Toxicity of Neem’s oil, a Potential Biocide against the Invasive Mussel Limnoperna fortunei (Dunker 1857)

The golden mussel Limnoperna fortunei (Dunker 1857) is one of the most distributed Nuisance Invasive Species (NIS) in South America, and a threat of great concern for the industry of the area. In this study, we carried out toxicity tests made with a Neem’s oil solution with L. fortunei larvae and benthonic adults (7, 13 and 19 ± 1 mm). Tests with non-target species ( Daphnia magna , Lactuca sativa and Cnesterodon decemmculatus ) were also made with the aim to evaluate the potential toxicity of the Neem’s solution in the environment. The LC 100 of Neem’s solution obtained for larvae was 500 µl/L, a value much higher than the one obtained for D. magna and C. decemmaculatus . Thus, we recommend that it should not be used in open waters. However, since the adults were killed in 72 h and the larvae in 24 h, this product can be used in closed systems, in man-made facilities.


INTRODUCTION
Nuisance Invasive Species (NIS) are a major worldwide environmental problem (Mack et al. 2000, McGeoch et al. 2010, chiefly in freshwater systems (Sala et al. 2000, Strayer 2010. Bivalve mollusks are considered an aggressive NIS, both for their ability to affect the structure of the environment and the availability of resources in invaded ecosystems (i.e. ecosystems engineers, sensu Strayer et al. 1999), and their ability to infest man-made structures (Darrigran 2002, Darrigran andDamborenea 2005).
The golden mussel Limnoperna fortunei (Bivalvia: Mytilidae) is one of the NIS most distributed in South America (Boltovskoy et al. 2006) and since its accidental introduction in 1991 (Pastorino et al. 1993) it has invaded streams and rivers in the Paraná basin, in at least four countries: Argentina, Brazil, Paraguay and Uruguay (Brugnoli et al. 2005, Darrigran 2002, Darrigran and Damborenea 2011, de Oliveira et al. 2006. Limnoperna fortunei infests all types of man-made facilities: treatment plants, irrigation channels, reservoirs, hydroelectric power plants and fisheries (Darrigran 2010, Pestana et al. 2008, Rolla and Mota 2010. The most common strategy to fight this kind of trouble is the use of biocides (Mackie and Claudi 2010). The main advantage offered by this strategy is that it can be engineered to protect the entire facility, from the intake to the discharge, without altering the operability of the facility (Mackie and Claudi 2010).
Chlorine is the most used biocide, mainly because of its low cost, although its deleterious effects on the biota are well known (Mackie and Claudi 2010). Others treatments (e.g. ozone, sodium hydroxide) are starting to be used in manmade facilities in South America (Mackie and Claudi 2010). Now there is an increasing pressure to avoid the use of toxic substances against nontarget organisms (e.g. IMO 2007), and the current tendency is to find environmentally friendly biocides (Qian et al. 2010, Yebra et al. 2004).
Indian's Neem tree (Azadirachta indica A. Juss) is one of the most studied trees in the world (Girish and Shankara Bhat 2008), due to its enormous potential as a source of pesticides, insecticides and organic agrochemicals (Brahmachari 2004). The oil from its seeds is used as fungicide and insecticide (Girish and Shankara Bhat 2008), and its medical use has been recently recommended by the World Health Organization (WHO 2007). Its principal active component is Azadichtin (Figure 1) (Schaaf et al. 2000). Azadichtin has low toxicity against non-target organisms and low persistence in the environment (Schaaf et al. 2000), both of which are desirable characteristics for a biocide.
Considering the problems caused by an infestation of L. fortunei in a man-made structure, and the potential of Neem's extracts, static bioassays have been proposed. With the aim to evaluate the acute toxicity of Neem's oil on L. fortunei, tests with three sizes of adults and the planktonic larvae were made. Also, tests with three non-target organisms, the cladocera Daphnia magna, the composite Lactuca sativa and the fish Cnesterodon decemmaculatus, were performed to evaluate the potential impact of Neem's oil in freshwater environments.

METHODS
Neem's solution consisted of a mix of 60% pure Neem's oil (Table I) and 40% Fatty Ethoxylated Alcohol of 9 mols as emulsifier (Table II). A generator was used to prepare the solution (Integración Química S.R.L. pers. comm.).
All living organisms of L. fortunei were captured in the Argentinean coasts of the Río de la Plata (34°49´S -57°56´W).
Adult mussels were collected manually and immediately transported to the GIMIP's laboratory (UNLP). Mussels were acclimated for 15 days in 10-L aquaria (conductivity = 1 ± 0.3 ms/cm, temperature = 23 ± 2 °C, and pH = 7 ± 0.5). The water was completely replaced the first three days and then partially replaced (5 L) every two days until the end of the 15-day acclimation period. The mussels were fed with fish food (TetraMin ® ) daily, except for the last two days (Pereyra et al. 2011).
Larvae were collected by filtering 1000 L of water throughout a 45-µm mesh plankton net. The samples were immediately transported to the lab, where the existence of enough larvae for each assay was checked.
Seeds of L. sativa (mantecosa variety) were obtained from commercial suppliers, with 98% of guaranteed germination.
Individuals of C. decemmaculatus were caught in streams of the zone, in fields with low or null utilization of agrochemicals. These individuals were transported to the laboratory (CIMA -UNLP) and held in 1200 L plastic pools (Pelopincho ® ), where they were daily fed with TetraMin ® fish food. For the toxicity test, juveniles born in the laboratory were used.
Tests with adults were performed choosing individuals by size with a digital caliber (0.01 mm precision) 24 h before each assay. Twelve mussels were placed on plastic Petri dishes and held in plastic containers with 500 ml of dechlorinated tap water and artificial aeration. After 24 h, the plastic Petri dishes with the mussels attached by byssal threads were transferred to other plastic containers with 500 ml of Neem's solution. Mussels that had not attached to the Petri dish after the first 24 h were considered unapt for the assay and thus discarded (Pereyra et al. 2011). Mortality was checked after 72 h. Failure to respond to external tactile stimuli was used as the death criterion.
The assays with larvae were made with 12 ml of Neem's solution and adding 10 to 15 larvae in every plastic Petri dish.  International standardized methods (USEPA 1996a, b, c) for all non-target species were used in these tests. All tests were made with at least five concentrations and one control.
For the tests with D. magna, individuals were exposed to the toxic for 48 h with dechlorinated tap water for solutions (conductivity = 1.05 ms/ cm; hardness = 215 mg/L de CO 3 Ca; alkalinity 180 mg/L de CO 3 Ca, pH range 7.5 ± 0.2). The tests were carried under controlled conditions (temperature 20 ± 2 °C; photoperiod = 16:8 light: darkness), and without feeding the neonates. Ten neonates were placed in every test tube with 10 ml of solution each. Mortality was checked 48 hours later. A neonate was considered dead when it remained at the bottom of the test tube and showed no signs of activity.
For the tests with L. sativa, plants were exposed to the toxic for 120 h in darkness and controlled temperature (22 ± 2 °C). Double distilled water was used as dilution water. Twenty seeds were placed into plastic Petri dishes containing a sterilized filter paper with 3 ml of solution to be tested for every dish. Plastic Petri dishes were placed in darkness for 120 h, and germination and elongation of the roots were evaluated at the end of the test.
For the tests with C. desemmaculatus, individuals were exposed to the toxic for 96 h (USEPA 1996c). Five juveniles were held in plastic containers with 500 ml of solution each. Juveniles were kept without aeration or feeding. Solutions were replaced daily until completing 96 h of exposure to the toxic. A juvenile was considered dead when it remained at the bottom of the plastic container showing no signs of activity.
For L. fortunei, the LC 50 were calculated using linear regression, previous transformation of the concentration with logarithm and the mortality to probit units (Finney 1978), using USEPA program Probit 1.5. The lowest concentration that effectively caused 100% mortality in exposed organisms was considered as LC 100 .
For D. magna and C. desemmaculatus, all the estimations of the LC XX were made with the probit program, considering the LC 1 as NOEC (the highest concentration of the product at which no significant differences as regards the control were seen), and the LC 10 as LOEC (the lowest concentration of the product at which a significant difference as regards the control was seen).
For L. sativa, the IC 50 (Inhibition concentration 50, i.e., the concentration at which 50% of inhibition as regards the control is observed) was obtained using linear regression with the logarithmic transformation of the concentrations, using the inhibition proportion concerning the control as the dependent variable. The NOEC and LOEC were calculated with a posteriori comparisons of Dunnet, previous ANOVA. Table III shows the results obtained for L. fortunei, Table IV shows those for D. magna and C. decemmaculatus, and Table V shows those for L. sativa. The LC 100 of L. fortunei larvae was 500 µl/L. The LC 100 of adults could not be estimated because the concentrations tested did not achieve 100% mortality.

RESULTS and DISCUSSION
This results show i) Neem's oil has acute toxicity against L. fortunei, and ii) the larvae are more vulnerable to the toxic than the adults. This agrees with what was previously reported about the vulnerability of the initial stages of development (Maroñas andDamborenea 2006, Van Benschoten et al. 1995   Working with tannins of Schinopsis balansae (Englers.), we have previously obtained the LC 50 of L. fortunei, for adult mussels of 13 and 19 ± 1mm, after 168 h of exposure to the toxic (Pereyra et al. 2011). In this study, although 100% mortality was not achieved (the maximum mortality reached was 80%, data not shown), a considerable reduction in the exposure time to the toxic was obtained.
Results show no overlap between the LC 50 of L. fortunei (both stages) and L. sativa (Tables III and  V), indicating that at these concentrations Neem's oil has no effect on L. sativa. There is overlapping between the LC 50 of L. fortunei and those achieved with D. magna and C. desemmaculatus (Tables III  and IV). In view of these results, we recommend not to use this product in open waters. Likewise, this product is recommended to be used in closed systems in man-made facilities.
It is important to note that the use of biocides to control macrofouling is one of the treatments to apply. The treatment to be used depends on the possibility of each section of the facilitie, the time of the year and the stages to control (Costa et al. 2008, Darrigran et al. 2007). Other techniques (e.g. ozone) are used to prevent macrofouling of bivalves, but are difficult to apply and expensive (Mackie and Claudi 2010). Using products like Neem's oil is a valid alternative in closed systems.   Another emerging perspective is that combined techniques (e.g. Costa et al. 2011) can be applied to improve the control measures in man-made facilities. Finally, the bioassays with biocides with L. fortunei carried out to date have been made in static conditions (e.g. Cataldo et al. 2003, Darrigran and Damborenea 2001, Pereyra et al. 2011). These kinds of assays have the advantage of their practicality and low impact, but the results should be tested in the industry. For example, Rolla and Mota (2010) have reported that the reduction in the spread of L. fortunei in the Paraíba River resulted, in part, from the elimination of the golden mussel from the transitory harbors with the use of chlorine. However, there are no reports on the effects of chlorine over the biota (Rolla and Mota 2010). This lack of communication between scientists and the people involved (e.g. stakeholders, the company, the community) has been observed in other areas of NIS control (e.g. Donlan et al. 2003, Gardener et al. 2010, Shanley and López 2009, and represents another problem to be solved in order to achieve effective control programs of the golden mussel.