An effective method for terrestrial arthropod euthanasia

A good death is an important part of a good life for all animals in captivity, and a lack of effective euthanasia methods for invertebrates has meant that the practice is not in tune with the theory. This gap in knowledge has gained significant attention recently (Cooper, 2011; Murray, 2012), as moral concern grows over just what constitutes suffering in these animals. Suitable euthanasia methods need to be not only effective and conducted in a way that causes minimal pain, but also relatively simple to perform, acceptable to the person conducting the procedure (American Veterinary Medical Association, 2007), as well as compatible with any research involved.

Battison et al. (Battison et al., 2000) published a new method of euthanasia for the American lobster (Homarus americanus) using injection of a saturated solution of potassium chloride (KCl). This method caused immediate immobilisation and death through circulatory arrest in approximately 1 min and therefore fits the criteria for a euthanasia procedure. Currently anaesthesia followed by immersion in fixative such as 70% ethanol has been cited as a preferred method (Pizzi, 2012), but it is inappropriate for microbiology (Cooper, 2011) or RNA extraction. Freezing is often suggested by private and professional keepers, but this compromises histological examination and is increasingly regarded to be inhumane (Pizzi, 2012) without prior anaesthesia to a depth at which recovery is impossible before death from freezing.

This study extrapolates the work performed by Battison et al. (Battison et al., 2000) to a range of terrestrial arthropod phyla. Our aim is to produce a universal method of arthropod euthanasia that is not only more effective than current methods, but is quicker, more economical and – importantly for transcriptomic and proteomic studies – preserves tissue quality.

Animal nervous systems contain ion channels permeable to potassium and sodium, which respectively control the flow of these ions across the cell membrane and subsequently the electrical potential. In the field of electrophysiology it is well known that neurones at a resting potential maintain a high intracellular potassium concentration and low intracellular sodium concentration relative to the extracellular environment. Addition of extracellular potassium ions (K⁺) in the form of KCl immediately depolarises the cell through neutralisation of the potassium concentration gradient across the cell membrane. This resulting change in potential (voltage) across the cell membrane causes opening of voltage-gated sodium channels. Thus, excess extracellular KCl causes exaggerated sodium influx that depolarises the cell and becomes toxic to it. The excess K⁺ remaining in the extracellular environment also prevents repolarisation (Takahashi et al., 1999). Therefore, application of excess K⁺ around the neurones of the thoracic ganglia in the form of KCl abolishes the neural input and results in circulatory collapse and then death.

Evolutionarily conserved across a wide range of arthropod orders, the arthropod ventral nerve cord – with some exceptions – is positioned on the ventral midline and exhibits a chain of serial ganglia. The anterior ganglia are the largest and have been referred to as the ‘invertebrate brain’ (Gullan and Cranston, 1994), but their main role is processing the large amount of sensory input from the head. Therefore, destruction of the brain in invertebrates principally affects the input of sensory information from the eyes and antenna but not from the rest of the body. Consequently, a method of euthanasia targeting the ganglia that control the vital functions would be a more appropriate way to successfully terminate arthropod life.

Humans are not the first species to use K⁺ ions as pharmacological tools; scorpions have evolved to utilise K⁺ ions. Parabuthus species possess potent venom peptides that act on many ion channels (including K⁺ channels). Yet they have also evolved a prevenom with a high concentration of K⁺ ions of approximately 80 mmol l^–1 (Inceoglu et al., 2003). A study on the effect of this prevenom on two insect species (Trichoplusia ni and Sarcophagia bullata) concluded it is an energy-efficient way of prey capture through paralysis (Inceoglu et al., 2003). We demonstrate that if K⁺ is delivered directly to the thoracic ganglia, its effects are rapid and avoid generating potential nociceptive action potentials.

We hereby propose the term ‘targeted hyperkalosis’ to describe the euthanasia of invertebrates through local injection of K⁺ to deliberately depolarise the thoracic ganglia and bring about rapid death. The direct derivation of targeted hyperkalosis is a local, high-potassium state that accurately describes the method.

All animals used in this study were euthanised to provide tissue for other research projects and, as such, the development of a suitable method of euthanasia was required. Although the species involved are not currently legally protected, they were maintained under the ethos of the UK Animals (Scientific Procedures) Act 1986 [A(SP)A 86].

All animals were acclimatised for 2 weeks to confirm health and nutritional status, and were weighed during total anaesthesia.

Prior to any dosing studies, six giant cockroach nymphs [Blaberus giganteus (Linnaeus 1758)] were anaesthetised in the manner described below and allowed to recover in fresh air to record a baseline recovery time.

Ten nymphs, with a mean ± s.d. mass of 3.25±1.03 g, were anaesthetised individually in a 670 ml anaesthesia chamber (Reed et al., 2011) that delivered carbon dioxide (CO₂) at 0.5 l min^–1 through a low range flow metre (Harvard Apparatus, Edenbridge, UK). Once anaesthetised, nymphs were abdominally injected with sterile Ringers solution at 100 μl g^–1 body mass, allowed to recover, and then monitored for signs of unusual behaviour and/or illness. The time taken for full induction (i.e. abolishment of the righting reflex) and any movement were recorded along with the recovery time. The time to full recovery was defined as time to the return of the righting reflex. This was important as it enabled us to establish when recovery was expected.

Rather than performing a full dose–response curve, three dose groups were selected to assess the initial efficacy of KCl as a method of euthanasia; this was to reduce the number of animals receiving a toxic but non-fatal dose. As a suitable euthanasia protocol needs to be effective for every animal every time, standard toxicological parameters such as LD₅₀ (lethal dose in 50% of population) and LD₉₀ (lethal dose in 90% of population) were not deemed necessary for this study. KCl was dissolved at 300 mg ml^–1 in sterile water, which avoids the need for an incubation step as conducted by Battison et al. (Battison et al., 2000) and eliminates the risk of precipitation if the solution is kept at cooler temperatures. The solution was sterile filtered through a 0.2 μm filter (Merck Millipore, Watford, Hertfordshire, UK) and aliquoted for storage in 1.5 ml centrifuge tubes (Fisher Scientific, Loughborough, Leicestershire, UK).

Blaberus giganteus were anaesthetised as previously described and injected with 300 mg ml^–1 KCl into the first leg sinus via the arthrodial membrane (Fig. 1) using a 25 G needle and 1 ml syringe (Fisher Scientific). Doses assessed were 10 μl g^–1 (N=5 animals, 2.72±1.26 g), 50 μl g^–1 (N=5 animals, 2.92±0.71 g) and 100 μl g^–1 (N=10 animals, 2.49±0.42 g). Double the number of animals was used in the higher dose group to confirm a 100% fatality rate.

In order to determine the effectiveness of KCl euthanasia in different terrestrial invertebrates, we tested the method in eight other invertebrate species from eight orders (Table 1).

Injections for Gryllus bimaculatus and Locusta migratoria were administered in the first leg sinus via the arthrodial membrane. In Hierodula membranacea, the injection was via the arthrodial membrane but in the second leg sinus.

Myriapoda species Scolopendra polymorpha and Narceus americanus were injected into the joint between the second and third segment along the ventral midline (Fig. 2). Phasmids were injected into the ventral midline at the junction between the first leg plate and adjacent ventral plate (Fig. 3).

Acanthoscurria cordubensis and Hadogenes troglodytes were selected based on the requirement for their venom glands. Acanthoscurria cordubensis were dosed centrally via the sternum (Fig. 4) whereas H. troglodytes were injected into the junction between the second leg coxa and the sternum, rostrally to the genital operculum (Fig. 5).

A rising concentration of CO₂ was effective at reversibly anaesthetising, and therefore immobilising, all species used during this study. Individuals from several insect species (particularly those from the orders Blattodea and Phasmidae) were observed to vomit during anaesthesia and occasional exaggerated limb movements were noted. Little other evidence of distress was observed.

Initial baseline data indicated that anaesthesia induction for B. giganteus took 1 min 28 s (±6 s) for full anaesthesia, with recovery taking 1 min 41 s (±13 s). After injection of 10% v/w of body mass of sterile Ringers solution whilst under anaesthesia, animals recovered in 1 min 46 s (±9 s). No unusual behaviour was noted in the 48 h observation time post injection and, subsequently, no further observations were made. The E. calcarata nymph given 100 μl g^–1 sterile Ringers solution into the ventral thoracic cavity recovered after 6 min and showed no detectable behavioural differences up to and beyond the 48 h observation time. This demonstrates that the large volume alone (10% v/w of body mass) does not cause any noticeable effect on survival or behaviour.

Battison et al. (Battison et al., 2000) used ultrasonography to confirm circulatory arrest in H. americanus; in this study the authors relied upon irreversible cessation of movement and sensation, which was defined as no movement or recovery during a 24 h period. Doses of 1% and 5% v/w KCl in B. giganteus were non-fatal; 5% v/w caused marked local paralysis of forelimbs and antennae, which persisted for over 1 min; 10% v/w of KCl caused instant paralysis and inward contraction of the limbs, as well as total abolishment of all nociceptive responses and non-recovery within 24 h. Therefore, 10% v/w was identified as the 100% effective dose for follow-up studies.

In G. bimaculatus, 10% v/w was immediately effective in all but one individual. However, this was a suspected inaccurate dosing because for such small species a 33 G needle is required to limit leakage and maintain accuracy. For L. migratoria, H. membranacea and E. calcarata, 10% v/w KCl euthanasia was effective in all animals. The only unusual observation made was the forceful autotomisation of a rear leg by one L. migratoria.

The Myriapoda unique body plan presented a challenge as nerve ganglia and heart tissue are duplicated throughout its length; this most likely brought about the results observed. The N. americanus responded to 10% v/w KCl with a wave of paralysis moving anterior to posterior with a maximum latency to death of 13 s. Over half of the study group (four animals) were deemed dead before removal of the needle. The passage of paralysis anterior to posterior was slower in S. polymorpha, but after a single dose of 10% v/w KCl none of the study group had died within 10 min. Immediate paralysis of the anterior segments was evident in all animals, while the posterior three segments remained active in all animals. A second dose of 10% v/w KCl was required and resulted in immediate cessation of all movement. As such, a 20% v/w KCl final dose was effective in all the S. polymorpha tested and was therefore proposed as the protocol for the Chilopoda class of Arthropoda.

The 10% v/w KCl dose in the H. troglodytes caused immediate paralysis and death in all animals signified by contraction of the limbs and ablation of responses to noxious stimuli (limb crush).

Initial studies on the Theraphosidae revealed that injection of 10% v/w into the central nerve ganglion was not possible due to volume limitations and a substantial back pressure was observed in the syringe. Although it was not possible at this time to record haemolymph pressure in the Theraphosidae prosoma – and the authors are unaware of any articles on the subject – reducing the dose to 0.5% v/w and leaving the syringe in place for 10 s resolved the issue. Thus 0.5% v/w can be injected into the central sternum as this is technically easier than reaching the ganglia through an arthrodial membrane.

Injection of 0.5% v/w administered centrally via the sternum was effective in ablating the nervous system and caused death in all animals. Injection of 0.5% v/w into the prosoma ganglia is terminal and non-recoverable in Theraphosidae; haemolymph can be collected via cardiac puncture but only for approximately 1 min as this appears to be when the circulation stops. During development of this protocol, we discovered that direct intra-cardiac delivery of 1% v/w KCl is also an effective euthanasia method for Theraphosidae spiders but it does not appear to work for Araneomorphae spiders (data not shown). Thus the proposed spider protocol is 0.5% v/w delivered directly to the prosoma ganglia.

Representatives of other insect orders can also be euthanised with 10% v/w KCl administered directly via injection to the thoracic ganglia through the sternal membrane between the forelegs in the ventral midline. Adult species tested were the lesser wax moth [Achroia grisella (Fabricius 1794); Lepidoptera], hover fly (unknown species; Diptera) and several beetles [Pachnoda marginata (Drury 1773), Smaragdesthes africana oertzeni (Kolbe 1895) and Dicronorrhina derbyana conradsi (Kolbe 1909); Coleoptera]. For euthanasia of the small insects, a 10 μl glass syringe (Hamilton Company, Reno, NV, USA) was used with a 25 G needle.

The data presented here are summarised in Table 2 as a list of proposed euthanasia protocols listed by insect order. For heavily armoured arthropods such as the crustaceans, the thoracic ganglia can be reached through the arthroidial membrane sinus (Battison et al., 2000). For terrestrial arthropods, the ventral midline should be used where possible.

The data presented here demonstrate that KCl dosing is a rapid and effective method of euthanasia in the arthropod species tested. With training, this method is easy to perform and causes immediate death, which allows rapid tissue collection for experimental transcriptomic and proteomic studies. An understanding of the species’ specific neural anatomy is crucial to performing this technique. As such, it is important that this is understood prior to working with such animals.

Before obtaining vertebrates for scientific study, it is vital that staff can identify potential suffering and are proficient in carrying out euthanasia. This is detailed in the A(SP)A 86 and we feel that this approach should be adopted for invertebrates as well. KCl is an ideal agent for euthanasia as it is cheap, effective, doesn’t require any special storage, has a very long shelf life and is safe to use. KCl is compatible with a wider range of pathological and research investigations than any of the current protocols (such as transcriptomics and proteomics), with the exception of primary neuronal culture due to the nature of action. However, there are still situations where KCl is impractical, such as for very small species (e.g. Drosophila sp.) or when culling large numbers of invertebrates, where tissue samples are not required. In these instances, anaesthesia followed by a confirmatory procedure such as immersion in fixative or rapid freezing should be performed, but only as long as the anaesthesia is deep enough so death occurs before recovery is possible.

 
• A(SP)A 86

  Animals (Scientific Procedures) Act 1986

 
• CO₂

  carbon dioxide

 
• KCl

  potassium chloride

 
• LD₅₀

  lethal dose in 50% of population

 
• LD₉₀

  lethal dose in 90% of population

 
• v/w

  volume by weight

Thanks to Dr Carol Trim for critical review.