Chemosphere 185 (2017) 1019e1029

Contents lists available at ScienceDirect

Chemosphere

journal homepage: www.elsevier.com/locate/chemosphere

Lethal and sub-lethal effects of commonly used anti-sea lice formulations on non-target edwardsii larvae

* Paulina Gebauer a, , Kurt Paschke b, d, Claudia Vera b, Jorge E. Toro c, Miguel Pardo c, d, Mauricio Urbina e a Centro i~mar, Universidad de Los Lagos, Casilla 557 Puerto Montt, Chile b Instituto de Acuicultura, Universidad Austral de Chile, Casilla 1327, Puerto Montt, Chile c Instituto de Ciencias Marinas y Limnologicas, Facultad de Ciencias, Universidad Austral de Chile, Valdivia, Chile d Centro FONDAP de Investigacion en Dinamica de Ecosistemas Marinos de Altas Latitudes (IDEAL), Chile e Departamento de Zoología, Facultad de Ciencias Naturales y Oceanograficas, Universidad de Concepcion, Chile highlights

Cypermethrin, deltamethrin, and azamethiphos affected 100% crab larvae at concentrations lower than used against sea-lice. Hydrogen peroxide at the concentration used as an anti-sea lice treatment had lethal and sub-lethal effects on M. edwardsii zoea I. Repeated exposure to azamethiphos (0.0625e0.5 mgL 1) increased mortality, but did not affect zoea I developmental time. Chronic exposure to hydrogen peroxide (187.5e1500 mg L 1) had a lethal effect on larvae. article info abstract

Article history: The pesticides used by the salmon industry to treat sea lice, are applied in situ via a bath solution and are Received 8 February 2017 subsequently discharged into the surrounding medium. The effects of cypermethrin, deltamethrin, Received in revised form azamethiphos and hydrogen peroxide were assessed on the performance of Metacarcinus edwardsii 10 July 2017 larvae, an important crab for Chilean fishery. All larvae were dead or dying after 30 min of exposure to Accepted 19 July 2017 cypermethrin and after 40 min to deltamethrin at concentrations 100 and 20 times lower (0.15 and Available online 24 July 2017 0.1 mgL 1, respectively) than the concentrations and exposure times recommended by the manufacturers Handling Editor: Jim Lazorchak (CRM) to treat sea lice. Azamethiphos affected all larvae at a concentration 10 times lower than CRM. Hydrogen peroxide had the lowest detrimental effects, but at the CRM, 100% of the larvae were affected. Keywords: Sub-lethal effects, i.e prolonged developmental time, were observed at concentrations lower than CRM. Cypermethrin Repeated exposure to azamethiphos (0.0625e0.5 mgL 1) and hydrogen peroxide (188e1500 mg L 1) had Deltamethrin effects on survival. In conclusion, the pesticides used against parasitic copepod tested here, negatively Azamethiphos affect non-target larvae. Due to the product’s characteristics, the lethal effects of the pyre- Hydrogen peroxide throids probably are restricted to the time and area of application, while the action of azamethiphos may Organophosphate extend to a wider area. Current data are insufficient to accurately dimension the effects of these com- Crustacean larvae pounds in the field. More research is required to evaluate the consequences of prolonged developmental times and/or reduction in appendage mobility, so as the effects of these compounds on the pelagic and benthic communities. © 2017 Elsevier Ltd. All rights reserved.

1. Introduction farming is susceptible to various bacterial, viral, and parasitic dis- eases, due to the high density of cultivation, requiring the use of In the Southern Hemisphere, salmon and trout are intensively both antibiotic and antiparasitic treatments (Burridge et al., 2010). farmed in the fjords and channels of Chilean Patagonia. Salmon Among them, sea lice pose a serious threat to productivity in both salmon and trout farming (Burridge et al., 2014). In global monetary terms, costs associated to sea lice control equate to around 0.19 V * Corresponding author. kg 1 of salmon produced (Costello, 2009). In Chile, the E-mail address: [email protected] (P. Gebauer). http://dx.doi.org/10.1016/j.chemosphere.2017.07.108 0045-6535/© 2017 Elsevier Ltd. All rights reserved. 1020 P. Gebauer et al. / Chemosphere 185 (2017) 1019e1029 predominant parasite is the copepod Caligus rogercresseyi. Sub-lethal effects, such as changes in locomotion, were reported Control measures have included the application of chemical com- after exposing adult American lobster to azamethiphos (Abgrall pounds (Aaen et al., 2015), such as organophosphates (since 1981) et al., 2000), and prawn (Palaemon serratus) to deltamethrin specifically azamethiphos (since 2013), synthetic pyrethroids (e.g. (0.6 ng L 1)(Oliveira et al., 2012). In several species of free-living cypermethrin and deltamenthrin, used since 2007 (Aaen et al., copepods, both swimming and feeding behaviours are altered by 2015), and hydrogen peroxide to a lesser degree. Large volumes concentrations of 5 mg L 1 hydrogen peroxide (Van Geest et al., of these pesticides are discharged into the Chilean marine 2014) and 0.098e0.32 mgL 1 cypermethrin (Barata et al., 2002). ecosystem; for example, 677 kg of active compound cypermethrin Damage by oxidative stress as protein carbonylation were reported and 197 kg of deltamethrin were discharged in 2012 (Helgesen in H. americanus as a result of chronic exposure to 61 ng L 1 aza- et al., 2014). Despite these levels of discharge, very little is known methiphos, and lipid oxidation was reported in Penaeus monodon about the potential negative effects on aquatic life. after 4 days of exposure to 0.1 mgL 1 deltamethrin (Dorts et al., These pesticides have a different mechanism of action. Organ- 2009). Both types of damage were observed in Procambarus clar- ophosphates are neurotoxics that inhibit the enzymatic activity of kii after exposure to 0.005 mgL 1 cypermethrin for 24 h (Burridge acetylcholinesterase (AChE), the enzyme responsible for acetyl- et al., 2014). These previous studies indicate a range of both lethal choline hydrolysis (Intorre et al., 2004). Pyrethroids act on neural and sub-lethal effects on non-target organisms from the applica- transmission by interfering with sodium channels (Miller and tion of the various families of pesticide chemicals. Adams, 1982), resulting in motor neuron depolarisation and re- Our aim was to assess the effects of commercial pesticides, petitive discharge in the nerve endings, which may lead to paralysis commonly used in the treatment of C. rogercresseyi on fish farms, on and death (Crane et al., 2011; Haya et al., 2005). Hydrogen peroxide the larval performance of the South American crab species acts as an oxidising agent generating gaseous oxygen, which re- M. edwardsii. This commercially important crab species is abundant mains trapped inside the cuticle of the . These chemical on the coasts of Southern Chile, accounts for nearly 80% of the compounds are applied in situ by enclosing net pens and bathing artisanal crab fishery, and is an important component of the local fish using a tarpaulin. After completion of the chemical treatment, economy (Chilean national fishery statistics, SERNAPESCA, the compound is then, released into the surrounding seawater. 2011e2014). The effects of acute (based on the manufacturers’ Dispersion and dilution rates of the chemicals depend on the recommendations, time and concentration) and chronic exposure characteristics of each pesticide and the hydrographic character- to pesticides on the first larval stage (zoea I) of M. edwardsii were istics of the area. A residual chemical compound concentration investigated. Developmental time and survival were used as persists in the area around the treated pens, which is likely to response variable; we also estimated the 50% lethal and effective impact non-target species. (Burridge et al., 2014; Ernst et al., 2014). concentrations. Given the generality and broad action spectrum of Specifically, concentrations of cypermethrin between 0.185 and the formulations used to treat sea lice in Chilean salmon-farming, 0.218 mgL 1 have been measured 25 m away from the net pen we hypothesize cypermethrin, deltamethrin, azamethiphos and treated, 25min after final treatment (Hunter and Fraser,1995; Willis hydrogen peroxide will negatively affect the developmental time et al., 2005), while concentrations of deltamethrin between 0.020 and survival of the zoea I of M. edwardsii. and 0.040 mgL 1 have been detected 4e5 h after treatments with this compounds (Ernst et al., 2014). For azamethiphos concentra- 2. Materials and methods tions between 1 and 25 mgL-1 have being measured from 1 to 1000 m away from the application area (Ernst et al., 2014). 2.1. Larvae collection The majority of research assessing the lethal effects of pesticides used on sea lice have been conducted on adult of non-target Ten ovigerous M. edwardsii females were kept in containers with crustaceans, determining lethal and/or effective concentrations circulating seawater until larvae hatched (~12 C, 32 PSU for 1 (LC50,EC50) (see more detailed Table 1 Ernst et al., 2001; Mayor month). Recently hatched zoea I were collected and actively et al., 2008; see more detailed Table 5 Van Geest et al., 2014). swimming larvae were selected for the exposure to cypermethrin ® ® ® However, very little is known about the lethal effects of these (Betamax ), deltamethrin (Alpha Max ), azamethiphos (Calfree ), ® chemical compounds on the early life stages of decapod crusta- and hydrogen peroxide (Hyperox ). ceans. It is useful to note that LC50 has been calculated for American lobster (Homarus americanus) larvae for cypermethrin (Burridge et al., 2000a, 2000b; Pahl and Opitz, 1999), deltamethrin 2.2. Acute exposures (Burridge et al., 2014), azamethiphos (Burridge et al., 2000b, 2014; Pahl and Opitz, 1999), and hydrogen peroxide (Burridge et al., Four independent experiments were carried out, in which 2014), providing some indication of the expected effects. different groups of zoea I larvae (2800 larvae per experiment) were exposed to each of the four compounds. For each compound

Table 1 Nominal concentrations of cypermethrin, deltamethrin, azamethiphos and hydrogen peroxide, and exposure times for the acute experiments in M. edwardsii larvae. CRM: concentration recommended by the manufacturer. Exposure times for each compound are the recommended by the manufacturers for treatment of C. rogercreseeyi.

Compounds Treatments-Concentrations

Control (seawater) A1 A2 A3 A4 A5 (CRM) A6 A7

Cypermethrin 0 0.15 0.5 1.5 5 15 30 60 (mgL 1, 30 min) Deltamethrin 0 0.1 0.25 0.5 1 2 3 8 (mgL 1, 40 min) Azamethiphos 0 1 3 10 30 100 300 500 (mgL 1, 30 min) Hydrogen peroxide 0 50 100 300 750 1500 2000 3000 (mg L 1, 20 min) P. Gebauer et al. / Chemosphere 185 (2017) 1019e1029 1021

(experiment), a working solution was prepared in UV-sterilised 2.4. Statistical analysis seawater and filtered using a 1 mm filter. The required compound solution volume was extracted by repetitive pipette and added to The effect of each chemical compound on the developmental the glass exposure containers filled with 3 L filtered and sterilised time and mortality of the first larval stage (zoea I) was assessed seawater, mixed for 2 min with a glass rod, in order to reach the using analysis of deviance, considering a gamma distribution for corresponding concentrations before larvae were exposed. In each the developmental time and a binomial distribution for the mor- experiment, different larvae groups were exposed to one of seven tality. Pairwise comparison were carried out using Tukey test, different concentrations (7 replicates per concentration, implemented in the R library multicomp (Bretz et al., 2016). The comprising 50 larvae each; 350 larvae per concentration) of each of homogeneity of variance was assessed using Levene’s test. The 50% the four compounds, plus a control group. The exposure times were lethal and effective concentrations for the chemical compounds based on the time recommended by manufacturer for treatment were also calculated, adjusting the survival data to a sigmoidal against C. rogercreseeyi (Table 1); the control group experienced the dose-response curve to calculate the concentration at which 50% of same handling as other treatments but it was only exposed to the larvae died (LC50) or were died or dying (EC50). The analyses seawater. were carried out using R, version 3.2.4. Following exposure, larvae were transferred to clean containers of filtered seawater and rinsed three times (one minute each time). 3. Results Immediately after rinsing, and before being placed into the follow- up containers, the larvae were observed and registered into one of 3.1. Acute exposure four groups: a) dead; b) dying A, those lying on the bottom of the container with weak appendage movements only observable under 3.1.1. Cypermethrin stereoscopic microscope; c) dying B, those on the bottom of the Larvae exposed to cypermethrin for 30 min presented 100% container, unable to swim but with active appendage movement, mortality at 0.5, 1.5, 5, 15 (CRM), 30 and 60 mgL 1.At0.15mgL 1, all observable without stereoscopic microscope; and d) alive (i.e. larvae were classified as dying A, and 78.09± 1.53% of these died actively swimming larvae). within 24 h after exposure while the rest remained dying. After Replicates were kept in 800 mL follow-up containers filled with 48 h, accumulated mortality for this treatment was 98.83± 2.90% seawater, filtered at 1 mm, without aeration, at 15 C, 32 PSU and (Fig. 1) reaching 100% mortality at 96 h post-exposure. Control 12:12 light:darkness photoperiod. Daily seawater and food (freshly mortality was 21.48± 3.78% after 24 h, and were the only larvae that hatched Artemia sp, nauplii) were changed. Larvae were counted survived (34.44± 3.03%) and moulted to zoea II (7.77 ± 0.11 days). and classified as alive, dying (dying A or B), or dead, and the presence of exuviae was checked. The experiments were monitored 3.1.2. Deltamethrin until all larvae were classified as either dead or moulted to larval After 40 min of exposure to 0.1, 0.25, 0.5 and 1 mgL 1 delta- stage zoea II. methrin, 100% of M. edwardsii larvae were classified as dying B (unable to swim but with observable active appendage movement). At the three highest concentrations (2, 3, 8 mgL 1), ~33% of the 2.3. Chronic exposures larvae died, and the remaining presented only weak appendage movements (dying A). All larvae in the control swam actively. There fi Chronic exposure experiments were only carried out for aza- were signi cant differences in mortalities between groups 24 h c2 ¼ < methiphos and hydrogen peroxide, since 100% of the larvae post-pesticide exposure ( (48) 50.68; P 0.0001), and pairwise ± exposed to the pyrethroids (i.e. cypermethrin and deltamethrin), comparisons (Tukey) revealed four groups: a) control (3.50 0.96% m 1 were either dead or dying 24 h post-exposure at the lowest con- mortality), b) the treatments 0.1 and 0.25 gL (between ± ± centrations (0.15 and 0.1 mgL 1, respectively). Each experiment 9.70 1.58% and 11.10 1.68% mortality), c) treatments 0.5 and m 1 consisted of 1750 larvae divided into four concentrations (Table 2), 1 gL (~50% mortality), and d) the three highest concentrations ± ± and a control group consisting of clean seawater. Each treatment (between 89.80 1.64% and 91.66 1.51% mortality). The LC50 ± m 1 and control group contained seven replicates with 50 zoea I larvae calculated at 24 h was 1.252 0.046 gL , roughly 37% lower than m 1 per replicate. the concentration recommended by manufacturer (2 gL ). The rearing conditions (i.e. water changes, feeding, selection, However, it is worth noting that, while the LC50 is an estimate based and larvae classification) remained the same as for acute exposure. on mortality, 100% of the larvae were negatively affected by this Larvae were exposed daily over a seven-day period at the con- concentration. Over the same period, dead and dying larvae were centrations and exposure times indicated in Table 2. This chronic found in all treatment groups, and the accumulated mortality was exposure period is close to the moulting time of zoea I to zoea II, so 100% in all concentrations 48 h post-exposure (Fig. 2). The larvae ± the exposure of only zoea I stage was assured. Each replicate group from the control group were the only survivors (35.09 3.37%), and ± was exposed independently over the full period of the experiment. moulted to next stages at 8.37 0.14 days.

3.1.3. Azamethiphos The larvae in the control, 1 and 3 mgL 1 azamethiphos treat- Table 2 ments swam actively post-exposure, while the larvae exposed to Nominal concentrations of azamethiphos and hydrogen peroxide, and exposure 10, 30, 100, 300, and 500 mgL 1 azamethiphos were all dying after times for the chronic experiments in M. edwardsii larvae. CRM: concentration rec- m 1 ommended by the manufacturer. 30 min of exposure. In the 10 and 30 gL treatments, larvae were in category dying B, moving actively the appendages but Compound Treatments-Concentrations completely unable to swim or capture food. Larvae exposed at Control (seawater) C1 C2 C3 C4 100 mgL1 presented weak appendage movement and curled 1 Azamethiphos 0 0.0625 0.125 0.25 0.5 telson, and those exposed to 300 and 500 mgL concentrations (mgL 1, 30 min) were classified as dying A. At 24 h post-exposure, there was sig- Hydrogen peroxide 0 188 375 750 1500 fi c2 ¼ ni cant larval mortality among the treatments ( (48) 49.77; (mg L 1, 20 min) P < 0.0001). Tukey pairwise comparisons identified two groups: a) 1022 P. Gebauer et al. / Chemosphere 185 (2017) 1019e1029

Fig. 1. Cumulative percent of dead and dying larvae 24 and 48 h following 30 min exposure to cypermethrin at different concentrations. Arrow indicates the concentration recommended by the manufacturer (CRM). Data represented at means ± SE.

Fig. 2. Cumulative percent of dead and dying larvae 24 and 48 h following 40 min exposure to different concentrations of deltamethrin. Arrow indicates concentration recom- mended by the manufacturer (CRM). Data represent means ± SE. P. Gebauer et al. / Chemosphere 185 (2017) 1019e1029 1023

0, 1, and 3 mgL 1, with a mean mortality of 39.13% and b) the other 3.2. Chronic exposure treatments (from 10 to 500 mgL 1), with a mortality that varied between 91 and 100%. The LC50 at 24 h post-exposure was 3.2.1. Azamethiphos 2.84 ± 0.39 mgL 1. The effect of the two lowest azamethiphos Mortality of M. edwardsii larvae 24 h after first exposure to concentrations (1 and 3 mgL 1) 24 h post-exposure can be clearly azamethiphos showed significant differences between the control 2 seen in the number of dead and dying larvae. These treatments and treatment groups (c (30) ¼ 121.97; P < 0.001). The greatest showed 60% of larvae dead and dying versus 34% in the control increase in mortality for all concentrations occurred 48 h post- 2 group (c (48) ¼ 42.18; P < 0.0001). The EC50 calculated for 24 h post- exposure, after the second exposure to azamethiphos (Fig. 5A). exposure (dead þ dying) was 0.94 ± 0.15 mgL 1. Differences in survival between control and the treatment groups At 48 h post-exposure, mortality in the 1 and 3 mgL 1 treat- remained throughout the experiment, with the exception that day- ments increased to 30e58%, reaching roughly 80% accumulated seven survival in the 0.5 mgL 1 treatment differed from the other mortality compared to only 38% in the control group (Fig. 3A). In the treatments, which coincided with the first day of moults to zoea II. control group, 27.89± 1.73% of the larvae survived, moulting to zoea The developmental time of zoea I in the control group was II after 7.34 ± 0.05 days. In treatments 1 and 3 mgL 1, only 1.71% and 9.51 ± 0.07 days; this presented no difference with the treatments 3.04% of the larvae reached zoea II after 7.4 ± 0.24 and 7.11 ± 0.11 (F(4-30) ¼ 1.31; P > 0.05), which ranged between 9.37 ± 0.23 and days, respectively (Fig. 3B). Significant differences were observed in 9.87 ± 0.11 days, at 0.5 and 0.00625 mgL 1, respectively (Fig. 5B). the total zoea I mortality between treatment and control groups The survival of the first larval stage was affected significantly 2 2 (c (18) ¼ 26.95; P < 0.0001), while no difference was shown in (c (30) ¼ 104.56; P < 0.0001). The control had the highest survival 1 developmental time (F(2-11) ¼ 0.917; P > 0.05). (78.28± 2.22%), followed by the 0.0625, 0.125, and 0.25 mgL treatments, with survival roughly 57%; the percent survival of the 0.5 mgL 1 treatment was 45.71 ± 2.66.10%. (Fig. 5B).

3.2.2. Hydrogen peroxide 3.1.4. Hydrogen peroxide Between 16.28 ± 1.97 and 27.42± 2.39% of larvae were affected After 40 min of exposure, 100% larvae at 50, 100 and 300 mg L 1 (dead þ dying) (Fig. 6), significantly different from the control 2 hydrogen peroxide and control swam actively, while 85% of the (c (30) ¼ 51.46; P < 0.001) 24 h after the first exposure to hydrogen larvae exposed to 750 mg L 1 swam actively and the remaining 15%, peroxide. Following two exposure at 1500 mg L 1, over 50% of the lay on the bottom but still presenting active movement of their larvae were affected, while at lower concentrations appendages. The larvae exposed to 1500, 2000, 3000 mg L 1 were (187.5e750 mg L 1), between 22.2% and 35.1%. After the third classified as dying B. The percentage of affected larvae hydrogen peroxide exposure of 375 and 750 mg L 1, more than 60% (dead þ dying) 24 h post-exposure, for concentrations between of the larvae were affected. Four groups were distinguished sta- 0 and 750 mg L 1, ranged between 16.27± 1.89% and 22.71± 2.05%. tistically (Tukey test) at 72 h: a) the control, with 15.5% of larvae The 1500 mg L 1 treatment recorded 41.43± 2.50% of larvae affected, b) the 187.5 mg L 1 treatment, with 24.9%, c) the 375 and affected, with 23% dying. The 2000 and 3000 mg L 1 treatments 750 mg L 1 treatments, with roughly 62%, and d) the 1500 mg L 1 affected more than 50% of the larvae. At 48 h post-exposure, 39.2% treatment, with 88.6% larvae affected (Fig. 6A). Close to 100% of the of larvae in the 1500 mg L 1 treatment group (CRM) were nega- larvae were affected after the fourth exposure to 1500 mg L 1 and tively affected, whilst 69.80± 2.30% and 98.90± 0.50% of larvae in the four groups described above were maintained. No larvae were the 2000 and 3000 mg L 1 treatments were affected, respectively active after 6 exposures at 375 and 750 mg L 1. After 7 exposure, 1 (Fig. 4A). The EC50 (dead þ dying) calculated 48 h post-exposure 1.43% of larvae (zoea I) in the 187.5 mg L treatment moulted to was 1269.61 ± 41.33 mg L 1. Comparing larval mortality 72 h and zoea II (9.22 ± 0.33 days). The rest of the larvae remained as zoea I 96 h post-application showed few variations. The accumulated for a maximum of 22 days. The control larvae moulted to zoea II on mortality presented significant differences between treatments average after 9.51 ± 0.22 and showed 78.28± 2.2% survival (Fig. 6B). 2 1 (c (48) ¼ 150.99; P < 0.001): concentrations up to 750 mg L showed no differences (Tukey test P > 0.05), with a mean mortality 4. Discussion of 30.2± 2.07%. The accumulated mortality of the other treatments, ranged between 41.9± 2.5% and 98.3± 0.6%. The EC50 (dead þ dying) Zoea I larvae of M. edwardsii, are negatively impacted by all the after 72 h was 1130.19 ± 42.14 mg L 1 hydrogen peroxide, and commonly used pesticides for the copepod parasite C. rogercresseyi slightly lower after 96 h (1036.25 ± 43.44 mg L 1). (deltamethrin, cypermethrin, azamethiphos and hydrogen Both zoea I developmental time and survival were affected by peroxide) tested here. M. edwardsii is of crucial socioeconomic exposure to the different concentrations of hydrogen peroxide (F(6- importance in southern Chile, and as its distribution coincides with 2 41) ¼ 319.06; c (48) ¼ 107.16; P < 0.0001, Fig. 4B). The developmental the salmon farming industry, it will likely be frequently exposed to time of the control larvae was 8.54 ± 0.08 days, and those of the 50, these pesticides. The concentrations and exposure times recom- 100, and 300 mg L 1 treatments were 8.65 ± 0.09, 8.2 ± 0.08, and mended by the manufacturers have lethal effects on the zoea I 8.63 ± 0.08 days, respectively (Tukey test P > 0.05). Larvae exposed under acute exposure, with cypermethrin, deltamethrin, and aza- to 750, 1500, 2000, and 3000 mg L 1 hydrogen peroxide suffered methiphos producing 100% mortality. Acute exposure to hydrogen delayed development, moulting to zoea II after 11.58 ± 0.11, peroxide at 1500 mg L 1 (CRM) generates both lethal (~67% mor- 11.78 ± 0.12, 12.13 ± 0.13, and 12 days, respectively, these concen- tality) and sub-lethal effects (delayed moult to zoea II in 38% of the trations showed significant differences with the control and 50, 100 surviving larvae). Chronic exposure to hydrogen peroxide at con- and 300 mg L 1 treatments (Tukey test, P < 0.05). There were sig- centrations between 187.5 and 1500 mg L 1 provokes high mor- 2 nificant differences in survival between treatments (c (48) ¼ 107.16; tality, but has no effect on the developmental time of the few P < 0.0001), and pairwise comparisons (Tukey) revealed three surviving larvae. groups: a) 0, 50, 100, and 300 mg L 1 treatments (55.89± 2.01% Chronic exposure of M. edwardsii zoea I for 7 days to azame- survival), b) 750 and 1500 mg L 1 treatments, roughly 35% survival, thiphos at the concentrations used in this study have no effects on and c) 2000 and 3000 mg L 1 treatments (11.4 ± 0.94 and 0.28%, developmental time, but do increase mortality by approximately respectively). 20%. Exposure to 0.5 mgL 1 for 30 min daily for 7 days led to a ~50% 1024 P. Gebauer et al. / Chemosphere 185 (2017) 1019e1029

Fig. 3. A) Cumulative percent of dead and dying larvae 24 and 48 h following 30 min exposure to different concentrations of azamethiphos. B) Developmental time (days, bars) and survival of the first larval stage (zoea I) (triangle). Arrow indicates concentrate recommended by the manufacturer (CRM). Data represent means ± SE. mortality; the same percentage was recorded in the acute experi- M. edwardsii larvae died in less than 24 h post-exposure at a con- ments at 1 mgL 1 after 24 h post-exposure. centration 10 times lower (1.5 mgL 1) than the manufacturer’s The two pyrethroids, cypermethrin and deltamethrin, were recommendation (15 mgL 1), and 48 h post-exposure at a con- extremely toxic for M. edwardsii larvae, as has been described for centration 100 times lower (0.15 mgL 1). This chemical compound other crustaceans. In the case of cypermethrin, 100% of the was lethal and acted quickly against M. edwardsii larvae. Its high P. Gebauer et al. / Chemosphere 185 (2017) 1019e1029 1025

Fig. 4. A) Cumulative percent of dead and dying larvae 24, 48, 72, and 96 h following 20 min exposure to different concentrations of hydrogen peroxide. B) Developmental time (days; bars) and survival of the first larval stage (zoea I) (filled triangle). Arrow indicates concentration recommended by the manufacturer (CRM). Data represent mean ± SE.

toxicity has been reported in larvae of Homarus americanus (Burridge et al., 2014), with an estimated LC50 ranging from 0.0006 1 (Burridge et al., 2000a, 2000b; Pahl and Opitz, 1999), with a LC50 to 0.0017 mgL after 24 h exposure, and an estimated LC50 of varying between 0.06 and 0.66 mgL 1 depending on exposure time 0.0035 mgL 1 after one hour exposure. In the case of M. edwardsii 1 (48 h and 5 min, respectively) and larval stage. In addition, a LC50 of larvae, exposure to a concentration of 0.1 mgL for only 40 min 0.17 mgL 1 after 4 h exposure has been reported in larval stages of caused 100% mortality after 48 h of exposure. Although difficult to fresh-water copepods (Wendt-Rasch et al., 2003). The effect of compare the sensitivity of different species, due to differences in deltamethrin has been assessed in American lobster larvae their mass and exposure conditions, the available information 1026 P. Gebauer et al. / Chemosphere 185 (2017) 1019e1029

Fig. 5. A) Percentage of dead larvae after 30 min daily exposure to different concentrations of azamethiphos, at 24, 48, 72, and 96 h. B) Developmental time (days; bars) and survival of the first larval stage (zoea I) (filled triangle) after daily exposure to different concentrations of azamethiphos over 7 days. Data represent means ± SE. P. Gebauer et al. / Chemosphere 185 (2017) 1019e1029 1027

Fig. 6. A) Percentage of larvae affected (dead þ dying) after 20 min daily exposure to different concentrations of hydrogen peroxide, at time points 24, 48, 72, and 96 h. B) Developmental time (days; bars) and survival of the first larval stage (zoea I) (filled triangle) after daily exposure to different concentrations of hydrogen peroxide over a seven-day period. Data represent means ± SE. 1028 P. Gebauer et al. / Chemosphere 185 (2017) 1019e1029 indicates that concentrations below manufacturer recommenda- Fraser, 1995). In a study carried out in Scotland, the highest con- tions produce lethal effects in the larval stages of the crustacean centrations of cypermethrin (0.187 mgL 1) were measured 25 m species assessed. (downstream) from the treated net pen, 25 min after application 1 For azamethiphos, a similar LC50 (1e3 mgL ) was obtained for (Hunter and Fraser, 1995), Willis et al. (2005) estimate similar M. edwardsii zoea I as for larval stages I, II, III and IV of Homarus values for cypermethrin 25 min after final treatment (0.218, 0.197 americanus (Burridge et al., 1999); however the exposure period and 0.195 mgL 1). Furthermore, the cypermethrin concentration was much longer in the American lobster (48 h vs. 30 min). This remains above 0.074 mgL 1 up to 30 min after release (Hunter and indicates a greater sensitivity in M. edwadsii larvae, which may be Fraser, 1995), and cypermethrin levels around pens 4e5 h after due to differences in the larvae mass (dry weight, M. edwardsii: application were measured between 0.020 and 0.040 mgL 1 (Ernst 15 mg vs. H. americanus: 899 mg, Pandian, 1970). As with cyper- et al., 2001). Concentrations between 0.180 and 0.018 mgL 1 of methrin and deltamethrin, negative effects are found well below deltamethrin have been measured 1 and 100 m, respectively from the concentration recommended by the manufacturer (100 mgL 1). their application point (Ernst et al., 2014), and azamethiphos con- The effect of hydrogen peroxide on crustacean larval stages has centrations of approximately 25 and 1 mgL 1 were measured 1 and also been studied in H. americanus larvae; with the LC50 24 h post- 1000 m, respectively, from their application point. These data exposure for stage I larvae being calculated as 1637 mg L 1 with one indicate that cypermethrin concentrations, which this study hour’s exposure (Burridge et al., 2014). This value is very similar to demonstrated to be lethal to M. edwardsii larvae (0.15 mgL 1), can that for M. edwardsii zoea I (1642 mg L 1) with 20 min exposure. persist in the plume for at least 25e30 min. Deltamethrin con- 1 Although the LC50 found is higher than the recommended centrations 0.1 mgL may be present in the immediate sur- 1500 mg L 1 used to treat C. rogercresseyi, immediately following roundings of the net pen where the product has been applied, hydrogen peroxide exposure at this concentration, 100% of the indicating that the direct effect of these pyrethroids is restricted in M. edwardsii larvae were unable to swim or feed. If this condition application time and area. For azamethiphos, the distance of action were to occur in the field, the affected larvae would almost could extend up to 1000 m, presenting concentrations close to the 1 certainly die, because they would either precipitate, be carried estimated EC50 (0.94 ± 0.15 mgL ) at that distance. away by the currents, or become vulnerable to predators. The lethal The direct lethal effects of pyrethroids on crustacean larvae are effects reported in the present study may therefore underestimate probably restricted in time and space to the application point, but the effects of hydrogen peroxide on non-target species under nat- the action of the organophosphate azamethiphos may affect a ural conditions. larger area of the water column. However, there is insufficient in- In Chile, bathing with pyrethroids (cypermethrin and delta- formation to dimension the real effects of the repeated application methrin) and the organophosphate azamethiphos is coordinated of these compounds over time and their potential synergistic ef- through a National Fisheries and Aquaculture Service (SERNA- fects with other products. Further studies are needed on the effects PESCA) application timetable, normally consisting of roughly 8 days of these chemicals on the meroplankton, and on their potential of pyrethroids or azamethiphos application and then 8 days accumulation in sediments, in order to examine broader implica- without application. Under this regime, non-target organisms may tions for both pelagic and benthic communities. suffer prolonged exposure to chemical compounds. Chronic expo- In conclusion, chemical compounds used to treat C. rogercressegi sure of M. edwardsii larvae to azamethiphos concentrations ectoparasite have detrimental effects on the survival and devel- 0.5 mgL 1 provoked a 24% increase in mortality. Thus, parasite opmental time of M. edwardsii larvae, and potentially on many control may cause greater harm to non-target species than has been other non-target species. This is due to un-specificity and broad previously thought. Furthermore, several net pens may be treated action spectrum of the formulations used to treat sea lice. Further in a relatively small area, increasing the potential for repeated or research is needed on the pesticide impacts on non-target species chronic exposure. A large part of Chilean salmon-farming occurs in to assess the broader effects of these compounds on marine eco- fjords, which often contain more than one salmon farm. If the water systems and move towards a sustainable and competitive aqua- exchange rate is low, these zones could be exposed to both repeated culture industry. and chronic effects. Repeated exposure to azamethiphos for short periods causes additional negative effects in the American lobster Compliance with ethical standards (Burridge et al., 2000a), and the possibility of cumulative effects in non-target species resulting from different treatments cannot be The authors declare no conflict of interest. All applicable inter- discounted in these environments. national, national, and/or institutional guidelines for the care and Pyrethroids have low solubility in water (Burridge et al., 2010; use of were followed. Zhou et al., 1995) and adhere to particulate matter, making them less available in the water column and more likely to accumulate in Acknowledgements the sediment (Mayor et al., 2008). Although pyrethroids are highly toxic pesticides, their direct effects in the aqueous phase are This work was supported by FIPA 2014e65. We are grateful to probably manifested in a restricted time and space around the Ignacio Retamal for technical assistance with the rearing experi- treated net pens. Due to their high affinity with organic matter, ments. The authors would like to thank Dr.(c) Karen Middlemiss for pyrethroids would represent a greater risk for the benthic com- reviewing the language of this manuscript. munity and organisms that ingest particulate matter. On the other hand, azamethiphos has low absorption and low Kow, meaning that References it persists for longer time in the water column than pyrethroids (Ernst et al., 2014). These azamethiphos characteristics indicate that Aaen, S.M., Helgesen, K.O., Bakke, M.J., Kaur, K., Horsberg, T.E., 2015. Drug resistance e its aqueous toxicity presents a higher risk, and sediment accumu- in sea lice: a threat to salmonid aquaculture. Trends Parasitol. 31, 72 81. Abgrall, P., Rangeley, R.W., Burridge, L.E., Lawton, P., 2000. Sublethal effects of lation is a less important route for its disappearance from the water azamethiphos on shelter use by juvenile lobsters (Homarus americanus). column (Ernst et al., 2014). Aquaculture 181, 1e10. Rapid cypermethrin dispersion and dilution occur after release Barata, C., Medina, M., Telfer, T., Baird, D.J., 2002. Determining demographic effects of cypermethrin in the marine copepod Acartia tonsa: stage-specific short tests from salmon pens, and a single application of the product may versus life-table tests. Arch. Environ. Contam. Toxicol. 43, 373e378. 2 result in a plume of up to 1 km (Ernst et al., 2001; Hunter and Bretz, F., Hothorn, T., Westfall, P., 2016. Multiple Comparisons Using R. CRC Press. P. Gebauer et al. / Chemosphere 185 (2017) 1019e1029 1029

Burridge, L.E., Haya, K., Zitko, V., Waddy, S., 1999. The lethality of Salmosan (Aza- Chemistry: Water Pollution, Part M, vol. 5, pp. 305e341. methiphos) to American lobster (Homarus americanus) larvae, postlarvae, and Helgesen, K.O., Bravo, S., Sevatdal, S., Mendoza, J., Horsberg, T.E., 2014. Deltamethrin adults. Ecotoxicol. Environ. Saf. 43, 165e169. resistance in the sea louse Caligus rogercresseyi (Boxhall and Bravo) in Chile: Burridge, L.E., Haya, K., Page, F.H., Waddy, S.L., Zitko, V., Wade, J., 2000a. The bioassay results and usage data for antiparasitic agents with references to ® lethality of the cypermethrin formulation Excis to larval and post-larval stages Norwegian conditions. J. Fish. Dis. 37, 877e890. of the American lobster (Homarus americanus). Aquaculture 182, 37e47. Hunter, R.P., Fraser, N., 1995. Field Monitoring of the Effects of Cypermethrin as Burridge, L.E., Haya, K., Waddy, S.L., Wade, J., 2000b. The lethality of anti-sea lice GPRD01. Rep. Grampian Pharm. Ltd.. Res. Div. Ref GP95033. ® ® formulations Salmosan (Azamethiphos) and Excis (Cypermethrin) to stage IV Intorre, L., Soldani, G., Cognetti-Varriale, A.M., Monni, G., Meucci, V., Pretti, C., 2004. and adult lobsters (Homarus americanus) during repeated short-term expo- Safety of azamethiphos in eel, seabass and trout. Pharmacol. Res. 49, 171e176. sures. Aquaculture 182, 27e35. Mayor, D.J., Solan, M., Martinez, I., Murray, L., McMillan, H., Paton, G.I., Killham, K., Burridge, L., Weis, J.S., Cabello, F., Pizarro, J., Bostick, K., 2010. Chemical use in 2008. Acute toxicity of some treatments commonly used by the salmonid salmon aquaculture: a review of current practices and possible environmental aquaculture industry to Corophium volutator and Hediste diversicolor: whole effects. Aquaculture 306, 7e23. sediment bioassay tests. Aquaculture 285, 102e108. Burridge, L.E., Lyons, M.C., Wong, D.K.H., MacKeigan, K., VanGeest, J.L., 2014. The Miller, T.A., Adams, M.E., 1982. Mode of action of pyrethroids. Insectic. Mode Action ® ® acute lethality of three anti-sea lice formulations: AlphaMax , Salmosan , and 3e27. ® Interox ParamoveTM 50 to lobster and shrimp. Aquaculture 420, 180e186. Oliveira, C., Almeida, J., Guilhermino, L., Soares, A.M., Gravato, C., 2012. Acute effects Costello, M.J., 2009. The global economic cost of sea lice to the salmonid farming of deltamethrin on swimming velocity and biomarkers of the common prawn industry. J. Fish. Dis. 32, 115e118. Palaemon serratus. Aquat. Toxicol. 124, 209e216. Crane, M., Gross, M., Maycock, D.S., Grant, A., Fossum, B.H., 2011. Environmental Pahl, B.C., Opitz, H.M., 1999. The effects of cypermethrin (Excis) and azamethiphos quality standards for a deltamethrin sea louse treatment in marine finfish (Salmosan) on lobster Homarus americanus H. Milne Edwards larvae in a lab- aquaculture based on survival time analyses and species sensitivity distribu- oratory study. Aquac. Res. 30, 655e665. tions. Aquac. Res. 42, 68e72. Pandian, T.J., 1970. Yolk utilization and hatching time in the Canadian lobster Dorts, J., Silvestre, F., Tu, H.T., Tyberghein, A.-E., Phuong, N.T., Kestemont, P., 2009. Homarus americanus. Mar. Biol. 7, 249e254. Oxidative stress, protein carbonylation and heat shock proteins in the black Van Geest, J.L., Burridge, L.E., Fife, F.J., Kidd, K.A., 2014. Feeding response in marine tiger shrimp, Penaeus monodon, following exposure to endosulfan and delta- copepods as a measure of acute toxicity of four anti-sea lice pesticides. Mar. methrin. Environ. Toxicol. Pharmacol. 28, 302e310. Environ. Res. 101, 145e152. Ernst, W., Jackman, P., Doe, K., Page, F., Julien, G., Mackay, K., Sutherland, T., 2001. Wendt-Rasch, L., Friberg-Jensen, U., Woin, P., Christoffersen, K., 2003. Effects of the Dispersion and toxicity to non-target aquatic organisms of pesticides used to pyrethroid insecticide cypermethrin on a freshwater community studied under treat sea lice on salmon in net pen enclosures. Mar. Pollut. Bull. 42, 432e443. field conditions. II. Direct and indirect effects on the species composition. Ernst, W., Doe, K., Cook, A., Burridge, L., Lalonde, B., Jackman, P., Aube, J.G., Page, F., Aquat. Toxicol. 63, 373e389. 2014. Dispersion and toxicity to non-target crustaceans of azamethiphos and Willis, K.J., Gillibrand, P.A., Cromey, C.J., Black, K.D., 2005. Sea lice treatments on deltamethrin after sea lice treatments on farmed salmon, Salmo salar. Aqua- salmon farms have no adverse effects on zooplankton communities: a case culture 424e425, 104e112. study. Mar. Pollut. Bull. 50, 806e816. Haya, K., Burridge, L.E., Davies, I.M., Ervik, A., 2005. A review and assessment of Zhou, J.L., Rowland, S.J., Mantoura, R.F.C., Harland, B.J., 1995. Influence of the nature environmental risk of chemicals used for the treatment of sea lice infestations of particulate organic matter on the sorption of cypermethrin: implications on of cultured salmon. In: Hargrave, B. (Ed.), Handbook of Environmental KOC correlations. Environ. Int. 21, 187e195.