks (Kehayias 2006; Kehayias et al. 2008), as well as larger designs that incorporate inexpensive materials with a modular light box and fluorescent lighting (Watson et al. 2002). Light traps can be quite large, up to one metre in height (McLeod and Costello 2017), though smaller traps may be as or more effective (Meekan et al. 2001). Designs for small, simple, inexpensive traps and light source customisation to collect phototactic plankton are lacking, especially considering the recent broad availability of selective wavelength LED components.

Light traps may capture species that are absent in other sampling efforts (Jones 1971). For example, Salmincola sp. are freshwater Lernaeopodid copepods, known as natural parasites of Brook Trout and other salmonids (Fasten 1912). However, these copepods have been recently recognised as a problem in regions where they have likely been introduced along with host fishes (Kamerath et al. 2009) and in altered systems including reservoirs (Hargis et al. 2014; Monzyk et al. 2015; Lepak et al. 2021). Salmincola californiensis is of particular concern because of high infection prevalence and parasite burden on threatened Chinook Salmon (Oncorhynchus tshawytscha) and associated impairment of the fish in Oregon, USA (Herron et al. 2018; Neal et al. 2021). Yet, copepodids (the infectious, free-swimming lifestage) of this species were not detected after four years of extensive tow netting in multiple upper reservoirs of the Willamette Basin (Murphy et al. 2019a) where Salmincola sp. adults were common on salmon (Monzyk et al. 2015). Unlike sea lice (family Caligidae), Salmincola sp. (commonly known as gill maggots) permanently anchor onto their host during development. Their development includes a non-swimming nauplius, a motile copepodid and attached chalimus and adult stages and they are restricted to freshwaters throughout all lifestages (Kabata and Cousens 1973; Murphy et al. 2020). Although adult female Salmincola sp. are easily observed on parasitised fish, the infectious copepodid stage has not been captured in the wild. Rather, this stage was described in the laboratory through hatching egg sacs from adult females (Kabata and Cousens 1973). Capturing copepodids could provide early detection where sensitive fish are rarely handled, offer insight into the conditions that are exacerbating infection rates and may eventually lead to methods for management and control. There remain challenges to capturing and quantifying species exhibiting primarily vertical (not horizontal) movements, such as copepodids of the parasitic genus Salmincola (Poulin et al. 1990) and to producing a sufficient number of traps to detect their presence and distribution (Doherty 1987), assuming they may be patchy in space and/or time (Herron et al. 2018).

To figure out how we might capture S. californiensis copepodids, we conducted extensive laboratory observations and revised the literature about sampling for infectious stages in other groups of parasitic copepods. We noted, as had previous studies for related species, that S. californiensis copepodids were negatively buoyant and frequently rested on the bottom of laboratory tanks outside of active bouts of swimming (Poulin et al. 1990). Based on the findings of Penston et al. (2004, 2008) for sea lice, we initially focused on mid-water tows and benthic sleds for copepodid capture. However, such sampling failed to yield copepodids from reservoirs where high levels of infection were documented (Monzyk et al. 2015). After noting positive phototaxis exhibited in the lab and the design by Novales Flamarique et al. (2009), we focused on the design and construction of customised light traps as we did not find replicable published designs that would allow capture from below (but see Tranter et al. 1981). Generally, leaving openings at the bottoms of traps would allow water (and captured material) to flow out of traps during retrieval unless there was a closing mechanism. Noting the limited information available on the customisation of light sources, on designs for traps with bottom openings (important for captures during the process of vertical migration) and on designs minimising mesh (important for species that are easily damaged or difficult to remove and to avoid clogging during phytoplankton blooms), we designed a novel small, robust trap with customisable lighting that can be run in vertical transects. A non-mesh trap was important because of the presence of structures and debris (e.g. stumps) that could damage mesh traps, as well as phytoplankton blooms that cause clogging of mesh cod ends. Designs using acrylic that fit these criteria were cost prohibitive for the extensive trapping we anticipated for this difficult-to-capture infective lifestage of the parasitic copepod. Finally, the commercialisation of designs for aquaculture has limited the information available to guide light trap construction. Here, we describe how we built, set and collected samples from our low-cost, robust light traps that can be deployed in series from the water surface to depths of at least 60 m and that can be outfitted with selective wavelength LEDs, based on target species. We also provide field test data from three reservoirs in the upper Willamette River Basin in Oregon, USA, to demonstrate the efficacy of these traps.

We captured S. californiensis copepodids from the wild (Cougar Reservoir, OR, USA, November and December 2018) using our novel light traps. These first captures of Salmincola sp. relied on prototype traps constructed using clear, glass funnels set into the bases of 5-gallon (19 l) buckets with sealed lids. Although successful, these traps were large and easily damaged. During associated field trials, many of the underwater lights we tried were ineffective, not modifiable and cost prohibitive. Our study sites are used for human recreational purposes and valuable equipment is more likely to be disturbed or removed. Thus, we explored PVC-based traps to focus on improving portability and durability while maintaining a cost that would be reasonable to deploy at a large scale and that could accommodate loss. The final trap design is based on an 18 cm length of 4-inch (10.16 cm) diameter Schedule 40 PVC pipe that is easy to source in the plumbing section of local hardware stores (Suppl. material 1: Table S1). We were unable to reproduce the closing mechanism in the trap described by Novales Flamarique et al. (2009), even after contact with the authors, so we focused on a simple design that would create a vacuum, thus retaining contents upon retrieval. Lights were placed in water-tight jam jars and either secured to the top lid or secured at the top of the trap by having their latch seated in a cut-out at the top of the pipe (Fig. 1). The light beam was directed downwards through a glass funnel placed in the bottom lid. We designed the traps to use rechargeable AA batteries because they are easy to source and minimise disposable waste after intense use.

Figure 1. 

Trap design A trap schema and component placement. Light is placed under the lid facing down towards the open inward-facing funnel at the bottom (entrance 8 mm at narrowest point, 75 mm diameter at the widest) B views of traps including standard (left), deep (centre) and funnel, light colour and deployment loops (right). The rubber weight is used on both trap configurations to ensure vertical orientation, but is pictured on the deep trap here, for reference.

We successfully captured S. californiensis copepodids using our light traps, but did not detect them in concurrent tow or Van Dorn traps samples (Table 1). In total, we collected 532 light trap samples across all reservoirs and sampling dates and 53 of the traps captured our target organisms. For the traps that captured S. californiensis copepodids, the number of copepodids captured ranged from 1 to 169 with a median of 2. Non-target organisms were also captured in our light traps (Suppl. material 1: Table S2). Most of them were cladocerans (primarily Daphnia spp.) and calanoid and cyclopoid copepods. Other organisms caught in lower numbers were various aquatic insects, aquatic mites, ostracods, amphipods and larval fish. Captures of all organisms in light traps were highly variable depending on reservoir, collection date and trap depth. Interestingly, Salmincola copepodids were captured at shallower and warmer conditions than expected, based on documented copepodid survival (Murphy et al. 2020; Suppl. material 1: Table S2).

Here, we describe light traps that successfully attract and capture Salmincola copepodids from lentic systems with the potential to be used to examine patterns of both spatial and vertical distribution of this species in our study reservoirs. In order to sample for this ‘needle in a haystack’, we present a robust low-cost trap design that can be widely deployed, including at depths with pressures of 5.9 bar and in environments that could damage mesh-based traps (e.g. reservoirs with large stumps and woody debris). The simple design eliminates the need for complex and expensive closing mechanisms and is constructed from easily-obtained materials. By establishing methods to capture this poorly-understood copepodid life stage, we will be able to gain valuable ecological knowledge about its density and distribution over time and space. This information should help us to understand the current problem of high infection prevalence and intensity and may allow us develop remediation strategies. In addition, the use of our light traps provide monitoring data to prompt ecological discoveries (e.g. use of optimised traps as potential method of control and management of copepodids).

We focused on 48-hour deployments because it increases trap captures and allows for us to set strings at multiple sites during our sampling window. Captures are thus possible throughout the day and night periods. Even so, we would expect most captures to happen at night, when the lights provide a greater difference in illumination from background natural levels. We did not focus on new moon periods, though moonlight could modify capture efficiency (Hickford and Schiel 1999). Light intensity is also likely important and we use bright lights because other species of parasitic copepod are more attracted with increasing intensity (Novales Flamarique et al. 2009).

Inexpensive traps can be important to achieve research goals with limited funding. We expect these traps to be especially useful for rare or poorly-understood species where large-scale deployment may be necessary. They are also ideal for areas with boat traffic where traps may be stolen or damaged. The option to deploy traps at depth may provide greater captures, as other studies have found increased catches near benthic areas (Tranter et al. 1981; Tor et al. 2009). Their small size can make plankton easy to damage by netting and pumping (Beers et al. 1967); damage is less likely with light trapping methods, especially using rigid traps, because taxa are not forced against a net. Inexpensive lights that function below 30 m are likely to be useful in an even broader context, as additional sampling and surveying methods rely on light for attraction or visibility. While here we use 445–450 nm wavelength LEDs, a wide variety of options are available. Spectral sensitivity for S. californiensis at 445–450 nm would not be unexpected given that other freshwater taxa have shown similar maxima (Forward 1988). Longer wavelengths (550 nm) have been documented as the maximum sensitivity for marine parasitic copepodids (Bron 1993) and peak sensitivity to 475–525 nm wavelengths is common for taxa in estuarine, coastal and open ocean habitats, though some taxa may have peak sensitivity as low as 360 nm (Forward 1988).

Future studies could explore the efficiency of these traps, especially for other taxa and the changes in capture composition with alternative LEDs and/or differing LED intensities. Blue and green wavelengths are likely ideal for a broad range of taxa and transmit well in water (McLeod and Costello 2017). Using a microcontroller, while not pursued in our final design, would allow for lights in series, patterns in illumination and the fine tuning of light intensity. Shorter duration deployments could explore day versus night efficiencies and may allow for the use of smaller batteries and possible further downscaling of the design presented here.

Outside of light trapping, we expect that the instructions (Supplement) for underwater light construction and rubber weights may be broadly applicable. Our description of how to assemble LED light sources for use underwater allows for researchers to customise wavelengths and target desired taxa. Low-cost selective wavelength LEDs may also be useful for other underwater applications (e.g. video). Weights that use discarded materials (bike tubes) and sand can aid in balancing equipment underwater where cost or vandalism are of concern. In our case, the overall affordability of the trap and components allows for the extensive trapping needed to capture and map the occurrence of rarely-observed species and lifestages, such as the copepodids of S. californiensis. In general, increasing the number of traps that can be deployed within or across sites can aid in the spatial comparisons of plankton distributions needed in studies of ecology and life history. Effective trapping of zooplankton, parasitic and free-living, may be critical to early detection of invasions or novel ecological dynamics.

Thank you to Chee Sing Lee for assistance in LED electronic assembly and design. The authors acknowledge US Army Corps of Engineers, JPL 19-03-SYS, the USGS OR Cooperative Fish and Wildlife Research Unit and Oregon State University for funding and technical support during this research. We especially thank Emilee Mowlds and Erik Swanson for trap construction and field deployment as well as Todd Pierce for trap deployment, recovery, boating and logistics. Laboratory testing would not have been possible without Ryan Koch and the Smith Farm facilities. James Peterson, Carl Shreck, Crystal Herron, Justin Saunders and Mike Kent provided discussion leading to successful trap design. David Noakes and OSU Surplus provided materials and additional inspiration. The views expressed in this publication are solely those of the authors and do not reflect official policy of the U.S. Government nor endorse any products or commercial services.