The effectiveness of electrified fencing using copper electrodes for slug (Airon spp.) control with direct electric current and voltage

Terrestrial slugs of the genus Arion are native to Europe, where they represent an important agricultural pest. This study determined the effectiveness of electrified fencing on barrier crossing by slugs. The fences were placed in glass insectariums. DC (direct current) voltages (2, 4, 6, 8 and 10 V) applied across the fence were tested together with limiting electric current values (0.1, 0.01 and 0.001 mA). Four categories of behavior were identified for slugs in the presence of an electrif ied fence: the animals (1) were stationary, (2) were moving but avoided the electrodes, (3) touched the electrodes and/or attempted unsuccessfully to cross the electrified fence and (4) succeeded in crossing the electrified fence. The effect of the applied voltages and the limiting current values on slug movement was highly significant. Forty-one percent of slugs crossed the fence at the lowest applied voltage, whereas only 1% of slugs succeeded in crossing at the highest voltage. The lowest limiting current values resulted in the most frequent fence crossings. For larger voltages and limiting currents, the frequency of slug crossings was effectively zero at fence regions positioned on the glass walls of the insectariums. In conclusion, this method of slug prevention may be highly effective, environmentally friendly and may result in deterrence, not death, of terrestrial slugs, a ubiquitous pest responsible for significant economic damage in agriculture. Additional key words: behavioral events; crossing; pests; repellent; terrestrial mollusks.


Introduction
Herbivorous terrestrial slugs and snails are destructive agricultural pests.They cause economic damage to a wide variety of plants, including vegetables, forage crops, tree fruits, shrubs, flowers, ground green cover, and newly sown lawngrasses (Port and Port, 1986;South, 1992;Faberi et al., 2006).Moreover, these mollusks play an important role in transmitting and spreading diseases to cultivated plants and in decreasing plant host resistance (Ohlendorf, 1999).Slugs of the genus Arion are an especially significant agricultural pest in Europe (Frank, 1998), North America (Hammond et al., 1999) and Asia (Ahmadi, 2004).Slugmediated damage is typically contained through the use of molluscicides because these chemicals are generally considered to represent the most effective measure for terrestrial mollusk control (Godin, 1983).
However, the widespread use of biocidal chemicals is rarely innocuous.Growers and farmers often experience difficulty controlling these pests with conventional bait pellets containing methiocarb and metaldehyde (Salvio et al., 2008).The efficacy of these pellets can be markedly degraded in wet conditions (Hata et al., 1997), resulting in unsatisfactory levels of pest control.Importantly, poison baits can be toxic to other nontarget soil invertebrates, (such as the beneficial carabid beetles), to birds and to mammals such as shrews and field mice (Hagin and Bobnick, 1991;Purvis, 1996).There is no antidote for metaldehyde; long-term skin exposure of levels > 50 mg kg -1 leads to dermatitis, whereas eye exposure can cause conjunctivitis (Purvis, 1996).Acetylcholinesterase inhibitor-based pellets are highly toxic and have caused the deaths of pets and humans.Carbamates are believed to be the most potent class of molluscicides (Miller et al., 1988;Radwan et al., 1992), but they can be toxic to warm-blooded organisms (Maddy et al., 1977).Clearly, there is a substantial impetus to develop selective and toxic molluskdeterrent compounds that have minimal effects on other species.
The development of effective alternatives to conventional molluscicides, particularly those that could be used in an integrated control strategy, would reduce plant losses, improve plant quality, and offer a sustainable strategy for controlling slug and snail pests with reduced molluscicide input.The development of alternative snail and slug control methods compatible with integrated pest management (IPM) strategies used to control other pests would help satisfy increasing market demands for ornamental plants and edible crops grown under environmentally responsible production methods (Schüder et al., 2003).
An alternative approach to poisoning snails/slugs chemically or biologically (i.e., the use of the slugparasitic nematode Phasmarhabditis hermaphrodita [Schneider]; Grimm, 2002) would prevent their access to target plants.Copper barriers are believed to create an electrical current when they react to snail or slug secretions (Schüder et al., 2003).
The fundamental biophysical parameters that determine the efficacy of electrical fencing as a deterrent to slug movement was investigated in present research.The aim of this work was to assess the interactions of terrestrial slugs (in particular Arion slugs) using electric current and voltage and to determine electrical parameters that are sufficient to block the passage of slugs past an electrical fence.The results of this work established recommended values that limit the amplitude of the electric current, thereby minimizing power consumption (electrical fences would mostly be battery powered) and potential harm to the operator while at the same time providing a highly effective slug deterrent.

Material and methods
Experiments were performed at the Laboratory of Entomology, Dept. of Agronomy, Biotechnical Faculty, University of Ljubljana.Four hundred five slugs, mainly representatives of Arion lusitanicus J. Mabille and Arion subfuscus (Draparnaud), were collected beneath the hedges, near the compost heap and near the brook at the laboratory field of the Biotechnical Faculty in Ljubljana (46°04' N, 14°31' E, 299 m a.s.l.).The slugs collected had an average body weight of 4.2 ± 2.3 g (n = 50) and were of various ages and lengths because we wanted to analyze a comprehensive sample of outdoor slug behaviors (Laznik et al., 2010).
An Agilent 3631A power source was used as a DC voltage or current source.The instrument as a voltage source and tested DC voltages of 2, 4, 6, 8 or 10 V was used.For each applied voltage, three limiting current values were tested: 0.1, 0.01 and 0.001 mA.Notably, it was not necessary for a limiting value of the set current to be reached when the slug touched or crossed the electrodes of the fence.This issue is explored in more detail in the Discussion section.
The slugs were placed in a glass insectarium (widthlength-depth 500-350-400 mm).At the bottom of the insectarium, three regions (area 200 mm × 140 mm) were delimited with an electrical fence.In the insectarium 3 slugs were placed, one slug in each region.The fence was constructed from electrodes made of copper adhesive tape (thickness 0.1 mm, width 10 mm) separated by 10 mm.The DC voltage source was applied to the electrodes.The behavioral response of the slugs was visually observed for approximately 10 minutes at each set of applied voltages and selected maximal currents.Each experiment was repeated three times in one day.In a subset of experimental paradigms, experiments were repeated the following day.In total, each experiment was repeated nine times.
The typical behavioral responses of the slugs during the experiment were classified in terms of four events, as described in Table 1.The numbers used to index the events were used to quantify the analysis.For example, the index value for event 2 is 2. To perform the data analysis, these index values were used as the values of the response variable "event."For ease of presentation, the term "the events" will also be used at times to refer to this response variable.For example, to calculate the average value of the events for a given combination of voltage and current, the index values were averaged for all of the slugs observed at that combination of voltage and current.
A two-way analysis of variance (ANOVA) was carried out to evaluate the differences in the response of Arion slugs to 15 different combinations of applied voltage and limiting electric current.The same (ANOVA) method was used to evaluate the time taken to cross the fence.Before analysis, each variable was tested for homogeneity of variance and the values transformed by arcsine square root if necessary.Duncan's multiple range test (p ≤ 0.05) was used to analyze the differences between individual treatment means (Hoshmand, 2006).All statistical analyses were performed using Statgraphics Plus for Windows 4.0 (Statistical Graphics Corp., Manugistics, Inc., Rockville, Maryland, USA).The data are presented as untransformed means (Trdan et al., 2006).

Individual analysis: tests of the differences between individual means
Results showed that higher current values produced less influence of voltage on the events (see Table 2).At the largest limiting current values (0.1 mA), the influence of voltage on the events was minimal.The statistical differences in the events at a current of 0.1 mA ranged from 2.63 ± 0.23 at the 2 V level to 2.56 ± 0.16 at the 8 V level.However, when the current was limited to 0.001 mA the set source voltages played a much more significant role.The most significant value of the events (3.37 ± 0.19) was observed at the lowest source voltage (2 V), whereas the lowest value of the events, 2.56 ± 0.17, was seen at the highest source voltage (10 V).Statistically significant differences between the events were also obtained for a limiting current value of 0.01 mA, where the average values ranged from 2.96 ± 0.12 at 6 V to 2.37 ± 0.16 at 8 V.
The effect of limiting current on the average crossing times was not statistically significant.A difference observed at voltages of 8 V and 10 V can be attributed to a single event (Table 2).
The analysis of the average values of the events and the average crossing times at different applied source voltages (Table 3 1 The slug is placed inside the fence but does not move. differences for average crossing times (p = 0.0488).The values ranged from 4.60 ± 1.25 at 8 V to 6.25 ± 0.77 at 6 V of applied voltage.The values at 10 V were disregarded owing to a single confounding event.Significant differences between the event means were also obtained.The means at the higher applied voltages (2.49 ± 0.10 at 8 V and 2.56 ± 0.09 at 10 V) differed significantly from the means at the lower applied voltages (2.97 ± 0.13 at 2 V, 2.90 ± 0.10 at 4 V, and 2.91 ± 0.08 at 6 V).At the higher applied voltages, the slugs often tended to move inside the fence without touching it (index value 2).At the lower applied voltages, the slugs almost always tended to touch the electrodes but did not cross them (index value 3).Analysis of the average value of the events and the average crossing time at different values of limiting electrical current (Table 3) revealed no signif icant differences in average crossing times (p = 0.4369), whereas signif icant differences were found for the average values of the events (p < 0.0000).At the two highest values of limiting current, the slugs often tended to move around the enclosure but avoided the electrodes (average event values of 2.54 ± 0.08 at 0.1 mA and 2.73 ± 0.07 at 0.01 mA), whereas at the lowest 6.33 ± 0.91 bA (n = 9) 6.00 ± 1.73 aA (n = 3) -8 4.00 ± 0.00 aA (n = 1) 7.00 ± 0.00 aB (n = 1) 4.00 ± 2.00 aA (n = 3) 10 9.00 ± 0.00 cA (n = 1) -p-value 0.0462 0.7578 0.9410 -: no slugs had crossed the copper adhesive tapes.

Discussion
Prior investigations have shown that copper barriers present an effective mechanical and physiochemical barrier for slugs because an electrochemical reaction between the copper and the snails' or slugs' secretions results in small electric currents (Sullivan and Cheng, 1976;Schüder et al., 2003).Our results show that both the source voltage and the amount of electric current that flows through the slug during the touching and/or crossing of the electrical fence affect the efficacy of the barrier.At the lowest limiting current (0.001 mA), 31% of the slugs crossed the fence, whereas only 10% of the slugs crossed the fence at the highest limiting current.The present results show that the amplitude of the source voltage plays a critical function in slug behavioral control.Only 1% crossing was detected at the highest voltage (10 V), whereas 43% of slugs crossed the fence at the lowest applied voltage (2 V).It should be noted that if the limiting current was set at a certain value, it did not necessarily have to be reached during the time that the slugs were crossing or touching the electrodes.
As shown in previous studies, mollusks take up copper through ingestion (Berger and Dallinger, 1989), and directly through the foot, at least in the form of copper compounds (Ryder and Bowen, 1977).We hypothesized that by taking up copper sulfate and/or other copper compounds through the foot on contact with the patina of the fence, the slug experienced the toxic sublethal effects of copper (Cu 2+ ) or its salts.The currentand voltage-dependence of slug behavior observed in our study suggests that electrolytically produced heavy metal cations (Cu 2+ ) and local changes in pH represent aversive stimuli that act as a behavioral deterrent in parallel with the effects of the electric field (the voltage drop across the animal) itself.Consistent with this hypothesis, exposure to Cu 2+ was shown to represent a sublethal stressor in pleurocercid snails (Paulson et al., 1983) and was toxic for bivalves and polychaetes as well as gastropods (Kidwai and Ahmed, 1999;Cheung et al., 2002).Smith et al. (2009) showed that metal chelation disrupts the crosslinking of macromolecule polymer gels in gastropod mollusks.These adhesive gels carry relatively high percentages of negatively charged amino acids (Pawlicki et al., 2004) that exhibit high-affinity EDTA-sensitive binding of divalent cations such as Ca 2+ and Mg 2+ (Smith et al., 2009) and should therefore be strongly affected by electrolytically produced protons and Cu 2+ .Accordingly, heavy metals such as Cu 2+ impact gastropod chemoreception, resulting in decreased feeding behavior, foot paralysis and death (Cuthbert et al., 1976).
As shown by Davies and Hawkins (1998), molluscan mucus contains mostly water.Water is therefore an important resource for the slug and has a strong influence on its behavior.In addition to the toxic effect of the copper patina itself, it may also be possible that copper chlorides (Rickett and Payer, 1995), urates (Bernardi et al., 2009) and other compounds present in the patina produce a dehydration of the mucus.Such dehydration would stress the animal and compel it to minimize contact with the dehydrating substrate.
Another study was conducted to determine whether polytetrafluoroethylene could be used to confine slugs Deroceras reticulatum (O.F.Müller) to an experimental plot.Polytetrafluoroethylene proved to be 100% effective in containing slugs within a restricted area over a 17-h period and was more convenient to apply and more effective than copper foil as a means of retaining slugs in plots (Symondson, 1993).Coto and Saunders (1987) evaluated four methods for keeping slugs inside experimental units.The methods consisted of four barriers [car grease, grease plus salt, glue (Resistol 5000) plus salt and an electric wire fence] placed around the units, which were planted with beans.Resistol 5000 plus salt gave good results in the laboratory but not in the field.The study concluded that none of the barriers could be recommended to retain slugs inside specific units and that if any physical barrier is to be used, overpopulation inside the units should be avoided.However, small units could be used if the experiment was performed in the field and with natural populations.
Chemoreception represents one of the most sensitive behavioral modalities of gastropods exposed to pollutants (Brown, 1976;Cuthbert et al., 1976).It remains to be seen whether exposure to an electric field and/or current affects the chemoreceptive sensors located in the tentacles and the foot, the orientation responses and the feeding behavior of gastropod slugs and snails.Moreover, the effects of the heavy metal that forms the fence, relative to the effect of the exposure to electricity, remain to be determined.Finally, we note that the efficacy of copper barriers might be reduced in the field due to moisture, rain and corrosion (Godin, 1983).The level of corrosion depends on chloride concentration, oxygen availability and acidity of the soil and possibly also on other factors (Afonso et al., 2009).
The impact of copper corrosion on the current flow, voltage drop across the barrier electrodes and effectiveness of the barrier as a slug repellent in the field remains to be determined in future studies.
Our investigation has clearly demonstrated the effectiveness of electrical fence barriers with DC voltages of 8-10 V and with electric currents between 0.01 and 0.1 mA on the behavior of slugs (Arion spp.), an important agricultural pest.The method can be regarded as an important alternative to limacides, substances that can be harmful to people (especially children) and the environment.From this point of view, we believe that electrical deterrence of slugs is ecologically advanced and that it is also ethical because it does not kill the slugs but only prevents them from damaging the crops.Because of the current high price of the copper tapes, this method is most likely to be used by gardeners and by some growers of high-value crops.

Table 1 .
Definitions of behavioral events occurring during the experiments

Table 2 .
Electrical barrier for Airon slugs 897 Average values of the events and average time (min) taken by slugs (Arion spp.) to cross copper adhesive tapes during exposure to different combinations of limiting current and source voltage.Capital and lower-case letters correspond to the grouping of means by Duncan's multiple range test (p < 0.05) for DC electric current-limiting value and voltage, respectively.Between rows, data with different letters differ statistically