Distribution and food chain assimilation of New Zealand mud snails in Spirit Lake, Washington, USA

Mount St. Helens National Volcanic Monument was designated by the U.S. Congress in 1982 to conserve the landscape for natural regeneration, scientific research, education, and cultural resource preservation. However, this designation has not eliminated threats from the introduction of non-native species. The non-native New Zealand mud snail (NZMS), Potamopyrgus antipodarum, was first observed in 2016 along the SW shore of Spirit Lake at the foot of Mount St. Helens, despite the lake’s closure to public recreation and isolation from other known sites harboring NZMS. Our study mapped native and non-native snails on aquatic macrophytes in Spirit Lake, analyzed NZMS eDNA in Spirit Lake and surrounding waters, measured stable isotopes in snails and their food sources, and analyzed rainbow trout (Oncorhynchus mykiss) gut contents from a twenty-year survey to examine the patterns of spatial distribution, habitat occurrence, and resource use. Our results show that NZMS colonies were likely first established along the SW shore of Spirit Lake in 2015, and presently remain largely confined to the vegetated littoral zone along this same shoreline. The native snail species Gyraulus deflectus and NZMS co-occur on multiple macrophyte species, and δ¹⁵N and δ¹³C isotope data reveal they are consuming the same food sources, but no evidence was seen for competitive exclusion. The abundance and frequency of NZMS found in rainbow trout gut contents have increased since 2015 with a significant portion undigested. In addition, stable isotope analysis shows a negligible trophic tie between snails (both NZMS and G. deflectus) and rainbow trout, which may signal longer-term impacts on fish populations. Characterizing this invasion spatially and temporally elucidates the factors facilitating and hindering the spread of NZMS in a relatively young and dynamic subalpine lake ecosystem closed to public recreation, and may inform current and future management decisions.

Methods
eDNA Sampling

On two dates in October 2019 we gathered water samples from 18 sites for eDNA testing. Sites included nearshore waters along the SW shore of Spirit Lake (where NZMS had been observed annually since 2016), tributary streams flowing into Spirit Lake along this shoreline, nearby isolated ponds, and downstream in Coldwater Creek and Coldwater Lake. Flow rates in the tributary streams entering Spirit Lake were generally low in October (0.5-90 L s-1 measured in 2008; Gawel et al. 2018). Flow rates exiting the tunnel outlet and into Coldwater Creek were approximately 2800 L s-1 at the time of sampling (Gawel et al. 2018; US Geological Survey 2024).  At each site two replicate 1000 mL samples were collected in sterile one-liter Nalgene bottles following published protocol (Goldberg and Strickler 2017). Samples were placed on ice and filtered within 12 hours of collection. Both replicates were analyzed for 13 of the sampling sites with single samples run for the remaining 5 sites, in addition to 2 field blanks. Filtered samples and field blanks were delivered to Dr. Caren Goldberg at Washington State University for analysis using published methods (Goldberg et al. 2013).

Aquatic Macrophyte and Snail Sampling

Sampling for aquatic macrophytes with attached snails occurred on multiple dates during the summers of 2021 and 2022. Samples of common and abundant macrophyte taxa (Myriophyllum spicatum, Charales (Class, Charophyceae), Ceratophyllum demersum, Potamogeton amplifolius, and filamentous algae) were gathered using a plant rake at water depths ranging from 1 to 5 m. Macrophyte samples were floated in water in a clear plastic tub and agitated by hand for 30 s to release any attached snails from the vegetation. Snails were allowed to sink to the bottom and separated from the vegetation. Subsequent periodic examination of the vegetation after agitation showed this method to be very effective at separating snails from vegetation, with < 1% of total snail abundance remaining on any vegetation sample. Any additional snails found in rechecked samples were not included in sample totals for method consistency. Macrophytes were then removed by hand, identified, and vegetation volume was estimated using a graduated beaker. The area of vegetation sampled was estimated using the diameter of the rake head, which was rotated on the pole axis to collect a circular sample of macrophytes. The remaining solids and detached snails were then transferred to plastic containers on ice for transport to University of Washington Tacoma for analysis. Snails were separated from any remaining solids, identified, and manually counted under a dissecting microscope. A subset of macrophyte and snail samples collected in 2021 were kept for stable isotope analysis.

Fish Sample Collection and Gut Content Analysis

Annual sampling of rainbow trout began at Spirit Lake in 2000 using hook and line and gillnet methods (Blackman et al. 2018). The majority of these surveys took place in Duck Bay in the SW quadrant of the lake. In 2021 additional hook and line sampling was conducted in locations paired with macrophyte surveys to provide more insight into snail distribution throughout Spirit Lake. Rainbow trout caudal fins from fish collected in 2021 (n= 42) were removed, transported on ice, and dried for stable isotope analysis. Gastrointestinal tracts (fish guts) were collected from destructive hook and line sampling, gill netting, and opportunistically from catch-and-release sampling when fish perished unintentionally. Rainbow trout gut samples were transported to our lab in 90% ethanol where they were dissected, with contents removed from the entirety of the gastrointestinal tract and stored again in 90% ethanol. Snails were then separated from the gut contents, identified, and counted under a dissecting microscope. P. antipodarum and G. deflectus individuals from a subset of rainbow trout guts sampled in 2021 containing both species (n = 10 fish) were stored after counting in 90% ethanol for stable isotope analysis. Midge larvae (Chironomidae) were also present in the gut contents of 9 of these 10 trout, and they were also stored in 90% ethanol for stable isotope analysis.

Arthropod and Amphibian Sample Collection

Aquatic insect larvae samples (Ephemeroptera, Plecoptera, Trichoptera, Odonata), aquatic mite samples (Hydrachna species), and amphibian tadpole samples (Anaxyrus boreas) were collected from benthic (log mat and littoral) substrates in the summers of 2018 and 2021. Arthropod specimens were scraped off submerged substrates with a modified brush and picked off substrates with forceps, and amphibians were captured with handheld dip nets. Emergent Ephemeroptera were collected from emergence traps (amphibious emergence trap - black and white, BugDorm, MegaView Science Co.) over logs and the lake surface. All samples were transported on ice, and stored frozen until they were prepared for stable isotope analysis.

Stable Isotope Sample Preparation

Macrophyte samples were dried at 40 °C to a constant weight, and then pulverized in a mortar and pestle. Aquatic arthropod samples that were large enough were dissected, and head, thorax and legs were preferentially subsampled for analysis. Smaller arthropods, including mites and midge larva, were kept as intact individuals, and whole snail bodies were carefully extracted from their shells. Tadpole tails and caudal fins were subsampled and sonicated in a 2:1 chloroform and methanol solution twice for 30 minutes to remove lipids. Caudal rainbow trout fins were wiped with methanol to clean surface contaminants. The distal 1-2 mm tip of the fin was subsampled for analysis. All animal samples were dried at 40 °C to a constant weight. Macrophyte and animal samples were weighed into tin capsules to a target mass of 1-3 mg, and multiple individuals (n = 3-5) were combined into composited samples for the snails and the smaller arthropods. 

Stable Isotope Sample Analysis

All samples were analyzed for carbon and nitrogen stable isotope ratios (reported as δ¹³C and δ¹⁵N values) at the University of Colorado Earth Systems Stable Isotope Laboratory using an elemental analyzer coupled to a Thermo Delta V Isotope Ratio Mass Spectrometer. Sample δ¹³C and δ¹⁵N values are reported relative to Vienna Pee Dee Belemnite and air N2 standards, respectively.

Stable Isotope Data Analysis

We assumed that the isotopic values of all samples were seasonally comparable and reflected the spring and summer growth season (May-September) in Spirit Lake. We used elemental C and N content to verify that invertebrate samples (comprised of chitin, proteinaceous materials, and soft tissues) were comparable among individuals; weight percent C/N ratios were within the range of 3.2 – 5.5 (except for 6 samples with higher ratios that were not isotopically anomalous and thus retained in the dataset). The tadpole tail and fin samples were fleshy and assumed to be primarily composed of skin, muscle, and connective tissues, and all C/N ratios fell within a range of 4.2 - 4.3. The rainbow trout fin tip sample C/N ratios were relatively invariant (range = 2.9 - 3.4), which suggested there were not substantial differences in tissue composition. We confirmed that the larval and emergent Ephemeroptera δ¹³C and δ¹⁵N values were not significantly different (t-test, p > 0.05) and combined these into a single dataset for that taxon. We used data collected from the few 2021 macrophyte samples to supplement a published Spirit Lake 2018 macrophyte dataset (Shinneman et al. 2024) and confirmed that the isotopic values were comparable, with only one δ¹⁵N value slightly outside the range of the published values. We also used isotopic values for Spirit Lake periphyton reported in Shinneman et al. (2024). We used average isotopic trophic discrimination factors (δ¹³C = 1.5‰, δ¹⁵N = 1.3‰) calculated from a snail feeding study (Li et al. 2018) to directly compare snail isotopic values to those of their most abundant diet sources (periphyton and macrophytes). Since this was not a comprehensive dietary analysis for either snail species, we did not attempt to quantitatively estimate the contributions of diet sources with an isotopic mixing model. We used isotopic trophic discrimination factors determined experimentally from trout (δ¹³C = 1.0‰, Jensen et al. 2012 and δ¹⁵N = 1.7‰, Heady and Moore 2012) to directly compare trout and snail values and demonstrate the improbability of snails as a substantial assimilated diet source. Our study was not a dietary analysis for the rainbow trout and future research may focus on determining the relative contributions of diet sources in Spirit Lake.