Challenges in eDNA detection of the invasive European green crab, Carcinus maenas

The early detection of invasive species is essential to cease the spread of the species before it can cause irreversible damage to the environment. The analysis of environmental DNA (eDNA) has emerged as a non-harmful method to detect the presence of a species before visual detection and is a promising approach to monitor invasive species. Few studies have investigated the use of eDNA for arthropods, as their exoskeleton is expected to limit the release of eDNA into the environment. We tested published primers for the invasive European green crab, Carcinus maenas, in the Gulf of Maine and found them not species-specific enough for reliable use outside of the area for which they were designed for. We then designed new primers, tested them against a broad range of local faunal species, and validated these primers in a field study. We demonstrate that eDNA analyses can be used for crustaceans with an exoskeleton and suggest that primers and probe sequences must be tested on local fauna at each location of use to ensure no positive amplification of these other species.


Introduction
Invasive species, introduced intentionally or by accident, can cause irreversible damage to the environment, threaten marine and freshwater ecosystems by outcompeting native species, thus decreasing biodiversity, and even threatening and impacting human health (Darling and Mahon 2011). 20-30% of all introduced species have caused major damages to their new environments, leading to over $120 billions of damages each year, and potential solutions to identify the most cost-effective way to repair and prevent these damages are still being investigated (Pimentel et al. 2001;Epanchin-Niell 2017;Pimentel et al. 2005). Early detection of an invasive species is crucial to preserve biodiversity and prevent environmental damages as many eradication methods can be costly and cause harm to native wildlife. Therefore, a reliable method of identifying and tracking of invasive species is necessary (Harvey et al. 2009;Gherardi et al. 2011;Jerde et al. 2011;Simberloff et al. 2012).
Monitoring marine species, particularly in fisheries management, is often accomplished through catch and release techniques (Cooke et al. 2006;Pollock and Pine 2007). However, these observations are sometimes inaccurate due to limited access to Abstract The early detection of invasive species is essential to cease the spread of the species before it can cause irreversible damage to the environment. The analysis of environmental DNA (eDNA) has emerged as a non-harmful method to detect the presence of a species before visual detection and is a promising approach to monitor invasive species. Few studies have investigated the use of eDNA for arthropods, as their exoskeleton is expected to limit the release of eDNA into the environment. We tested published primers for the invasive European green crab, Carcinus maenas, in the Gulf of Maine and found them not species-specific enough for reliable use outside of the area for which they were designed for. We then designed new primers, tested them against a broad range of local faunal species, and validated these primers in a field study. We demonstrate that eDNA analyses can be used for crustaceans with an exoskeleton and suggest that primers and probe sequences must be tested on local fauna at each location of use to ensure no positive amplification of these other species. the respective areas (e.g., marine protected areas), required taxonomic expertise due to morphological similarities between species, or limited time or funding for the respective detailed surveys (Cooke et al. 2006;Polonco Fernández et al. 2021; Thomsen and Willerslev 2015). Capture detections often rely on bottom trawling, which can cause habitat destruction and possible bycatch of unrelated species. Underwater visual censuses and photography or video surveys can be problematic due to environmental conditions (e.g., light levels) and spatial coverage of these surveys, thus suffering from biases toward particular species. Furthermore, some habitats (e.g., rocky areas with changing benthic characteristics) may be too costly to access with traditional visualization gear (Afzali et al. 2021;Danielsen et al. 2005;Danovaro et al. 2014).
The detection of environmental DNA (eDNA) has reliably been applied in terrestrial and aquatic environments to detect native and invasive species (Pilloid et al. 2013;Rusche et al. 2007;Taberlet et al. 2012). DNA is continuously released by organisms into their respective environment (Lawson-Handley 2015), and can be isolated from water, soil, or air, and then amplified by quantitative real-time polymerase chain reaction (qPCR) using specific primers and a fluorescent dye or a specific fluorescent probe. The amount of eDNA can be quantified using standard curves of a known DNA concentration. The DNA is shed by living or deceased organisms (e.g., from skin and bodily excretions), as well as extracellular DNA from cell death or destruction (Deiner and Altermatt 2014; Thomsen and Willerslev 2015;Strickler et al. 2014;Taberlet et al. 2012). Once exposed to the environment, eDNA begins to degrade due to chemical hydrolysis and microbial activity. Additional abiotic factors, including temperature, pH, UV radiation, and salinity can adjust this degradation rate, specifically by altering enzymatic activity which degrades DNA (Strickler et al. 2014). eDNA can be preserved in water from hours to weeks, or even years in ice, depending on the abiotic factors and the aggregate eDNA released (Baker et al. 2018;Balasingham et al. 2017;Dejean et al. 2011;Willerslev et al. 2014).
Species-specific detection by eDNA and qPCR has been attempted for multiple invasive species like Carcinus maenas, the European Green crab (Bott et al. 2010;Crane et al. 2021;Roux et al. 2020). This species is a prime example of an invader that has caused significant damage worldwide. Native to Europe and North Africa, it is now established throughout the coasts of North America, Australia, South Africa, and Asia (Ahyong 2005;Klassen and Locke 2007) and is predicted to even invade Antarctica (Aronson et al. 2015). C. maenas populations have been increasing along the coasts of the United States and other countries (Gharouni et al. 2017;Yamada et al. 2005), which has led to destruction of eelgrass beds and an increase in coastal erosion (Garbary et al. 2014;Infantes et al. 2016). Potentially due to C. maenas' high adaptability, it quickly becomes the dominant species of crab, outcompeting native species in the intertidal once it is established, and thus impacting the distribution and abundance of a multitude of species, specifically bivalves and other crabs (Bott et al. 2010;Jensen et al. 2002).
Using eDNA analysis can aid in the early detection and tracking of C. maenas before it has become established in a new location. Methods of trapping the species are still possible, however this relies on the species being present in the exact location of the traps and actually being caught. eDNA analyses provide an additional method of tracking the species without relying on successful capture of the species itself. Furthermore, it could be easily implemented into other tracking methods, and can be used in conjunction with eDNA detections of other species without the need for additional equipment.
Carcinus maenas' wide range and genetic variability between populations of C. maenas may require different primers for the detection of each C. maenas population (Darling et al. 2008;Darling and Mahon 2011;Jeffery et al. 2017;Roman and Palumbi 2004). Primers and probes for qPCR specific to C. maenas have previously been designed and implemented in detection methods in Australia and have proven to amplify only C. maenas DNA and not that of other local Australian species (Bott et al. 2010;Roux et al. 2020). However, it is still unknown whether these primers can be used in other locations for the detection of C. maenas due to the genetic variability and the presence of diverse local species. The recently published primers for C. maenas in the Gulf of Maine (Crane et al. 2021) have not been validated with local species and were only tested in silico.
The goal of this study was to develop and validate qPCR assays for use in detecting C. maenas in the Gulf of Maine, USA. Furthermore, we determined whether primers and probes for C. maenas in the Gulf of Maine can also be used in other, genetically different populations of this species.

Animal collection
Multiple species were collected by hand in the intertidal zone in Biddeford Pool, Maine, USA 43.442292° N, 70.341244° W, Carcinus maenas (European green crab), Hemigrapsus sanguineus (Asian shore crab), Pagurus longicarpus (Longwrist hermit crab), Balanus improvisus (Bay barnacle), Asterias forbesi (Forbes sea star), Modiolus modiolus (Northern horse mussel), Mytilus edulis (Blue mussel), Littorina littorea (Common periwinkle), Caprella mutica (skeleton shrimp), Palaemon elegans (Rockpool shrimp) and Nucella lapillus (Atlantic dog whelk). Specimen of Homarus americanus (American lobster), Cancer irroratus (Rock crab), and Cancer borealis (Jonah crab) were obtained from local lobster fishermen in Biddeford, Maine. Male C. maenas were also collected by hand and by traps in Kejimkujik Seaside National Park, Nova Scotia, Canada (43°50′32.3″N 64°50′04.1″W); Placentia Bay, Newfoundland, Canada (47°49′07.1″N 54°01′08.8″W); and Sandgerdi, Iceland (64°2′17.2284″N 22°43′16.8096″W) and transported in coolers by car (crabs from Maine and Canada) or air cargo (crabs from Iceland) to the Marine Science Center of the University of New England in Biddeford, Maine. Crabs from locations other than Maine were held in separate 300 L tanks in a flow through sea water system and fed an assortment of fish ad libitum once a week. Effluent from the system was filtered through 1 µm filters and sterilized by 2 UV filters (QL-40 Lifegard Ultraviolet Sterilizer) prior to discharge. The entire system was housed in a 4-m diameter pool as a secondary containment. The system was permitted and inspected by the Maine Department of Marine Resources (permit numbers S2013-007, S2014-007, S2015-008, S2016-008, S2017-007). Hemolymph samples were taken from decapod crustaceans by inserting a needle into the arthropodial membrane of the 4th walking leg. In molluscs, echinoderms, and the barnacle a piece of tissue was excised with bleach-sterilized scissors. DNA was extracted from the hemolymph or tissue using the Qiagen DNeasy Blood and Tissue Kit following the manufacturer's protocol.
Carcinus maenas-specific primers and probe published by Bott et al. (2010) were tested with the extracted DNA from multiple crustacean species by qPCR on a Stratagene Mx3005x thermocycler following the parameters specified in Bott et al. (2010): 95 °C 15 min, 40 cycles of 95 °C 15 s, 60ºC 1 min. Carcinus maenas-specific primers and probe published by Crane et al. (2021) were tested with the extracted DNA from several crustacean species by qPCR on a Stratagene Mx3005x thermocycler following the parameters 95 °C for 3 min, followed by 40 cycles of 95 °C for 30 s and 60 °C for 30 s.

New primer design
Sequences of the COI gene of Carcinus maenas (GenBank Accession number JQ306003.1), Hemigrapsus sanguineus (KT209545.1), Homarus americanus (KU564525.1), Cancer irroratus (MG320501.1), Cancer borealis (KY250734.1), Callinectes sapidus (MH235922.1), and Crangon crangon (KT209555.1) were aligned using the online Multalign tool (http:// multa lin. toulo use. inra. fr/ multa lin/) and regions of higher inter-specific nucleotide variability in the alignments were identified. Organisms for nucleotide alignment were chosen based on genetic relativity to C. maenas and population in the Gulf of Maine (i.e., organisms that are commonly found in the same location as C. maenas were chosen). GenBank sequences were compared to at least two other GenBank sequences to ensure no major nucleotide differences. Multiple degenerate forward and reverse primers were designed manually, along with TaqMan MGB-FAM qPCR probes. Primers were added as 1 µl of 100 nM stock solution in 40 µl GoTaq qPCR Master Mixture (Promega, Madison, WI, USA), TaqMan MGB-FAM probe was added as 1 µl of 100 µM stock solution. qPCR conditions were 15 min at 95 °C; then 40 cycles of 15 s at 95 °C and 1 min at 62 °C. The last segment was one cycle for 1 min at 95 °C, 30 s at 62 °C, and 30 s at 95 °C. The primers and probe which yielded amplification for C. maenas were then tested for specificity on a broad range of other local species, as well as validity of these primers on C. maenas from other populations (Table 1). If DNA amplification occurred before 40 cycles within qPCR, the sample was deemed to have positive amplification of DNA, if amplification curves did not reach the threshold (CT) at cycle 40 this was considered negative amplification and interpreted as no DNA present in the sample. For every qPCR, negative (no DNA added) and positive controls (DNA isolated from green crab tissue) were run along with the isolated DNA from the collected samples.
To quantify eDNA concentration, a standard curve with cycle thresholds (CT) from a dilution series (1x, 10x, 100x, 1000x) of isolated DNA from the species was used. DNA for these methods was isolated from the liver of C. maenas using the Qiagen DNeasy Blood and Tissue Kit and the concentration was determined using a NanoDrop 2000 spectrophotometer.

Validation in the field
To test whether the designed primers can be used to detect eDNA of C. maenas in the field we collected sea water samples from the dock in the harbor of Wells, Maine, USA (43.320093° N, 70.563395° W) during a one-month span between September and October 2020. A remotely operated vehicle (BlueROV2, BlueRobotics, Torrance, CA) with an attached GoPro camera was deployed to record bottom fauna and confirm that C. maenas was present at the sample site before beginning the study. Ninety minutes before the high tide, 1000 mL water samples were collected using a Niskin bottle every 0.6 m starting at the surface and moving towards the bottom at 5 m. Water samples were stored in sterile Nalgene bottles on ice and filtered in the lab within 2 h (see methods below). As a control, a cooler blank with 200 mL of deionized water was brough to the field to check for contamination. This, along with 200 mL of deionized water as a blank sample from the lab was filtered before the collected water samples to ensure there was no DNA contamination from the equipment used. The water collection with depth and negative control water samples were collected on 5 different days during a one-month period.

Filtration methods and DNA extraction
To extract eDNA from seawater a system of four 300 mL filter funnels with sterile 47 mm 0.45 cellulose nitrate filters connected to a vacuum pump (Gast DOA-P7004-AA) was constructed. This system was contained in a light proof box and sterilized by 7-Watt UV light for ten minutes before each water filtration (Ravanat et al. 2001). Sea water samples were filtered, and filters were stored in individual sterile 2 mL Eppendorf tubes at − 80 °C until DNA isolation, if isolations could not be performed immediately. DNA from filters was isolated using the Qiagen DNeasy Blood and Tissue Kit following the manufacturer instructions, with minor modifications. 180 µL of buffer ATL and 20 µL Proteinase K were added to the filter and incubated at 56 °C for 10 min. 200 µL of buffer AL was added to the samples and they were incubated again at 56 °C for 10 min. After incubation, 200 µL of ethanol was added and the tubes were centrifuged for 1 min at 8000 rpm to separate the liquid from the filter. The liquid was pipetted to a DNeasy Mini spin column, and the column was washed with 500 µL buffer AW1 and 500 µL of buffer AW2. DNA was eluted with 200 µL buffer AE. The eluted DNA was then stored at − 80 °C.
Quality assurance/quality control (QA/ QC) The following quality control steps were performed to ensure no contamination of samples: Before sampling, all collection Nalgene bottles were sterilized with 10% bleach for 10 min and rinsed with deionized water. Once sterilized, Nalgene bottles were tightly closed and placed in a sealed, bleach-sterilized container for transport to the field. In the field, Nalgene bottles were removed from storage as needed using fresh gloves. After collection of samples, Nalgene bottles were immediately placed on ice to slow down the degradation rate of collected eDNA. All samples were filtered immediately after returning to the laboratory, within 2 h of sample collection.
The filtration system set up within the laboratory was contained in a box to minimize contamination from other sources. The system was soaked with 10% bleach for 10 min and rinsed with DI water after filtering each sample. It was then further sterilized with a UV light for an additional 10 min. All filters and collection tubes used were individually packaged and sterile. After filtration, filters were either directly used for DNA isolation or stored at -80˚C until DNA isolation. Following DNA isolations, samples were stored at -80˚C until qPCR analysis. All samples were amplified in duplicates. Control samples of deionized water, as well as artificial sea water, were filtered before experimental samples and analyzed through qPCR to corroborate that the filtration system was not contaminated.

Statistics
Field test data was analyzed through Type III ANOVA and Kruskal Wallis tests to test for differences in eDNA by sample depth. All data are shown as mean ± standard deviation. All analyses were performed using R Studio v. 1.3.1093 (R Core Development Team 2020).

New primer design
Primers and probes within the COI gene published by Bott et al. (2010) tested on C. maenas, and related species including H. sanguineus, C. irroratus, and H. americanus yielded positive amplification of all tested species, rather than solely C. maenas (Fig. 1A). Positive amplification of C. maenas occurred at a CT of 25, the other crustaceans tested did have positive amplification as well (CT 33). Dissociation curves showed clean peaks, indicating that the primers amplified only one DNA product, as intended. Primers and probes within the COI gene published by Crane et al. (2021) tested on C. maenas, and related species including H. sanguineus, Pagurus sp, H. americanus, Caprella mutica, and Palaemon elegans yielded positive amplification of all tested species, rather than solely C. maenas (Fig. 1B).
New primers had to be designed for the amplification of solely C. maenas through the alignment of the COI gene of four crustaceans. A total of 5 forward primers, 4 reverse primers, and 2 probes for qPCR were designed (Supplemental Material Table 2) and tested in varying combinations. The primers and probe that yielded robust amplification of C. maenas DNA were: forward primer 5′-AAT ATT GGG AGG GCC AGA TAT AG-3′.
These sequences are not identical to their corresponding sections of the COI gene of C. maenas to account for differences in annealing temperature between the primer and probe (Fig. 5 in supplemental material).
Amplification of DNA of only C. maenas occurred early in the qPCR, at a CT of 23.05 ± 0.5 (n = 5; Fig. 2

Carcinus maenas populations
The optimized primer and probe sequences were tested on C. maenas from the Gulf of Maine,  Newfoundland, Nova Scotia, and Iceland. The COI gene was aligned between populations to note any nucleotide differences between them (Fig. 3). DNA isolated from hemolymph of crabs from each of these populations yielded a positive signal in the qPCR and a clean single peak in the respective dissociation curve. C. maenas from the Gulf of Maine yielded a CT of 23.05 ± 0.5 and Nova Scotia crabs yielded a CT of 23.34 ± 1.4, while samples of C. maenas from Newfoundland and Iceland had CT's of 36.17 ± 3.7 and 38.8 ± 1.68 (Table 1). CT values from C. maenas of Newfoundland and Iceland were significantly different from Maine and Nova Scotia (ANOVA F 3,10 = 50.47, p = 2.41E-6).

Validation in the field
The remotely operated vehicle (BlueRov2) diving at the sampling site in the harbor of Wells, ME, showed a small number of C. maenas at our testing site. This site was chosen due to the easy access and anecdotal information on a large C. maenas population presence. Within a 15-min dive covering about 100 m 2 of benthos we spotted 5 C. maenas and 2 lobsters (Homarus americanus) walking on sandy ground. Despite this relatively low local abundance, their eDNA was detected in the water column through qPCR amplification (surface to 5 m deep) of the water column. However, amplification throughout the depth of the water column was variable depending on day of sample collection (Fig. 4). The only depth which had continuous amplification was the deepest depth of 5 m. Despite this, no other clear trend of eDNA detection (measured in qPCR cycle threshold) through the water column could be detected as water conditions (temperature, salinity, currents) might have varied between sampling days in this highly dynamic harbor estuary system. Negative control samples of filtered DI water did not amplify, demonstrating no cross contamination of DNA samples due to the filtration equipment, and all positive control samples successfully amplified. Detecting eDNA is frequently used for the detection of rare and invasive species (Dejean et al. 2011;Hinlo et al. 2017;Jerde et al. 2011). For instance, eDNA analysis was used to determine the distribution of the endangered Chinook salmon (Oncorhynchus tshawtscha) (Laramie et al. 2015). This type of analysis was also used for the detection of bull trout (Salvelinus confluentus) and brook trout (Salvelinus fontinalis) to better understand their population dynamics and showed that especially when distinguishing between species of the same genus specificity of the primers is of most importance to avoid false positives (Wilcox et al. 2016). Similarly, to detecting rare species, invasive species can be easily detected by eDNA, as shown for the invasive oriental weather loach (Misgurnus anguillicaudus) in Australia (Hinlo et al. 2017), the invasive Burmese python (Python bivittatus) (Piaggio et al. 2014), the cane toad Rhinella marina (Tingley et al. 2017), the invasive Northern Pacific seastar, Asterias amurensis, and the invasive European green crab, Carcinus maenas, in Australia (Bott et al. 2010).
In the Gulf of Maine, USA, C. maenas has become a destructive invasive species, leading to environmental and biodiversity damages. Eelgrass (Zostera marina L.) destruction in southern Maine has been linked to increased bioturbation by this species. This destruction of local eelgrass is of utmost concern due to the importance of this community on the local habitat. Eelgrass is one of the most productive plant communities in the area and is a necessary habitat for fish and shellfish species, thus providing an abundant food source for local marine, including avian, species, and the economic food industry of southern Maine (Neckles 2015; Orth et al. 2006a, b). Eelgrass also provides oceanographic importance by stabilizing bottom sediment and regulating wave action, prolonging the process of coastal erosion (Orth et al. 2006a). With the increase in devastation of this plant community due to C. maenas activity has come the increase in costal erosion and loss of habitat for numerous species (Neckles 2015). C. maenas has also impacted native species due to predation. Juvenile rock crabs (Cancer Fig. 4 Amplification of C. maenas eDNA from water samples collected for a total of 5 m. Each day is plotted separately, n = 2 per depth for 10/16/2020, 10/19/2020, and 10/20/2020. Temperatures were recorded at the surface and might differ at depth. X represent datapoints that include one or more replicates with no amplification irroratus) and soft-shell clam (Mya arenaria) are highly consumed by green crabs and have decreased in population due to this predation. This not only harms the biodiversity of local habitats, but also the economy (Tan and Beal 2015;Griffen and Riley 2015). Efforts to monitor the arrival of this destructive species in new areas is thus of utmost importance to aid early eradication of C. maenas before causing irreversible damage to coastal ecosystems. eDNA analyses can be a helpful tool for this if used with the appropriate methods and species-specific primers.
Currently, there are two different sets of primers and probes published for eDNA detection of Carcinus maenas, one by Bott et al. (2010), and one set by Crane et al. (2021). Bott et al. (2010) developed a series of species-specific primers for eDNA detection of multiple species found in Australia, including C. maenas. The C. maenas-specific primers are based on the COI gene, a standard target for DNA barcoding (for review see e.g., Krishnamurthy and Francis 2012;Kress et al. 2014). These primer sets were based on extensive testing against related fauna and the use of methods easily applicable at other locations. Furthermore, these primer and probe sequences were used in a follow-up study to analyze assay performance on the west coast of Canada. The assay was able to accurately detect presence and absence of green crabs in these areas (Roux et al. 2020). Thus, based on its sensitivity, we tested these primers in another geographical region to see if this assay could be applied to the east coast of the United States as well. There is considerable variability of the COI gene sequence among the C. maenas populations worldwide (Roman and Palumbi 2004;Brian et al. 2006). We tested the primers and probe published by Bott et al. (2011) on C. maenas caught in Maine, USA, as well as on multiple other species found in the Gulf of Maine. The described C. maenas-specific primers lead to gene amplification by qPCR not only in C. maenas, but also in the Asian shore crab, Hemigrapsus sanguineus, the Rock crab, Cancer borealis, and the Jonah crab, Cancer irroratus. H. sanguineus is also an invasive crab which competes with C. maenas in the rocky intertidal (Griffen 2016). We did not test these primers on further local species as it was clear that these primers do not provide the species-specificity required in our test area. While the primers designed by Bott et al. (2010) show great species specificity in the Australian and the Pacific Coast of Canada fauna, using the same primers the Gulf of Maine may cause false positives when testing for the presence of C. maenas in the presence of other crustaceans.
The primer and probe set to detect eDNA developed by Crane et al. (2021) allowed for the detection of C. maenas eDNA in the laboratory setting. Furthermore, that study showed that the various life stages of crabs release different amounts of eDNA. Through their analysis of the effect of life stages, including oviparity, results showed that eDNA sampling during spawning events could increase detection rates, which is of utmost importance when interpreting eDNA results in the field. However, this study was not able to detect eDNA from sediment samples, despite the presence of C. maenas above the sediment. While we did not test our primers and probe on sediment samples, in water samples C. maenas eDNA was detected along the depth profile, showing that our primers and probe are suitable for detecting C. maenas in the field. We tested our primer and probe extensively against local species, while Crane et al. (2021) tested their primers in silico. As we show in Fig. 1B, in our hands using traditional qPCR the Crane et al. (2021) primer and probe set is not species specific, as it detects multiple other native and invasive crustacean species as well (H. sanguineus, Pagurus sp., H. americanus, C. mutica and P. elegans). We did not test whether these primers and probe are species specific in an actual ddPCR. However, ddPCR is typically more sensitive than qPCR (Taylor et al. 2017), which would lead to detection of even smaller amounts of eDNA from other species.
Testing for C. maenas eDNA along depth profiles showed variability of eDNA detection per day, with detection ranging from qPCR CT values of approximately 28 to 40, with 40 corresponding to no detection of the species. The most consistent eDNA amplification of C. maenas occurred closest to the benthos at 5 m depth. However, positive amplification was not consistently shown at the depths preceding this at 4.4 m and 3.8 m. While a larger number of water samples showed eDNA presence at warmer temperatures (15.6ºC vs. < 14.7ºC) we don't interpret this as an effect of temperature on eDNA recovery, as the differences are too small. The sampling location in the Wells harbor is a highly dynamic estuarine system with four rivers (Webhannet River, Pope Creek, Depot Brook, and Blacksmith Brook) draining through a marsh area into a harbor. The estuary is protected by a peninsula and drains into the ocean through a narrow channel. The average tidal range at this harbor is about 2 m (Monserrat et al. 2014). This dynamic system leads to significant water mixing, potentially either leading to an omnipresence of C. maenas eDNA, or a dilution below detection level, or any scenario in between, dependent on currents, tides, and other parameters. Whether a more pronounced stratification of the eDNA signal can be found in deeper more stratified water with less mixing remains to be shown and should be addressed in future studies. Furthermore, eDNA release has been linked to life stages and spawning events (Crane et al. 2021;de Souza et al. 2016). Our depth profile study took place in September and October, during which C. maenas larvae are still present in the estuarine system. The presence of the larvae could have increased eDNA yield, thus leading to more detectable concentrations of DNA. Therefore, our depth profile in this highly dynamic system should not be interpreted as the ultimate tool for detecting detailed spatial presence or absence of C. maenas, especially in absence of a true negative control (which in this setting is not possible to obtain). We see this part of the study more as a proof of concept, showing that detection of C. maenas with these primers is possible in the field.
Our results indicate that species-specific eDNA primers for species that are distributed world-wide, need to be assessed carefully in the appropriate context. It is thus suggested that primer and probe sequences be tested on related local species when using them for eDNA analysis in several locations. C. maenas has successfully invaded South Africa, Japan, Pacific Coast of USA, Canada, Tasmania, Argentina, and is predicted to invade Antarctica (Aronson, 2015;Carlton and Cohen, 2003;Cohen et al. 1995;Hidalgo et al. 2005;Roman 2006). With this vast range comes variability in genetics and haplotypes. For example, between Canada and New York, USA there are 6 haplotypes alone and the differences in genetic makeup are consistent with minimal gene flow between particular populations (Roman and Palumbi 2004;Williams et al. 2015). Thus, it is necessary to test the primer and probe sequences on each population of C. maenas as well to ensure species specificity of primers. Our newly designed primer and probe sequences were additionally tested on C. maenas of three other populations found in Nova Scotia, Newfoundland, and Iceland. The population from Nova Scotia showed amplification of DNA at a CT of 23.3 ± 1.4, similar to that of C. maenas in the Gulf of Maine with a CT of 23.05 ± 0.5. Green crabs from Newfoundland and Iceland, however, yielded amplifications at a CT of 36.17 ± 3.7 and 38.8 ± 1.68, respectively. This wide range of CT values is possibly due to DNA degradation attributed to age of the samples, and not the difference in genetic makeup due to a known lack of nucleotide differences within the location of base pairs of primers and probes. It should be noted that there is only one nucleotide difference found in the genome of the Icelandic population within the location of the forward primer (Fig. 3). These samples were stored in a -80ºC freezer for approximately three years prior to this test, while the samples of crabs from Nova Scotia and the Gulf of Maine were new (samples taken immediately before DNA isolations and qPCR). Comparing samples from crabs from Nova Scotia that were stored for 3 years with freshly sampled DNA from Nova Scotian crabs showed CTs of 32.27 ± 1.25 vs 23.34 ± 1.4, respectively. Consequently, we interpret the lower CT values in crabs from Iceland, Newfoundland, and Nova Scotia not as a lower performance of the new primers, but as caused by deterioration of the DNA in the freezer. We conclude that we do get positive amplification of DNA from all four populations of C. maenas tested with our new primer and probe combination. However, based on our findings the primer and probe sequences need to be tested against the local fauna in Iceland, Newfoundland, and Nova Scotia, to make sure that the primers don't detect a species that is not located and already tested in Maine.
In conclusion, we have developed and validated qPCR primers for species-specific eDNA detection and shown that they need to be tested against the local fauna to prevent false positives. We tested and verified our primers against 23 different species from 5 different phyla and feel that this captures a broad enough range to conclude species specificity. It was questioned earlier whether crustaceans leave a detectable eDNA signal in the water, due to their exoskeleton and subsequent potentially reduced eDNA release (Dougherty et al. 2016;Tréguier et al. 2014), another study showed significantly lower eDNA shedding rates in crustaceans compared to invertebrates without an exoskeleton (Andruszkiewicz et al. 2020). Our laboratory studies demonstrate that our primers and probe can detect eDNA C. maenas in small volumes of water from holding tanks in laboratory setting (Danziger et al. 2022). Furthermore, our field data show that despite the low concentration of eDNA release by crabs, green crab eDNA (potentially, but unlikely aided by the presence of larvae) can indeed be detected not only in small volume laboratory samples, but in water samples from the field at various depths, despite the highly dynamic water mixing in this estuary system. This method therefore shows that eDNA detection can be used for crustaceans with an exoskeleton and provides an additional tool to detect and monitor the spread of this highly invasive species without the need for visual detection of the species. This is especially important in invasive species with a nearly world-wide distribution in which locationspecific primers and probe combinations may be needed. Furthermore, for species with genetically distinct populations the species-specific primers need to be tested on all the respective populations. In our case, the developed primers could be used on green crabs from Maine, Nova Scotia, Newfoundland, and Iceland. However, whether the same primers would work on green crabs from other locales requires further testing.