by Brian Tomasik
First published: 6 Feb. 2017; last update: 14 Feb. 2017


While bivalves are probably less sentient than most animals of their size, they still sense their environments, show altered morphine levels in response to trauma, and adjust to changing environmental conditions.

Note: This page is a work in progress. I have not extensively researched this topic yet.


Mussel dishSome effective altruists who shun most animal foods embrace ostroveganism, largely based on Fleischman (2013): "The ethical case for eating oysters and mussels". While I agree that bivalves matter less per individual than most other animals that humans eat, I'm not convinced that bivalves can't suffer to some degree. Considering the numbers of bivalves that people eat in one sitting and the brutality of the ways bivalves are killed, I worry that bivalve suffering is not negligible.

Sensory system

Bivalves have a number of sensory abilities:

The sensory organs of bivalves are not well developed and are largely located on the posterior mantle margins. The organs are usually mechanoreceptors or chemoreceptors[...]. The chemoreceptor cells taste the water and are sensitive to touch. [...] The osphradium is a patch of sensory cells located below the posterior adductor muscle that may serve to taste the water or measure its turbidity[...]. Statocysts within the organism help the bivalve to sense and correct its orientation. Each statocyst consists of a small sac lined with sensory cilia that detect the movement of a mineral mass, a statolith, under gravity.[20][21] In the order Anomalodesmata, the inhalant siphon is surrounded by vibration-sensitive tentacles for detecting prey.[22]

[...] All bivalves have light-sensitive cells that can detect a shadow falling over the animal.[17]

Cascio (2017):

While scallops and clams are motile and have simple eyes and photoreceptor cells respectively, their movements are automated by light level shifts and response to physical stimuli. This is similar to a human pupil dilating due to shifting light levels which trigger iris nerve innervation from the parasympathetic nervous system, rather than out of a conscious and aware decision. [1, 2, 3] This is also similar to the Venus fly trap, which detects stimuli and clamps down in response. The mechanisms are different, but the involuntary reflex aspect are biologically analogous. These reactions allow some bivalves to be able to hide and crudely navigate[...].

Morphine studies

Sonetti et al. (1999)

In this section, I cite Sonetti et al. (1999), which found in a mussel "endogenous morphine [..] in specific tissues of these animals [that] appears to be involved in the response to physical trauma." In particular, that physical trauma was "cutting with a fine lancet the shell posterior adductor muscle." So cutting live mussels may be painful.

Cadet et al. (2002)

Cadet et al. (2002) studied the edible blue mussel (Mytilus edulis) and found that (p. 31)

Rapid exposure of whole mollusks to cold temperatures from maintenance at room temperature alters ganglionic opiate processes. We have [proved that] such treatment results in significantly enhanced levels of ganglionic opiate alkaloids that are time-dependent. [...] This study also demonstrates that both μ opiate receptors and opiate alkaloids are expressed under basal conditions, suggesting their constant use by the organism. With this in mind, it would appear that opiate signaling is involved in this organism’s response to thermal stress.

Cadet et al. (2002) refer (p. 31) to other studies on Mytilus mussels, including one which found that "Mytilus starvation also results in higher ganglionic morphine levels [26]." The authors say: "This has led us to surmise, given morphine’s general immune and neural down regulating actions, that it functions to limit the extent of excitation caused by an initial stressor, i.e. thermal, restoring homeostasis [13,21,23]."

If mussels are stressed by cold water, perhaps they're also stressed by hot water? If so, boiling them alive might be quite bad. Sadly, boiling alive is standard practice: "Mussels are alive, and you want them alive when you cook them."

Inducible defenses

Freeman (2002) describes a study on blue mussels, which I think is Leonard et al. (1999), although I haven't read the full text of that article. Freeman (2002) explains (p. 956):

George Leonard and co-workers hypothesized that inducible defenses might be important for blue mussels living in an estuary along the coast of Maine [...].

They found that mussels in the high-predation area had thicker shells and were more strongly attached to their substrate than mussels in the low-predation area [see the figure below]. [...]

These results show that molecules that are shed by crabs or by killed conspecifics induce blue mussel defenses. What are these chemical cues, and how are they sensed? These questions remain unanswered.

Of course, inducible defenses also occur in plants. In my opinion, this shouldn't dampen our appraisal of mussel sentience but should augment our appraisal of plant sentience.

Aren't bivalve behaviors just unconscious reflexes?

Cascio (2017) says:

A good way to think of bivalves is to relate it to if you were to hit a fresh human cadaver’s knee with a reflex hammer. If the body is not too decayed, the knee will reflex and jerk forward, because of localized stimuli processing. This obviously does not cause any pain to the cadaver, because the brain is disconnected from the nervous systems. [...]

If the argument against eating bivalves is that the conversation is fundamentally flawed because we shouldn’t ascribe human ideas of sentience to nonhuman animals, that argument would need to extend to plants and fungi as well, especially those that exhibit thigmonasty as they are, in a sense, crudely aware of external factors. If the argument is that motility or a motile stage indicates possible sentience, we would need to look at all fungi and algae that have a zoospore stage.

I have two different replies here.

1. Phylogenetic similarity

Bivalves are much closer phylogenetically to animals that we know are sentient. As Cascio (2017) notes: "There are mollusks that problem solve and display complex behavior, such as octopuses, while their cousins, bivalves, have no behavior other than reflexes." While I'm not familiar with the details, it's plausible that the common ancestors of bivalves and other present-day mollusks were more sentient than bivalves are today. We don't fully know how much cognitive machinery those ancestors had or how much bivalves have retained. Until we've exhaustively studied the behaviors and nervous systems of bivalves, we should bet that bivalves have somewhat more advanced cognitive capacities than plants/fungi do.

Fleischman (2013) acknowledges the possibility of vestigial sentience but replies that "This is very unlikely because pain is biologically expensive". I would reply that most vestigial structures are at least somewhat costly, but they're retained by evolution because they're hard to eliminate. Maybe bivalve ancestors used more complex brain processing, and some of that machinery was retained in bivalves because it was too hard for evolution to jump to a more efficient alternative nervous system, in a similar way as it's hard to make a large software codebase as efficient by incremental revisions as if one could start over from scratch.

Our understanding of bivalve nervous systems is not sufficient to rule out modestly complex neural computations. For comparison, we've mapped the 302 neurons of C. elegans, but these tiny creatures display an amazing array of complex behaviors that we're nowhere near being able to "read off" from their connectome. Meanwhile, oysters have thousands of neurons. There could be some mildly sophisticated cognitive processing hidden within that morass of nerve cells that we're not in a position to notice yet. And if these cognitive operations are vestigial, then they shouldn't necessarily show up in macroscopically visible behavior, in a similar way as you can't tell just from the fact that a program can add numbers whether the algorithm it uses to do that addition is very complicated or very simple.

2. A more cosmopolitan view of suffering

My second reply is that I do also feel that plants, fungi, and algae deserve to be seen as marginally sentient. The basic idea is to point out that, at a high level, we are all just stimulus-response machines—receiving inputs from the environment and producing changes in physiology and behavior in response. What's different among plants, bivalves, and humans is the degree of complexity involved. Humans engage in far more detailed processing of input stimuli into more complex mental representations, and have more sophisticated internal self-reflection and self-stimulation systems. But as far as I can see, there aren't discontinuities here. There's not a single cognitive architecture that's required before the "lights of consciousness" turn on, where before all was darkness. Rather, there's a vast space of possible mind designs, some with greater detail and self-reflectiveness than others. In another piece, I quote Daniel Dennett expressing a view that I share regarding the question of clam sentience:

It is a big mistake to think of consciousness as all-or-nothing. [...] Are clams conscious? Are fish conscious? Are vertebrates conscious? Are octopuses conscious? And I think that those are just ill-formed questions. Let's talk about what they can do in each case and what their motivational systems are, what emotional possibilities they have. And as we sort that all out, what doesn't happen, I don't think, is that we see emerging from the gradual fog a sort of sharp line at any point.

Cascio (2017) believes that bivalves have "no capability of having thoughts". But what are "thoughts"? Thoughts are patterns of neural (and, according to some philosophers, also somatic) activation that represent certain information. Thus, thoughts have degrees of complexity and self-awareness. There doesn't seem to be a principled level of complexity at which "thoughts" stop taking place.

Applying the above points to the morphine studies

David Cascio says regarding the morphine studies I mentioned above:

some research has shown that one species of bivalve has μ opioid peptide receptors. The question is, what is the entire purpose of mu opioid peptides? They are not just for pain. They are also used to suppress unnecessary biological functions in response to stress. Without a brain or a way to process/experience the pain, the peptides are very likely present as a neurotransmitter to alert the body to direct its focus away from the nonessentials and regulate its cardiovascular system.

My reply mirrors points #1 and #2 from the previous section.

Phylogenetic similarity

Given that some of the neurochemistry is shared between bivalves and mammals, and given that bivalves (unlike plants/fungi) have neurons, we don't fully know how much pain-like processing bivalve neurons are capable of, nor how much morphine acts on that neural processing.

A more cosmopolitan view of suffering

What is "pain" anyway? I agree with Minsky (1998):

I regard each emotional state to be a different arrangement or disposition of mental resources.

[...] the feeling of pain results from the engagement of certain special resources. [...] Presumably each common emotion involves arousing a variety of particular processes in different brain [and, Brian would argue, body] centers. These in turn will then affect how some other mental resources will be disposed.

[...] As I see it, feelings are not strange alien things. It is precisely those cognitive changes themselves that constitute what "hurting" is[...]. The big mistake comes from looking for some single, simple, "essence" of hurting, rather than recognizing that this is the word we use for complex rearrangement of our disposition of resources.

In other words, a bivalve's stress reaction—which alters the animal's allocation of resources in response to a problem—can be seen as one rudimentary component of pain. And when Cadet et al. (2002) surmise that morphine "functions to limit the extent of excitation caused by an initial stressor" (p. 31), this is a crude form of the idea that "morphine dampens pain" that we see in mammals.

Cascio (2017) says "Bivalves do not use endogenous opiates or opiate receptors to inhibit pain". But if we construe "pain" broadly, in the way I suggested, then there's legitimate room to dispute Cascio (2017)'s contention (even without appealing to uncertainty and giving bivalves the benefit of the doubt) because bivalves may indeed use morphine to inhibit a crude form of pain (i.e., stress reactions).

In more sophisticated animals like humans, stress responses co-occur with higher-level thoughts, memories, and verbal declarations about the noxious stimulus. But once again, it's not clear where to draw the line between simpler vs. more complex forms of stress responses, and a sliding scale of sentience seems more sensible to me.

How harmful is eating bivalves?

A full calculation of the consequences of eating bivalves vs. plant proteins is complex and requires analysis of the (positive and negative) impacts of various food production methods on wild-animal suffering. I haven't done this calculation, in part because the world's bivalve consumption isn't enormous, so this may not be the most important issue to work on.

Robinson et al. (2002) note that "bivalves are generally regarded as herbivores." Given that bivalves are both bigger and possibly less sentient than most zooplankton, maybe it's good to farm bivalves in order to reduce the amount of food available for zooplankton?

Anyway, I personally avoid bivalves because of uncertainty about how sentient they actually are and because large numbers of them are killed in brutal ways for a single meal. However, I would probably give a single bivalve less moral weight than a single insect, even if the insect is many times smaller. So, for instance, entomophagy seems much worse to me than eating bivalves.