Impacts of the invasive seaweed Asparagopsis armata exudate on rockpool invertebrates

The marine red algae Asparagopsis armata is an invasive species which competitive advantage arises from the production and release of large amounts of toxic compounds to the surrounding invaded area, reducing the abundance of native species. The main objective of this study was to evaluate the effects of this invasive seaweed on marine invertebrates by exposing the common prawn Palaemon elegans and the marine snail Gibbula umbilicalis to the exudate of this macroalgae. The seaweed was collected and placed in tanks, for 12 hours, in the dark in a 1:10 ratio. Afterwards the media containing its secondary metabolites was collected for further testing. Lethal and sublethal effects of A. armata were investigated. Biochemical biomarkers responses associated with energy metabolism (lactate dehydrogenase, LDH; electron transport system activity, ETS; content in lipids, proteins and carbohydrates) were analysed. The biomarker responses showed invertebrates’ physiological status impairment after exposure to low concentrations of this algae exudate. Highest concentrations of exudate significantly increased lipid content in both organisms. In the shrimp, protein content, ETS, and LDH were also significantly increased. On the contrary, these parameters were significantly decreased in G. umbilicalis. A behavioural impairment was also observed in G. umbilicalis exposed to A. armata exudate, with reduction in feeding consumption. These results represent an important step in the research of natural toxic exudates released to the environment and prospective effects of this seaweed in invaded communities under increasing global change scenarios. These results indicate that organisms have different energy requirements to deal with the stress caused by the macroalgae exudate. This comparative analysis may provide important insights into the heterogeneous effects of the A. armata exudate, driven by species-specific metabolic susceptibility patterns.


Introduction
Overall rapid globalization, and increasing trends of trade and travel, have been accelerating marine biological invasions, by transporting species to areas outside their native range. These non-indigenous species (NIS) may then be considered invasive, once they establish, spread rapidly, and proliferate and dominate the new habitat without the direct support of humans (Richardson et al., 2011). Asparagopsis armata Harvey, 1855 (Bonnemaisoniales, Rhodophyta) is a red seaweed, native to Western Australia, and nowadays distributed throughout Europe in the Atlantic and Mediterranean basin, where it is highly invasive (Otero et al., 2013). This seaweed possesses chemical defence mechanisms that are nuclear to their invasiveness, based on the synthesis and storage of an array of secondary metabolites, which include over 100 halogenated compounds such as haloforms, haloacids, and haloketones (O. McConnell & Fenical, 1977). These halogenated volatile hydrocarbons containing one to four carbons are known antifeedant and cytotoxic compounds, among others (reviewed in Pinteus et al., 2018), and the pungent aroma of these algae is attributed to an essential oil that is composed mainly of bromoform with smaller amounts of other bromine, chlorine, and iodine-containing methane, ethane, ethanol, acetaldehydes, acetones, 2-acetoxypropanes, propenes, epoxypropanes, acroleins, and butenones, stored in vacuoles within gland cells (Burreson et al., 1976). These compounds potent biological effects can induce significant changes in terms of native community composition (Paul et al., 2006a) favoring A. armata in a given niche. Due to the enclosed environment during low-tide, rocky pools are ought to be sensitive sites to the increase of released compounds by retained A. armata, which may ultimately present adverse effects for other organisms such as seaweed, vertebrates or invertebrates, leading to severe consequences for coastal ecosystems. To investigate potential ecological impairments caused by these compounds, key role species in the structure and functioning of costal ecosystems were used: marine invertebrates such as the common prawn Palaemon elegans and the gastropod Gibbula umbilicalis, which inhabit the upper intertidal zone on rocky shores where A. armata is often found attached to the substrate, or unattached (drifting), and therefore releasing its chemical exudates. In this work, the assessment of A. armata exudate effects over these marine invertebrates was performed addressing survival and sublethal effects through behavioural and biochemical responses.
Several behavioural parameters have been chosen as indicators of invertebrate health status. Feeding activity have been shown to be sensitive tools to assess the impact of contaminants at concentrations far below lethal levels, which often is related to a decrease of fitness and capacity to respond to contaminants and also often reflects alterations in movement capability, and that will reduce energetic input and metabolism (Cabecinhas et  Dehydrogenase (LDH) and Electron Transport System (ETS) activities may provide valuable information on the physiological status of the studied organisms. Therefore, the purpose of this research was to address the effects of A. armata exudated secondary metabolites on the survival, behaviour, and energetic metabolism of two marine invertebrates inhabiting rock pools, giving further ecotoxicological insight on this seaweed invasive strategy.

Test organisms
The common prawn Palaemon elegans and the sea snail Gibbula umbilicalis were collected from Carreiro de Joannes, a rocky beach in Peniche, central Portugal (39°21'18.0"N, 9°23'40.6"W), with no known sources of chemical contamination. The organisms were maintained during 7 d in the laboratory in natural seawater at 20 ± 1 °C, with a 16 h:8 h (light:dark) photoperiod in aerated aquaria. Shrimps were fed with mussel and snails fed with Ulva lactuca until used in experiments. Prior to testing, organisms were kept fasting for 24 h.

Asparagopsis armata collection and preparation of exudates
Asparagopsis armata was collected in Berlenga Island, Peniche, Portugal (39°25'03.0"N, 9°30'23.6"W), a marine protected area, by SCUBA. In the lab, after cleaned and sorted, four aquaria with 5kg of A. armata and 50L of natural filtered seawater (through 0.45 µm cellulose acetate membrane filters) were left in the dark at 20 °C. After 12h, the algae was removed, the water from the different aquaria was pooled and sieved for bigger particles, followed by a filtration through a 0.45 µm cellulose acetate membrane filter. This exudate was then kept in PET bottles at -20ºC until further use. This, due to the obvious composition constrains, constitutes the stock solution for all experiments and the 100% concentration.

Exposure setup
All experiments were conducted in a climatic room at 20 ± 1 °C, with a 16 h:8 h (light:dark), and experimental replicates consisted of glass vials with 60 mL and 750 mL exudate solution (or seawater in controls) for snails and shrimps, respectively, with one organism each.
Flasks were covered with a plastic mesh to prevent organism to escape and to assure constant submersion. Exudate solutions were renewed every 24h to avoid excreta accumulation and possible loss of volatile compounds. Exudate concentrations are presented as % of the exudate produced as described in 2.2.

Survival
After a range finding test, sea snails were exposed to increasing concentrations ranging from 1 to 15 % of exudate (1; 1.57; 2.47; 3.87; 6.08; 9.55; and 15%), and shrimp from 4 to 10% of exudate (4; 4.66; 5.43; 6.32; 7.37; 8.58; and 10%). Exposures lasted 96 h and mortality was recorded daily. During exposures, no food was added. Eight and five replicates per treatment were used for sea snails and shrimps respectively, including a control treatment with filtered seawater only.

Sublethal exposure for biomarker analysis
Information on the lethal effects was used to establish maximum concentrations and conditions for each independent sublethal test, using half the LC10 as the highest concentrations tested. Sea snails were exposed to increasing concentrations of exudate ranging from 0.04 to 0.87 % (0; 0.04; 0.07; 0.14; 0.25; 0.47; and 0.87%), and shrimp were exposed from 0.11 to 2.46 % of exudate (0; 0.11; 0.21; 0.39; 0.72; 1.33; and 2.46%).
Exposures lasted 168 h and 16 replicates per treatment were used for snails and 8 for shrimps, including a control treatment with filtered seawater only. At the end of the exposure period, snail's shell was broken with a vise, and soft tissues removed with forceps, weighed and kept on ice for operculum removal. Shrimps were sacrificed by decapitation and dissected. The abdominal muscle was rapidly isolated on ice and weighed. Tissues were maintained at -80 °C until further analysis.

Feeding activity
For the feeding activity assay, concentrations of the exudate were the same as used for the biomarkers exposure, using 8 replicates per treatment for both invertebrates and exposed for 96h. Organisms were fed with discs of Ulva lactuca with c.a. 10 cm 2 previously dried at 60°C for 48 h, weighed and re-hydrated just before adding to the medium (one disc per replicate). At the end of the 96 h exudate-exposed feeding test, the discs were rinsed in clean water, dried again, weighed, and the feeding was assessed by subtracting the algae final weight to its initial dry mass (mg).

Tissue preparation
Snails were processed as pools of two individually exposed organisms, with each pool being considered as one biological replicate for the biomarker analysis (N=8). For shrimps, the muscle tissue of each organism was processed individually and considered as one biological replicate (N=8). The replicate tissues of each invertebrate species were homogenized in potassium phosphate buffer (0.1 M, pH 7.4) at a proportion of 1:12 for G. umbilicalis and 1:10 (m:v) for P. elegans. The homogenate was then separated into different microtubes for the analysis of total protein, carbohydrate, and lipid content. The remaining homogenate was separated into two fractions centrifuged respectively at 1000 g for 5min (4 ⁰C) for ETS measurement and at 3000g for 5 min (4 °C) for LDH measurement. All aliquots were stored at -80 °C until further analysis.

Energy reserves
Carbohydrate, lipid, and total protein contents were measured according to the approaches outlined by De Coen and Janssen (1997,2003). The total carbohydrate content was determined in a reaction with phenol 5% and H2SO4 (95-97%), using glucose as standard and measuring absorbance at 490 nm (De Coen & Janssen, 1997). Lipid content was determined according to Bligh and Dyer (1959), using tripalmitin as standard and measuring absorbance at 400 nm. The total protein content was determined using the Bradford method (1976), with bovine serum albumin as standard, measuring absorbance at 600 nm. Following De Coen and Janssen (1997,2003), all energy reserves were transformed into their energetic equivalents (39.5 kJ g −1 lipid, 24 kJ g −1 protein, 17.5 kJ g −1 glycogen).

Energy metabolism related enzymes
Electron transport system (ETS) activity was determined following the method described by De Coen & Janssen (1997). The ETS activity was measured spectrophotometrically by adding NADPH solution and INT (p iodo-nitro-tetrazolium) to the sample and absorbance was read at 490 for 3 minutes. The oxygen consumption was then calculated using the stoichiometric relationship: 2 µmol of formazan formed = 1 µmol of oxygen consumed. The oxygen consumption rate was then converted into the energetic equivalent of 484kJ/mol O2 for average carbohydrate, lipid, and protein consumption combinations (Gnaiger, 1983).
The activity of LDH was measured following Vassault (1983) with adaptations of Diamantino et al. (2001). The process is based on the efficiency of LDH to convert pyruvate to lactate, in the presence of NADH, which results in NADH oxidation and consequent decrease in absorbance. The absorbance was read at 340 nm for 5 min. A molar extinction coefficient of 6.3 × 10 3 M cm −1 was used, and results were expressed as nmol min −1 mg protein −1 .

Statistical analysis
Significant differences between each treatment for biomarker analyses and behavioural parameters were studied using one-way analysis of variance (ANOVA) and differences to control were addressed by Dunnett's post hoc test. Normality was checked by Shapiro-Wilk test and homoscedasticity by Levene test. In case of non-normally distributed data, the Kruskal-Wallis test was applied followed by Dunn's post hoc test. Statistical analyses for biochemical and behaviour endpoints were performed with the software SigmaPlot (Systat Software, San Jose, CA) and LCs and the correspondent 95% confidence intervals and global fitting were done on GraphPad Prism version 7 for Mac (GraphPad software, San Diego, CA).

Energy metabolism related biomarkers
Regarding G. umbilicalis exposure to A. armata exudate, no significant differences were observed in the carbohydrate and protein contents (Fig. 4a,c)

Discussion and conclusions
There are some studies addressing the toxicity of seaweeds and its detrimental effects to invertebrates, but very few address toxicity of seaweed secondary metabolites, such as in the study of the effects of Ulva sp. exudate on the gastropods Littorina littorea and L.
obtusata (Peckol & Putnam, 2017 elegans. The reasons for such low A. armata exudate tolerance in both invertebrates are of some concern due to the well documented importance of these species to the functioning of rocky shore communities as principal microalgal consumers of seagrass biofilm (Orth & Van Montfrans 1984).
The comparison between 96 h dose-response curves revealed that P. elegans were more tolerant than G. umbilicalis with significantly higher LC50. In fact, the high sensitivity of G.
umbilicalis to this exudate is less obvious, given the well-documented tolerance of this species to extreme environmental conditions (Southward, 1958). Additionally, G.
umbilicalis have been found to be relatively tolerant to other contaminants In exposed Palaemon elegans there was an increase in lipids and proteins along with the increase of ETS and LDH activities. The increase in lipids was similar as discussed previously for the sea snails, but for P. elegans, the exudate also exerted a significant increase of protein content. This increase may also reflect an induction in protein synthesis for detoxification processes and other defence mechanisms (Smolders et al., 2003). This is also The present work was performed using exudates which, to obtain naturally, demand that the seaweed is placed in seawater at a given ratio and defined conditions, which will mostly represent what is exudated by A. armata in the nature. Ratios of 1:10 and over are commonly found in rock pools, which may remain enclosed from minutes to several hours during a tidal period. Other factors which may induce additional stress and compound release may influence more toxicity such as the case of temperature or hydrodynamics (Gschwend et al., 1985). Also, the exudates were prepared in the dark, and the production of volatile halogenated organic compounds (VHOCs) production rate tends to decrease under dark conditions (Bondu et al., 2008). Despite the difficulties to benchmark this stressor preparation and impacts with a real case scenario, the very high toxicity of this seaweed might not even reflect the worst case scenario of exposure to exudates from this macroalgae, especially in bloom events, in summer, in tide pools, where a body of water separates from the sea for hours during a tidal cycle, where seaweed concentrations are high and water dilution is little.
The present study is an important step in the research of natural toxic exudates released into the environment and the mechanisms on how they can affect the surrounding organisms and their mode of action in the invaded ecosystems.
To the present knowledge, this represents the first work to study the exudate per se and its impacts in costal invertebrates. Although, as stated, this exudate contains a myriad of compounds, its toxicity is attributed mainly to the halogenated compounds (secondary metabolites) produced and stored in vacuoles within A. armata's gland cells (Burreson et al., 1976). Additionally to the results and toxic effects seen here at high dilutions of the prepared exudate, to better understand the real impact on coastal environments, and specially in more exposed tidal pools, further studies should be made to understand the concentrations found and its variation, considering seaweed density and other biotic and abiotic factors to better address this seaweed chemical defences true impact in coastal environments. These impacts may probably be more extent compared to the ones here found due to the referred increased stress and conditions that may lead to an increased production of secondary metabolites and also densities that can be much higher, especially when considering seasonal seaweed stranding, and all other factors affecting these organisms in a high stress burden ecosystem.

Acknowledgements
This study had the support of Fundação para a Ciência e a Tecnologia (FCT) through the