Invertebrate Sentience: Urgent But Understudied

There is not, and may never be, an unambiguous and universally-accepted metric for whether an organism can suffer or not. Nonetheless, our perceptions of animal behavior and cognition are related to our decisions about using animals (Bilewicz, Imhoff, & Drogosz, 2011), and we are likely to describe animals we plan to use as less sentient, independent of actual differences (Bastian, Loughnan, Haslam, & Radke, 2012; Loughnan, Haslam, & Bastian, 2010). An accurate understanding of animal sentience and their ability to suffer, especially for extremely common animals like invertebrates, needs to be rooted in an accurate understanding of their behavior. However, our overall understanding of invertebrate cognitive capacity is severely limited from lack of data and reliable studies.

Why invertebrates?

Every year, humans raise or harvest at least trillions of shrimp, crab, krill, and insects from silkworms to crickets to bees (“FAOSTAT,” n.d.; Gilette, 2008; Tomasik, 2017).1Tomasik estimates almost 1 trillion silkworms used by humans. The global shrimp harvest (farmed and wild-caught) in 2008 was 6 million tons. Assuming about 45 shrimp a pound, that’s 500 billion shrimp. Humans also own a total of about 80 million beehives, and with about 40,000 bees in a hive, we get about 3 trillion bees directly managed by humans.

If we expand our moral circle to include populations directly shaped by humans, we get a vast number – including animals in cropland and fields, bugs killed by pesticides and cars, bugs used for pet food, and the invertebrates living in our houses, yards, and cities, and more.

Expanding our circle a step wider, wild invertebrates make up 99% of the animal species on Earth (Borrell, 2012). They are, by biomass and number, the vast majority of animal life. Most of these invertebrates are small animals such as ocean nematodes, zooplankton, or insects (Ray, 2017a). If these animals have even a fraction of the moral consideration of larger animals, we should be paying significant attention to their wellbeing.

Currently, people assign very low importance to harm suffered by invertebrates.  We are, in general, willing to do things directly to invertebrates that we would be very uncomfortable doing to large vertebrates – smashing them with books when they appear in our homes, boiling them alive, spearing them on fishhooks for other predators to eat, and more. On an institutional level, the National Institute of Health, a major grant-making entity in the United States, has extensive regulations for the treatment of vertebrate laboratory subjects, but none for invertebrates (“Office of Laboratory Animal Welfare |,” n.d.). On an even larger scale, many breeding shrimp on shrimp farms have their eyes removed by hand without anaesthesia (Gilette, 2008; Waycott, 2017).

There has been a very recent push by some scientific entities for an increased acknowledgement of octopus welfare, in light of their great cognitive abilities (Harmon, n.d.). Otherwise, most people correctly believe that most invertebrates have fewer cognitive capacities and are less similar to humans than vertebrates. That said, while we may never feel the need to give insects the same degree of humane treatment that we are inclined to give to chimpanzees or dogs, they may still feel enough pain that we should avoid allowing them to die in painful ways, or lead lives of misery.

How do we study sentience?

Different people and philosophies define what suffering is, or what kinds of suffering are meaningful or bad, in different ways. Researchers, however, have converged on a few experimentally verifiable behavioral attributes that seem to capture relevant aspects of possible sentience (Broom, 2007; Elwood, Barr, & Patterson, 2009). These include:

  • The degree to which an animal’s nervous and nociceptive systems are similar to a human’s (including identified nociceptors and brain structures, and the presence of natural opioids that moderate responses to damage)
  • Behavioral responses to damaging stimuli (across the whole body and/or at the specific body part)
  • The capacity to learn from negative stimuli
  • Having a wide repertoire of behaviors

These metrics are not perfect, but they capture useful facets of a meaningful definition of pain. As such, it seems like it would be a simple start to rigorously examine if a large variety of invertebrate species meet each of these. Some of the research that exists has been assembled (Ray, 2017b). Unfortunately, many of these answers are nonexistent in the research outside of a few select species, and much of the existing research is flawed, old, or scant. Much common ‘knowledge’ in the field seems to be based on impressions and not on controlled experiment – for instance, that insects don’t limp on damaged appendages (Eisemann et al., 1984), or that bivalves don’t learn. As such, we’re in the dark about making important decisions regarding invertebrate welfare.

Some might say that in a case like this, it’s best to be conservative and assume that a wide variety of organisms probably feel significant pain from a wide variety of stimuli (e.g. (Birch, 2017)). This might be prudent, but given how little attention is paid to the welfare of invertebrates of any kind, it seems best to begin addressing the problem where efforts are most likely to be effective, and where we can most clearly convince others that their efforts will be worthwhile. This indicates that we should look to areas that research suggests will be most valuable.

What research are we missing?

Unfortunately, a great deal.

There are major clades that have scarcely been studied. Some of these represent among the most common organisms on Earth. These include springtails, mites, nematodes other than Caenorhabditis elegans, and many types of zooplankton.

Within a taxon, what research exists is often sporadic. A few animals, like ants and bees and model organisms like Drosophila, C. elegans, and the Aplysia sea slug, have been studied in considerable depth. For all other species, much data is drawn from small numbers of individuals in a small handful of species, in inconclusive research. We are missing information on a wide number of echinoderms, nematodes, insects, crustaceans, annelids, molluscs, and other groups.

For example, even astute animal welfare activists often claim that damaging or killing bivalves (clams, oysters, etc) is probably ethically neutral, because their nervous systems are so simple (Cox, 2010).The last part is certainly true – the entire nervous system of the bivalve comprised of four nerve cords and six ganglia (Kraeuter & Castagna, 2001), and a small clam has perhaps 6000 neurons in total (fewer than a fruit fly) (Kohl & Jefferis, 2011). Nonetheless, I’m not aware of studies that have gone out to specifically assess bivalve ability to learn, damage response, etc. The literature consensus generally accepts that they do not (Crook & Walters, 2011), but it seems possible we wouldn’t know until we check. This may also vary between bivalves or in different life stages – while clams and oysters are generally sessile, scallops and some freshwater clams are motile, and bivalve larvae are motile.

A different kind of missing research is in which kinds of stimuli, especially stimuli likely often felt in the environment, produce pain-like responses. Electric shock and perhaps heat and mechanical force, for instance, tend to be somewhat better studied, but there is much less evidence on the effects of starvation, dehydration, osmotic stress, various kinds of predation and parasitism, air exposure (in the case of aquatic invertebrates), or toxins of various kinds (although some existing research on suffering caused by pesticides has been compiled (Eskander, 2017)). As such, it is difficult to determine which kinds of common invertebrate experiences are likely to be painful.

Are the results negative or just nonexistent?

A lack of published research is possibly weak evidence that an organism doesn’t exhibit a certain response.

To illustrate why, imagine you’re a scientist trying to come up with a study on which to publish a paper. You decide to see if sea cucumbers can be classically conditioned. You acquire half a dozen sea cucumbers and do some preliminary testing. You soon find that although they move away from electric shocks or being poked with a needle, they don’t seem to learn to pair them with a light stimulus. That wouldn’t be a very exciting paper, and you’re not sure which quality journal would accept it, so you scrap the whole project and look for something more interesting to study. You don’t publish anything.

In another example, nobody has ever studied sea cucumber response to pain, at least in any kind of formal setting. You consider the idea one day, but never follow up. You don’t publish anything.

In both cases, the literature doesn’t report anything. The lack of incentive to publish negative results, and thus the number of published negative results, is a major ongoing issue (Fanelli, 2011). This is especially unfortunate in the case of invertebrate suffering research. In these cases, negative results are crucial information for making informed decisions about animal welfare. When we search the literature and find no rigorous studies on sea cucumber learning, does that mean that attempts have been made to train sea cucumbers and have failed, or that simply nobody has investigated the question? We don’t know, and this story is repeated wherever knowledge is missing.

Pre-registration – a guarantee with a particular journal to publish a study, based on its methodology, before its results are known – is among several promising tools for reducing publication bias and ensuring accurate results reach the literature (Hopkins & Hopkins, 2015; Munafò et al., 2017). Publication of negative results on this topic would be very valuable, even in brief, such as letters to the editor.

Problems and conflicts in existing research: decapod learning as an example

For instance, the research on decapod crustacean learning and sentience is very unclear. much existing crustacean pain and learning research has been conducted using electric shocks and formalin injections as negative stimuli (e.g, Appel & Elwood, 2009; Denti, Dimant, & Maldonado, 1988). (Formalin and shocks are known to cause pain and inflammation in mammal models, and do elicit responses in crustaceans.).

However, Puri and Faulkes (2015) point out that both electric shocks and formalin have nonspecific effects on all cells, not only nociceptors or even only neurons (Puri & Faulkes, 2015). They write that “nociceptive behaviours triggered by such stimuli may represent abnormal responses of the nervous system rather than the workings of a nociceptive sensory system tuned to tissue damage by evolution.”

In other experiments, crabs treated with morphine appear to be less motivated to escape unpleasant environments – while in fact, morphine may have just reduced their overall movement (Barr & Elwood, 2011). Similar confusion has been described for the use of electric shocks during insect learning studies (Forman, 1984). These common experiments may not represent evidence for pain in the way that we are most interested in.

Even assuming that electric shocks to cause nociception, their role in nociception is uncertain. Studies find that crabs seem to learn, or make trade-offs for things they want, when given electric shocks (Appel & Elwood, 2009; Denti et al., 1988; Elwood & Adams, 2015). Other studies have found that crayfish can be trained to do some behaviors to avoid electric shocks but not others (Kawai, Kono, & Sugimoto, 2004), and that shore crabs don’t seem to be averse to electric shocks at all (Barr & Elwood, 2011).

Injection of morphine might reduce pain or aversion from noxious stimuli in crab (Lozada, Romano, & Maldonado, 1988) and mantis shrimp (Maldonado & Miralto, 1982), unless it just reduces their movement overall and pain isn’t a relevant factor (Barr & Elwood, 2011).

Crayfish seem to limp on and rub limbs injected with formalin (Dyuizen, Kotsyuba, & Lamash, 2012), and prawns are widely cited as grooming antenna that have been brushed with noxious substances (Barr, Laming, Dick, & Elwood, 2008), but a study with three other different species of prawn didn’t find any such behavior (Puri & Faulkes, 2010).

What do we make of such conflicting results? Granted, few of these studies are on the same species, so perhaps there is no conflict, but on the other hand, our intuition suggests that complex behaviors and neural capacities ought to be conserved between closely related species. In any case, it seems strange to imagine that some prawns and crayfish respond to the local site of an injury, but that other similarly-sized prawns in the same order do not. Are some of each set of studies incorrect? Do the small differences in stimuli between each experiment explain the difference (e.g., perhaps damage responses in limbs and antenna are generally different in crustaceans, or the specific choice faced affects the drive to avoid electric shock differently)?

We do not know, and little research is currently being done into these basic questions, although well-done research would quite possibly resolve our uncertainties. Worse, these studies are often referred to in isolation, without their confusing context, giving an incomplete picture of our current understanding.

A diversity of research would help us generate heuristics

With an estimated 8 million species of animal (“How many species on Earth? About 8.7 million, new estimate says,” n.d.) on Earth, a detailed study of which ones have properties associated with pain is unlikely to be completed any time soon. At the moment, though, we have so little data on invertebrate behavior that it’s difficult to even make inferences about organisms that haven’t already been studied (or had a close relative studied).

Some categories it would be valuable to have heuristics for include capacity for suffering between different classes, between sizes of animals (e.g.: there’s at least an 100-fold size difference between the smallest and largest beetles – do their different brain sizes imply they have different behavior?), between larval or pupal and adult animals, between social and nonsocial animals, and between animals in various environments or ecosystems. All this would require is an increased interest in researching the behavior of a variety of invertebrates, and recording the results. With this, we’ll be able to make better judgement calls about the lives, conditions, and moral relevance of all animals.

As a specific example, social insects tend to lead very different lives than solitary insects. On one hand, we might expect social insects to be more likely to show pain, since they’re more able to obtain friendly support from conspecifics by doing so. (Crook & Walters, 2011) On the other hand, social insects also do some behaviors that seem to be voluntary and contraindicate suffering (e.g. self-sacrifice) (Shorter & Rueppell, 2012). Studies here seems to be disproportionately conducted on ants and bees, and a wider range of comparisons of both social and non-social insects would help determine if this is a real trend or not. If it turns out that social insects are either less likely to, or have a lower capacity to suffer, this has implications for how we should treat domesticated bees, and how much attention we should devote to e.g. different kinds of insects.

How would this research help us treat animals more ethically?

It seems unlikely that our civilization will suddenly transition to treating insects and zooplankton with the greatest care and concern. That said, there is some precedent for humane treatment of invertebrates, and similar approaches might be used with organisms that are especially likely to suffer.

In Europe, for instance, a couple organizations have created ethical standards for the use of cephalopods. Decapod crustaceans were also under consideration for these standards by the European Food Standards Agency, but the notion was fought, and ultimately, octopuses were the only invertebrates given such protections (Birch, 2017). Jonathan Birch writes that because decapod crustaceans experimentally showed the criteria of sentience (because of the findings involving prawns with grooming damaged antennae), by applying the precautionary principle, all animals in the order Decapoda should fall under standards for ethical treatment (Birch, 2017) – at least until such a point as their capacity for suffering-like behavior is conclusively determined.

Populations of humans already consider invertebrate welfare important: Jain monks are famous for taking actions that avoid harming insects (e.g., wearing masks to avoid breathing in insects), and many vegans avoid eating honey or using beeswax or silk. Some people keep shrimp, spiders, snails, or other invertebrates as pets. The recent protections afforded to the octopus in Europe suggest that legal protections for other invertebrates are politically possible, although, of course, there will probably be more opposition for invertebrates used more widely by humans (e.g. from shrimp- and crab-fishing industries, which would face major costs by instituting more humane practices.)

Of course, these standards are only applied to animals already used by humans. When focusing on the vast numbers of wild animals, it might make sense to first direct attention towards large and clearly sentient animals that people are more inclined to empathize with – the large mammals that are the so-called “charismatic megafauna”, intelligent birds, etcetera. But the welfare of invertebrates may well be more significant than that of comparatively rare larger vertebrates, and if this seems to be the case based on further study, the welfare of small invertebrates will be a crucial part of practical and effective future efforts to reduce the suffering of wild animals.


Invertebrates represent a huge portion of the biosphere. Humans interact with invertebrates in many forms, and the scale is even vaster in the wild. Unfortunately, the ability of invertebrates to feel pain is understudied by scientific research – many major clades are missing, many important questions are unanswered, and existing research is sometimes old or contradictory. Although the results will always be subjective in some way, an accurate understanding of invertebrate behavior and capacity to suffer would have major implications for how we should treat them, and how much of our limited resources and time we should pay to improving their welfare. Researchers should be encouraged answer these questions.


Appel, M., & Elwood, R. W. (2009). Motivational trade-offs and potential pain experience in hermit crabs. Applied Animal Behaviour Science, 119(1), 120–124.

Barr, S., & Elwood, R. W. (2011). No evidence of morphine analgesia to noxious shock in the shore crab, Carcinus maenas. Behavioural Processes, 86(3), 340–344.

Barr, S., Laming, P. R., Dick, J. T. A., & Elwood, R. W. (2008). Nociception or pain in a decapod crustacean? Animal Behaviour, 75(3), 745–751.

Bastian, B., Loughnan, S., Haslam, N., & Radke, H. R. M. (2012). Don’t mind meat? The denial of mind to animals used for human consumption. Personality & Social Psychology Bulletin, 38(2), 247–256.

Bilewicz, M., Imhoff, R., & Drogosz, M. (2011). The humanity of what we eat: Conceptions of human uniqueness among vegetarians and omnivores. European Journal of Social Psychology, 41(2), 201–209.

Birch, J. (2017). Animal sentience and the precautionary principle. Animal Sentience, 2(16). Retrieved from

Borrell, B. (2012). One-fifth of invertebrate species at risk of extinction. Nature News.

Broom, D. M. (2007). Cognitive ability and sentience: which aquatic animals should be protected? Diseases of Aquatic Organisms, 75(2), 99–108.

Cox, C. (2010, April 7). It’s OK for Vegans to Eat Oysters. Retrieved November 27, 2017, from

Crook, R. J., & Walters, E. T. (2011). Nociceptive behavior and physiology of molluscs: animal welfare implications. ILAR Journal / National Research Council, Institute of Laboratory Animal Resources, 52(2), 185–195.

Denti, A., Dimant, B., & Maldonado, H. (1988). Passive avoidance learning in the crab Chasmagnathus granulatus. Physiology & Behavior, 43(3), 317–320.

Dyuizen, I. V., Kotsyuba, E. P., & Lamash, N. E. (2012). Changes in the nitric oxide system in the shore crab Hemigrapsus sanguineus (Crustacea, Decapoda) CNS induced by a nociceptive stimulus. The Journal of Experimental Biology, 215(Pt 15), 2668–2676.

Eisemann, C. H., Jorgensen, W. K., Merritt, D. J., Rice, M. J., Cribb, B. W., Webb, P. D., & Zalucki, M. P. (1984). Do insects feel pain? — A biological view. Experientia, 40(2), 164–167.

Elwood, R. W., & Adams, L. (2015). Electric shock causes physiological stress responses in shore crabs, consistent with prediction of pain. Biology Letters, 11(11).

Elwood, R. W., Barr, S., & Patterson, L. (2009). Pain and stress in crustaceans? Applied Animal Behaviour Science, 118(3), 128–136.

Eskander, P. (2017, July 11). An Analysis of Lethal Methods of Wild Animal Population Control: Invertebrates – Wild-Animal Suffering Research. Retrieved November 20, 2017, from

Fanelli, D. (2011). Negative results are disappearing from most disciplines and countries. Scientometrics, 90(3), 891–904.

FAOSTAT. (n.d.). Retrieved November 22, 2017, from

Forman, R. R. (1984). Leg position learning by an insect. I. A heat avoidance learning paradigm. Journal of Neurobiology, 15(2), 127–140.

Gilette, R. (2008). Global study of shrimp fisheries. Retrieved November 22, 2017, from

Harmon, K. (n.d.). Octopuses Gain Consciousness (According to Scientists’ Declaration). Retrieved November 4, 2017, from

Hopkins, D., & Hopkins, D. (2015, March 13). How to make scientific research more trustworthy. The Washington Post. Retrieved from

How many species on Earth? About 8.7 million, new estimate says. (n.d.). Retrieved November 8, 2017, from

Kawai, N., Kono, R., & Sugimoto, S. (2004). Avoidance learning in the crayfish (Procambarus clarkii) depends on the predatory imminence of the unconditioned stimulus: a behavior systems approach to learning in invertebrates. Behavioural Brain Research, 150(1-2), 229–237.

Kohl, J., & Jefferis, G. S. X. E. (2011). Neuroanatomy: decoding the fly brain. Current Biology: CB, 21(1), R19–20.

Kraeuter, J. N., & Castagna, M. (2001). Biology of the Hard Clam. Gulf Professional Publishing.

Loughnan, S., Haslam, N., & Bastian, B. (2010). The role of meat consumption in the denial of moral status and mind to meat animals. Appetite, 55(1), 156–159.

Lozada, M., Romano, A., & Maldonado, H. (1988). Effect of morphine and naloxone on a defensive response of the crab Chasmagnathus granulatus. Pharmacology, Biochemistry, and Behavior, 30(3), 635–640.

Maldonado, H., & Miralto, A. (1982). Effect of morphine and naloxone on a defensive response of the mantis shrimp (Squilla mantis). Journal of Comparative Physiology, 147(4), 455–459.

Munafò, M. R., Nosek, B. A., Bishop, D. V. M., Button, K. S., Chambers, C. D., du Sert, N. P., … Ioannidis, J. P. A. (2017). A manifesto for reproducible science. Nature Human Behaviour, 1(1), s41562–016–0021.

Office of Laboratory Animal Welfare | (n.d.). Retrieved November 4, 2017, from

Puri, S., & Faulkes, Z. (2010). Do decapod crustaceans have nociceptors for extreme pH? PloS One, 5(4), e10244.

Puri, S., & Faulkes, Z. (2015). Can crayfish take the heat? Procambarus clarkii show nociceptive behaviour to high temperature stimuli, but not low temperature or chemical stimuli. Biology Open, 4(4), 441–448.

Ray, G. (2017a, June 29). How Many Wild Animals Are There? – Wild-Animal Suffering Research. Retrieved October 2, 2017, from

Ray, G. (2017b, June 29). Which Invertebrate Species Feel Pain? – Wild-Animal Suffering Research. Retrieved November 17, 2017, from

Shorter, J. R., & Rueppell, O. (2012). A review on self-destructive defense behaviors in social insects. Insectes Sociaux, 59(1), 1–10.

Tomasik, B. (2017, January 29). Insect Suffering from Silk, Shellac, Carmine, and Other Insect Products. Retrieved November 22, 2017, from

Waycott, B. (2017, November 3). Will Ablation-Free Shrimp Become a Reality? Retrieved November 4, 2017, from


1 Tomasik estimates almost 1 trillion silkworms used by humans. The global shrimp harvest (farmed and wild-caught) in 2008 was 6 million tons. Assuming about 45 shrimp a pound, that’s 500 billion shrimp. Humans also own a total of about 80 million beehives, and with about 40,000 bees in a hive, we get about 3 trillion bees directly managed by humans.