Toxicity of clothianidin to common Eastern North American fireflies

Chemicals

We acquired clothianidin from Chem Service (West Chester, PA, USA; purity ≥ 98%), and prepared stock solutions of 2  × 10², 2  × 10³, 2  × 10⁴, and 2  × 10⁵ ng mL⁻¹ clothianidin in acetone (Sigma Aldrich, St. Louis, MO, USA, ACS reagent, purity ≥ 99.5%). Pure acetone served as a control. We stored stock solutions at 4 °C and allowed them to reach room temperature (20 °C) before applying them to soils for the assays.

Firefly collection and colony care

We ran toxicity assays on three separate cohorts of fireflies (Table 1): late-instar larvae from the Photuris versicolor species complex (Pt. Photuris), early-instar Pt. versicolor, and early-instar Photinus pyralis (Pn. pyralis). Both Pt. versicolor and Pn. pyralis are relatively large-bodied (6–20 mm adult body length), widespread firefly species found throughout Eastern North America, and their populations do not appear to be declining (Lewis, 2016). Because both species spend 1–2 years in the soil as larvae and feed on soil invertebrates (Pt. versicolor are thought to feed on a diversity of soil invertebrates whereas Pn. pyralis larvae are considered specialists on earthworms; (McLean, Buck & Hanson, 1972; Buschman, 1984; Lewis, 2016)), they likely experience chronic contact and oral neonicotinoid exposure in contaminated habitats.

Five of the late-instar Pt. versicolor were reared from eggs laid by a mated female collected in late July 2019 from the Bucknell University Chillisquaque Creek Natural Area (Montour Co, PA; 41°01′15″N, 76°44′53″W) , while the other 25 late-instar Pt. versicolor were wild-collected in summer of 2019 from multiple locations throughout Pennsylvania: Bald Eagle State Park (5 August; Centre Co, 41°00′44.0″N 77°12′54.3″W), Allegheny National Forest (24–25 June; Forest Co, 41°31′29.8″N 79°17′33.9″W), and Bucknell University Forrest D. Brown Conference Center (23–24 July; Union Co, PA; 40°57′28″N, 77°00′49″W). Specimens were collected in Bald Eagle State Park, Pennsylvania under permit SFRA-1907 to S. Lower and in Allegheny National Forest under firefly monitoring permit to the Pennsylvania Firefly Festival. Larvae were wild-collected at night by visually inspecting the ground for their faint glows. Larvae were identified to genus by external morphology (McLean, Buck & Hanson, 1972). After collection, we housed individual larvae in 16-oz clear plastic deli containers (11.5-cm diameter × 8-cm tall) lined with moist filter paper. Every 1–2 weeks, we provided each larva with one piece of cat food (Grain-Free Real Chicken Recipe Dry Cat Food, Whole Earth Farm™, Merrick Pet Care Inc., Amarillo, TX, USA), which had been softened in DI-water for 1 h (McLean, Buck & Hanson, 1972). After 24 h, we removed cat food and replaced the filter paper. Occasionally there was extensive fungal growth on the cat food, which could be fatal to Pt. versicolor larvae; in these instances, we gently wiped larvae with DI water and a delicate task wipe then transferred them to clean containers.

Early-instar Pt. versicolor and Pn. pyralis cohorts were reared from eggs laid in July 2020. On the evening of 10 July 2020, we collected 3 male and 2 female Pt. versicolor adults and 3 mated Pt. versicolor females. Flying Pn. pyralis males were collected and identified based on their characteristic “J” flash pattern (Lewis, 2016) while female Pn. pyralis were collected from nearby patches of short grass and were identified based on their flash pattern and similar morphology to the Pn. pyralis males (Lewis, 2016). Female Pt. versicolor were collected near Pn. pyralis females and identified based on their green-shifted flash color and morphology (Lewis, 2016). Additional Pn. pyralis males were collected to provision the mated Pt. versicolor females. We collected Pt. versicolor and Pn. pyralis in a residential area (State College, Centre Co, PA; 40°47′03″N, 77°52′25″W) into two separate 16-oz deli container “nurseries” kept at ambient room temperature (20–22 °C); each nursery contained a handful of moist sphagnum moss on top of moist soil (2-in deep; silt loam, collected from certified organic fields at the Russell E. Larson Agricultural Research Center at Rock Springs, PA, USA; 40°42′52″N, 77°56′46″W). Both Pn. pyralis females mated within a few minutes of collection.

Female Pt. versicolor and Pn. pyralis laid eggs within the following 3 days (50+ Pt. versicolor eggs and 100+ Pn. pyralis eggs; we did not attempt more accurate counts to avoid damaging eggs). Under ambient temperatures (20–22 °C), first-instar larvae of both species began to emerge three weeks after eggs were laid (5 August 2020). We kept Pt. versicolor larvae in the nursery chambers for two weeks, and then, after we observed significant cannibalism among larvae, moved them into individual soil-lined 1-oz polypropylene portion containers. As with larvae collected and reared from 2019, developing Pt. versicolor were fed moistened cat food (Grain-Free Real Chicken Recipe Dry Cat Food, Whole Earth Farm™, Merrick Pet Care Inc., Amarillo, TX, USA) in addition to pieces of freeze-killed Lumbricus terrestris (Josh’s Frogs, Owosso, MI, USA). As evidence of the hypothesis that Pn. pyralis larvae are specialist on earthworms, Pn. pyralis larvae did not feed on cat food but did feed gregariously on freeze-killed L. terrestris. Unlike Pt. versicolor, Pn. pyralis failed to thrive in isolation, so they were kept in the nursery chamber until starting the toxicity assay.

Toxicity assay on Late-instar Photuris versicolor

We started the toxicity assay with late-instar Photuris versicolor on 22 June 2020. We used 1-oz polypropylene portion containers containing 8 g of dry soil (same soil source as nursery chambers) for our assay containers. To the soil in each assay container, we added 0.5 mL of the appropriate clothianidin stock solution, allowed the acetone to completely evaporate, then added two mL of DI water to moisten the soil and to achieve clothianidin concentrations of 0 ng g⁻¹, 10¹ ng g⁻¹ soil, 10² ng g⁻¹ soil, 10³ ng g⁻¹ soil , 10⁴ ng g⁻¹ soil. We chose this concentration range (10¹–10⁴ ng g⁻¹ soil) to encompass the range of neonicotinoid concentrations in soil that have been measured in potential firefly habitats (Knoepp et al., 2012; Schaafsma et al., 2015; Pearsons et al., 2021).

After setting up assay containers, we weighed the late-instar Pt. versicolor and randomly assigned each to a particular clothianidin concentration (ensuring all larvae in each dose-set were sourced from the same location). All late-instar Pt. versicolor were over 12 months old at the start of the assay, and were over 10-mm long and >50 mg (Table 1). Each clothianidin concentration (0, 10¹ ng g⁻¹ soil, 10² ng g⁻¹ soil, 10³ ng g⁻¹ soil , 10⁴ ng g⁻¹ soil) was replicated six times (30 late-instar Pt. versicolor larvae in total). We recorded firefly status at 1, 4, and 24 h, and every day for an additional 99 d. Fireflies were categorized as dead (D), exhibiting a toxic response (T), or apparently healthy (A). A larva was assumed dead if it did not respond to gentle prodding with forceps. If a larva was flipped on its back and/or demonstrating repetitive twitching of its legs or head, it was recorded as exhibiting a toxic response (T). Fireflies were recorded as apparently healthy (A) if they exhibited a usual response to prodding from blunt forceps (Fig. 1A; quickly curled up on its side, often glowing). During the toxicity assay, we fed larvae once a week by carefully transferring individuals out of the assay containers into clean containers lined with moistened filter paper and containing a piece of moistened cat food. After 24 h, we returned fireflies to the assay containers and noted if the cat food had obvious signs of feeding (Fig. 1B). Feeding activity for each week was measured as a simple binary (0 = no obvious signs of feeding, 1 = obvious signs of feeding). At each status check, we noted if a firefly had constructed a protective soil chamber, then carefully dismantled the chamber to check larval status. Larvae often re-built soil chambers by the next day; if a larva built soil chambers on multiple consecutive days (feeding days as an exception), we noted this behavior as a “period of chamber formation.” Assay containers were kept in a dark drawer except when doing daily checks, and we misted containers with DI water as needed to keep the soil from drying out.

Toxicity assay on early-instar Photuris versicolor

The toxicity assay with early-instar Photuris versicolor was similar to the assay with late-instar larvae, except we added half the amount of soil (4 g) and half the volume of clothianidin stock solutions (0.25 mL) to each assay container to achieve the same clothianidin concentrations (0, 10¹ ng g⁻¹ soil, 10² ng g⁻¹ soil, 10³ ng g⁻¹ soil , 10⁴ ng g⁻¹ soil). All early-instar Pt. versicolor were less than 3 months old and weighed between 3 and 15 mg. On 17 Sept 2020, we started trials with early-instar Pt. versicolor (three replicates at each concentration, 15 larvae in total), feeding them cat food once a week and recording their status at 1, 4, and 24 h, and every day for 10 d, then twice a week for an additional 90 d. Unlike for late-instar Pt. versicolor, we fed early-instars by directly placing moistened cat food in the assay containers (we removed the food 24 h later after noting if food had been damaged [1] or not [0]).

Toxicity assay on early-instar Photinus pyralis

As with the early-instar Pt. versicolor assay, the Photinus pyralis assay was run in 1-oz polypropylene portion containers containing 4 g of soil with 0.25 mL doses of clothianidin stock solutions (to achieve 0, 10¹ ng g⁻¹ soil, 10² ng g⁻¹ soil, 10³ ng g⁻¹ soil , 10⁴ ng g⁻¹ soil). All early-instar Pn. pyralis were less than 3 months old and weighed between 0.6 and 2.4 mg. On 17 Sept 2020, we started the assay on early-instar Pn. pyralis, exposing larvae in sets of five (five larvae per container, three replicates at each concentration, 75 larvae in total), recorded their status at 1, 4, and 24 h, and every day for 10 d, then at least twice a week for an additional 20 d. We terminated the Pn. pyralis assay earlier than the Pt. versicolor assays due to an acarid mite infestation, which rapidly increased larval mortality across all doses. During the assay, we fed Pn. pyralis pieces of earthworm (L. terrestris) in the same manner that early-instar Pt. versicolor were fed cat food.

Statistical analysis

We performed all statistical analyses in R (v4.0.4) (R Core Team, 2021). For each firefly cohort, we calculated median toxic concentrations (TC₅₀) and median lethal concentrations (LC₅₀) at 24 h, 7 d, and 30 d of exposure using probit analysis (LC_PROBIT from the “ecotox” package; Robertson et al., 2017; Hlina et al., 2019); for TC₅₀ estimates, we included both sub-lethal and lethal responses, while LC₅₀ estimates were based on mortality alone. To assess long-term survivorship across clothianidin levels, we used the Kaplan–Meier method (“survival” functions SURVDIFF and PAIRWISE_SURVDIFF; Therneau, 2021; Therneau & Grambsch, 2000). To determine how clothianidin exposure affected firefly behavior, we used non-parametric Mann–Whitney U tests (WILCOX.TEST) to compare feeding frequency and soil-chamber construction across clothianidin doses; we made pairwise comparisons using Wilcoxon rank sum tests with continuity corrections (PAIRWISE.WILCOX.TEST). As firefly larvae reduce feeding before pupation (McLean, Buck & Hanson, 1972), we excluded the two feeding events preceding pupation for feeding assessments.

24 h, 7 d, and 30 d TC₅₀ and LC₅₀ estimates

Dose–response curves and estimated TC₅₀ and LC₅₀ indicate that Photuris versicolor and Photinus pyralis were surprisingly tolerant of exposure to clothianidin (Table 2 and Figs. 2–4). Reliable TC₅₀ and LC₅₀ estimates were limited by our small sample sizes and low acute mortality within the tested concentration range. Overall, TC₅₀ values ranged from 500 ng g⁻¹ to 2,000 ng g⁻¹ while LC₅₀ values exceeded our test limit (above 10,000 ng g⁻¹).

Firefly survival

Clothianidin exposure significantly reduced long-term firefly survival at high concentrations (Fig. 5). Between one and four hours after initial exposure, half of the late-instar Pt. versicolor larvae and 87% of the early-instar Pn. Pyralis larvae exposed to the highest clothianidin concentration (10,000 ng g⁻¹) began to exhibit toxic responses. By 24 h, all six late-instar Pt. versicolor exposed to the highest clothianidin concentration (10,000 ng g⁻¹) exhibited a toxic response (Fig. 2A); these larvae never recovered and died by day 84. Photuris larvae were somewhat tolerant to lower clothianidin concentrations (10 ng g⁻¹ or 100 ng g⁻¹) and neither late- nor early-instar larvae exposed to low concentrations had significantly lower 100 d survival probability compared to controls (Figs. 5A–5B). All Pt. versicolor in the control treatment either pupated (2 out of 6 late-instar larvae) or survived through day 100 (4 out of 6 late-instar larvae, all three early-instar larvae). Although the experiment was terminated at 30 d due to the mite infestation, early-instar Pn. Pyralis exposed to 1,000 ng g⁻¹ clothianidin showed marginally non-significant reduced survivorship (P = 0.07) while Pn. pyralis exposed to 10,000 ng g⁻¹ clothianidin showed significantly reduced survivorship (P < 0.0001) compared to controls (Fig. 5C).

Feeding behavior

Clothianidin exposure significantly reduced the number of times firefly larvae fed (Fig. 6). During the toxicity assays, no Pn. pyralis or Pt. versicolor larvae exposed to the highest clothianidin concentration (10,000 ng g⁻¹ soil) fed. Late-instar Pt. versicolor exposed to 1,000 ng g⁻¹ soil fed significantly less frequently than control larvae (χ²₄ = 16.3, P = 0.003), and early-instar Pn. pyralis larvae fed significantly less at higher doses (1,000 ng g⁻¹ and 10,000 ng g⁻¹) compared to the control or lower doses (χ²₁ = 12.4, P = 0.0004).

Soil-chambers, molting, and pupation of late-instar Photuris versicolor

The 14 late-instar Photuris larvae that survived as larvae through day 100 went through 1 to 5 periods of consecutive days when they regularly formed protective soil chambers (median = 2 periods) and spent anywhere from 1 to 20 total days in soil chambers (median = 9 d). Larvae exposed to 10,000 ng g⁻¹ clothianidin never constructed soil chambers while larvae exposed to 1 ppm clothianidin spent significantly fewer days in soil chambers than larvae exposed to 10 ng g⁻¹ (P = 0.01; Fig. 7).

Formation of protective soil chambers did not correspond with molting or pupation, and all recorded molting and pupation events occurred outside soil chambers, on the soil surface. Late-instar Pt. versicolor larvae only molted once or twice, irrespective of how frequently or for how long they built soil chambers (larvae that survived through 100 days; frequency: R²_adj = −0.09, F_1,10 = 0.10, P = 0.76; duration: R²_adj = −0.02, F_1,10 = 0.81, P = 0.39). Six of the thirty late-instar Pt. versicolor larvae pupated; five of which successfully eclosed within 35 d of starting the assay (two controls, one at 10 ng g⁻¹, two at 100 ng g⁻¹) and one which was unsuccessful (1,000 ng g⁻¹). The unsuccessful larva failed to shed its last-instar exoskeleton and died during the pupal stage. At 35 d, three of the larvae exposed to the highest clothianidin concentration (10,000 ng g⁻¹) were still alive, but none of these larvae ever entered a pupal stage. Of the individuals that successfully eclosed, three were lab-reared from eggs laid in 2019 (3 out of 5) while only two were wild-collected (2 out of 25).