North over Limestone

Lily Moth (Polytela gloriosae), family Noctuidae, Sri Lanka

Rocky and Grace

OP: How I created this visual magic (cr 阿梓蔗)

(a post I am not putting under a read more so that anyone who reblogs it can have the full text available regardless of what I do to my blog specifically in the future)

TLDR: We don’t know. Let me get that out of the way. We very much Do Not Know. I am going to go into detail about a number of studies, the conclusion of which, when taken together, is ‘we need to do more studies in a way that may not currently be possible.’

Now come with me on a journey where I say that again but take over 5000 words to do it.

The cold, wet facts: what we can be reasonably sure of about hagfish sex

Hagfish gonads are located in the peritoneal cavity, a space between the membrane that surrounds the internal organs and the membrane that lines the abdominal wall. Pictures of hagfish gonads can be seen in Gorbman 1990, Powell 2004, Martini 2013, Weinrauch 2015, Muramatsu 2024. While it starts as a paired organ, one gonad withers early in development, leaving them with a single functional gonad.

The gonad is very long. In immature hagfish, the whole gonad is undeveloped and undifferentiated – there is gonadal tissue present, but it’s not making any gametes, nor has it developed the structures to do so. At some point in their lives, which is currently assumed based on growth patterns to be several years after hatching, the gonadal tissue begins to develop. For the most part, when the anterior (towards the head) two-thirds of it develop to maturity, the tissue produces eggs. When the posterior (towards the tail) third develops to maturity, the tissue produces sperm. In some individuals, gametogensis occurs outside of these boundaries. This seems to be more common in larger hagfish, and one proposed explanation is that as hagfish age, gametogenic tissue expands past the border of undifferentiated tissue that usually separates the anterior and posterior sections of the gonad in order to increase the quantity of gametes the organism can produce.

In a very small percentage of the population (exact numbers are unknown and almost certainly vary based on subfamily, genus, and species, but in the larger and more recent population surveys of several species in the genus Eptatretus the incidence is estimated at below 1%), the entire gonad develops to maturity, with the anterior two-thirds producing eggs and the posterior third producing sperm simultaneously. Whether these individuals are self-fertile, or fertile at all, is currently unknown. In other individuals, there appears to be no gonadal development even after reaching the sizes that are assumed to indicate sexual maturity – however, it’s unclear how many of these individuals are actually adults who have not and will never sexually differentiate, as opposed to above averagely large juveniles or adult hagfish in a part of their reproductive cycle where their gonads look undeveloped to the naked eye. Some hagfish gonad developmental stages being difficult to distinguish from each other without a microscope is an important fact that will become relevant repeatedly.

People studying hagfish reproduction frequently sort hagfish into different stages based on the size and development of their gonads and gametes. In hagfish that produce eggs, these stages are primarily defined by the size of the eggs, which start small and numerous and then grow and reduce in number until a fully developed clutch of usually around 10-30 eggs is ready to be fertilized and laid. In hagfish that produce sperm, the stages are primarily defined by the size of the testicular follicles.

What exactly hagfish reproduction looks like behaviorally is mostly unknown. We don’t know how they fertilize their eggs or where those eggs are laid. Inshore hagfish (E. burgeri) are believed by some to have a synchronous spawning cycle, and the population has been reported to synchronously develop mature gametes and predictably migrate in association with this hypothesized mass spawning – however, the actual spawning has never been directly observed. Mature hagfish just seem to move en masse from the shallows to deeper waters, and when they come back, the females that were formerly full of mature eggs don’t have those eggs any more. An additional piece of evidence for the synchronous spawning hypothesis is that the only known protocol for collecting hagfish eggs for laboratory use, described by Ota K.G., Kuraku S., & Kuratani S. (2007), involves collecting mature wild E. burgeri at a specific time in the year and leaving them alone in tanks to do their unknown thing until eggs appear. The predictable, cyclical nature of their reproductive cycles allows evolutionary developmental biologists and other scientists to obtain hagfish embryos for study even without the knowledge of how those embryos come to be. In all the other species of hagfish I was able to find detailed reproductive data on, hagfish in most if not all stages of gonadal development were present in nearly every collection, which suggests that they are not synchronous spawners.

Hagfish are mostly deep water animals that live on and near the sea floor. They do not have complex eyes, but they do have eye patches that are sensitive to light. That means that they are affected by a major problem with studying life in the deep sea. Even when humans can get past the challenges of reaching the hagfish in their environment, in person or with machines, if we expose them to light so we can observe them we invariably alter their behavior. A sudden big light somewhere that’s too deep underwater for sunlight to penetrate is, understandably, alarming and disorienting for the animals that live there. We are very dissimilar organisms, and it makes collecting information about living hagfish difficult.

The studies: The sequence of claims that did or did not make it to the public and institutional consciousness about hagfish sex

We’ll start with a population survey of Pacific hagfish (E. stoutii) published in 1990 by Aubrey Gorbman, whose work is still cited on several government-run fishery websites and in multiple papers and books on hagfish. Gorbman assessed 100 individual Pacific hagfish and concluded that prior assertions that hagfish display protandry (all individuals differentiate as male first and then become female later in life) was based on misdiagnosis of developing ova as testicular follicles. He claimed instead that hagfish display protogyny and every juvenile will begin to develop ovarian tissue in the anterior section of the gonad when approaching sexual maturity. In some hagfish, ovarian tissue development proceeds through all of the identified stages, resulting in a sexually mature hagfish who produces eggs. In other hagfish, ovarian tissue development reverses and the posterior portion of the gonad develops into testes while the partially-developed ovarian tissue in the anterior of the gonad degrades, resulting in a sexually mature hagfish who produces sperm (but might still contain identifiable ovarian tissue in an early developmental or degenerating state, and permanently retains the “vascular and connective tissue framework” (317) used to support development of the anterior section of the gonad in that juvenile stage). And in a small portion of the population, the entire gonad develops into a reproductive organ that simultaneously produced ova and sperm. He based the claim of juvenile protogyny on the observation that all hagfish in his sample below a length of 20cm contained what he identified via microscope histology as differentiated ovarian tissue. Some specimens from 16 to 24 cm long contained intermixed ovarian and testicular tissue in the same section of the gonad, which he viewed as evidence of a transitional phase, as he did not find any larger hagfish with intermixed tissue. He did, however, find 3 hagfish with gonads that had fully developed along their length into ovarian tissue that produced eggs in the anterior and testicular tissue that produced sperm in the posterior.

In 2001, Davis et al. published their examinations of a small number of Atlantic hagfish (M. glutinosa). (they also did some experiments with injecting lamprey hormones into hagfish but I am interested in how hagfish sexual development works without encounters with scientists doing cyclostome HRT so I am only counting the results from the first part of the experiment). Their sample size was VERY small, and so I don’t think it’s reasonable to take it as reflective of wild populations, but what interests me about this study is that despite the small sample size they still identified multiple hagfish that contained both mature ovarian and mature testicular tissue. They also identified some hagfish with intermixed ovarian and testicular tissue, but lumped those in with the hagfish with undifferentiated gonadal tissue in their data set so I have no further information what that looked like or how many of them there were. But it’s worth noting.

In 2004 Powell et al. published an experiment measuring hormone concentrations within the gonads of M. glutinosa. Unfortunately they also made some choices with how to sample the gonads that interfered with their results. Their results indicated that estradiol and progesterone levels seem to fluctuate seasonally within hagfish gonads regardless of whether they produce eggs, sperm, both, or neither. However, all samples taken prior to November 2001 were only from the anterior region of the gonad (you may recall one of the known things about hagfish gonads is that usually a developed anterior produces eggs and a developed posterior produces sperm – if you only sample the anterior portion of the gonad of all your hagfish you will almost exclusively end up with ovarian or undifferentiated tissue and you won’t have a full picture of what’s happening in the organism, and if I am reading their graphs right that is exactly the result this produced) and when they started sampling the anterior, middle, and posterior section of the gonad, they found that the middle portion was usually indeterminate tissue, so they didn’t include that data in their analysis. I think that is a meaningful piece of information about the structure of hagfish gonads, but at least they mentioned this at all rather than keeping it out of the paper completely.

Overall, due to the limitations of their sampling methods and the fact that this experiment didn’t run for long enough to see if the hormone fluctuations were a consistent annual pattern, I don’t think this is sufficient evidence for a synchronous spawning cycle, especially because hormone levels did not seem to be connected to the stages of gamete maturation. Hormone levels actually seemed to be more connected to age as estimated from size – the smallest hagfish had the highest amount of progesterone and estradiol at nearly every month in the study. When combined with Gorbman’s hypothesis that all juvenile hagfish go through a period of ovarian development, this offers interesting possibilities. In lampreys, the other surviving group of agnathans, estrogen plays a role in the spawning cycles of both lampreys that produce eggs and lampreys that produce sperm. Lamprey ovarian tissue also has a higher amount of estrogen receptors than lamprey testicular tissue. (Sower & Baron 2011) Hagfish and lampreys diverged a very long time ago and their exact evolutionary relationship has historically been contentious due in part to the fact that hagfish, in the course of their evolution, shed many identifying vertebrate traits, such as having a bony spine. (Kuraku & Kuratani 2006, Ota & Kuratani 2006, Ota et al. 2007, Marlétaz et al. 2024, etc., ‘where do hagfish fit on a phylogenetic tree relative to other vertebrates’ is one of the most heavily researched things about them) However, they are confirmed to use several similar hormones and hormonal pathways which are common to vertebrates. Because of the structure of hagfish gonads, it would make sense to me that if high levels of estradiol and progesterone in juvenile hagfish play a role in the sexual maturation process for all hagfish regardless of sex, it might trigger some amount of ovarian development in all of them, even if they don’t all go on to produce viable eggs. However, this is idle amateur conjecture. Additionally, Atlantic hagfish and Pacific Hagfish belong to two separate lineages of hagfish (Myxine and Eptatretus) which are estimated to have diverged before the non-avian dinosaurs went extinct. (Kuraku & Kuratani 2006 say Cretaceous, Brownstein & Near 2024 say Triassic.) That’s a long time in which to develop different reproductive strategies, so we shouldn’t treat Atlantic and Pacific hagfish as interchangeable.

In 2013, Martini and Beulig make the claim that observations of protogyny in hagfish were likely a misdiagnosis of different timelines of sexual development. They suggest that hagfish have a gonochoric genetic sex system (the genome of hagfish is of interest to evolutionary developmental biologists, among others, so a few hagfish genomes have been cataloged, but I found no evidence that a genetic sex system has yet been identified. This doesn’t mean one doesn’t exist, just that it’s not a settled matter), and animals who will produce ova simply begin gonadal development earlier than animals who will produce sperm, creating a population of juveniles that, in earlier stages, only have ovarian or undifferentiated tissue. They also point out that immature hagfish ovarian tissue can sometimes be mistaken for undifferentiated hagfish gonadal tissue and vice versa. As evidence they submit the observations that the sex ratio of undifferentiated, female, and male hagfish in their study sample shifts as the hagfish get larger, changing from mostly undifferentiated or female at smaller sizes to closer to a 1:1 ratio of males and females at larger sizes, and that all hagfish in their study above a certain size could be identified as male or female upon macroscopic examination.

I do not uncritically accept their conclusions. The main issue I have is that while they do seem to have performed some microscope histology on some of their samples (this is my interpretation of the statement on page 2 that “Staging was verified by histological examination of representative members of each stage,” which to me indicates that they verified their overall ability to sort hagfish through macroscopic examination by confirming their diagnosis via microscope histology performed on one or more individual hagfish per stage), they didn’t do it on all of them, and they primarily categorized the hagfishes’ reproductive organs by macroscopic analysis, as in, by observing them with standard human vision. Several of the papers I read, including this one, note that early ovarian development can only be distinguished from undifferentiated tissue under a microscope. This means that if all hagfish do undergo some amount of ovarian development prior to sexual maturity as Gorbman and Powell et al. believe, the methods used in this study would not catch it. It would have been helpful for clarifying the mysteries of hagfish sexual differentiation if they used a microscope to check hagfish with testicular tissue for the telltale leftover vascular and connective tissue in the anterior section Gorbman described. It would have been particularly helpful considering that the preceding 3 studies had a low volume of hagfish with testicular tissue and some did not have any hagfish at the later stages of testicular tissue development at all, while Martini & Beullig acquired multiple hagfish in every stage of testicular development. This is a different species in a different part of the world from Gorbman’s study, but they are at least both genus Eptatretus, which makes them a little more comparable.

Fleury et al. (2021) seem to share Martini and Beulig’s conclusions about hagfish sexual differentiation to the point of also not doing microscope histology on immature hagfish gonadal tissue and diagnosing different reproductive stages through macroscopic histology alone. As such, while their study had by far the biggest sample size and included both Pacific hagfish and black hagfish (E. deani), I am not sure their numbers are as trustworthy as those provided by people who performed microscope histology on hagfish gonadal tissue. However, the sheer volume of hagfish involved in this study (thousands) means that microscope histology of all of them would be significantly more demanding than in the smaller studies, and diagnosis of mature hagfish gonads by macroscopic analysis is usually more reliable than diagnosis of immature gonads, so the information definitely isn’t worthless. It’s just not as comprehensive as I would like it to be.

There are issues with all of these studies. The first 3 I listed have very small sample sizes. This means, among other things, it’s completely possible that the conclusion that all juvenile hagfish contain ovarian tissue could have been an accident of sampling where they only managed to catch juvenile hagfish that were developing ovarian tissue, not because all hagfish do but because those specific ones did. This is one of the reasons it’s good to have a large sample size, because these sorts of coincidences can get louder and more likely the smaller your data set is. It’s also possible that the researchers in the first 3 studies were wrong about what they were seeing, as this is a recurring problem in hagfish gonad analysis. They could have misdiagnosed undifferentiated tissue as ovarian, and they could have been sampling from areas of the gonad that decreased the likelihood of identifying testicular tissue. It’s possible that the researchers who later dismissed the claims of protogyny in hagfish came to these same conclusions. It’s also possible, because they were affiliated with actual educational and scientific institutions and I am not, that these later researchers were able to examine more of the data from these past studies (more pictures of tissue than appear in the published papers, for example) and disagreed with the histological analyses these conclusions were based on.

However, if that did happen, they didn’t publish that information, and they haven’t responded to my emails yet. And the information they did publish on their methodology and the reasons for their beliefs about hagfish sexual differentiation isn’t enough to convince me. I am fully open to the possibility that Martini & Beullig and Fleury et al. are correct that hagfish sexual differentiation is genetic and hagfish are largely gonochoric with, as in many other gonochoric animals, a smaller percent of the population being intersex. However, I don’t think they’ve collected or provided the data necessary to settle that claim. I don’t think anyone has.

There are also several genera of hagfish. They split a very long time ago, and have over 80 identified species spread between them. They may all seem similar, as the hagfish bauplan needs little improvement or variation because they are perfect organisms, but it would be fundamentally absurd to assume that what’s true for one species of hagfish must be identically true for the rest. If we get a full picture of the reproductive developmental cycles of one species of hagfish, we will still only know how it works for that species of hagfish.

I do find it interesting that nearly every study, even those with a small sample size, apparently managed to capture individuals with both ovarian and testicular tissue. The exception is that Fleury et al. didn’t find any black hagfish with both ovarian and testicular tissue, but they also, as established, weren’t doing microscope histology, and black hagfish are weird for other reasons. Both Fleury et al. and multiple fishery websites (I haven’t yet been granted access to the population surveys that these claims are based on, but it seems worth mentioning because it matches up with Fleury et al.’s results) report a trend of catching notably more female black hagfish than males. No one is sure if this reflects the population-wide sex balance or something about the capture method results in more females than males. Black hagfish live much deeper than E stoutii, with some reports putting their range at up to 2,000 meters deep, which makes it less likely that we’re gathering samples that represent a full picture of what their lives and populations are like.

It would be, at this point, impossible for everyone who has published a hypothesis on hagfish sexual differentiation to be right. But due to the variations in methods used in the studies and the limitations of studying deep sea animals, it’s not easy to determine which hypothesis is most likely to be correct.

Why are people saying they change sex?

You may note that none of the studies I cited claim that hagfish change sex as adults. That’s because I haven’t been able to find any studies in the past 30 years that make that claim or provide physiological evidence for it. The prevailing modern models of hagfish sexual differentiation in papers published by researchers working with hagfish are protogyny or gonochorism. However, many fishery websites, aquarium websites, and other science communication sources report a range of sexual differentiation strategies (protogyny, protandry, serial bidirectional sex changing, environmentally influenced sex differentiation, gonochorism). Presenting a hypothesis without detailed information on the limits of our actual knowledge is an unfortunately common situation in science communication, made even more unfortunate by the fact that it’s possible that the beliefs about hagfish posted on fishery websites are representative of the beliefs about hagfish that are informing policy decisions about commercial fishing of hagfish. Incorporating inaccurate beliefs about an organism’s reproductive strategies and capabilities in decision-making about what level of human-inflicted mortality populations of that organism are able to withstand is not great, historically.

The structure of hagfish gonads does seem to have at least some similarities to structures seen in vertebrates that are known to change sex (see Cole 2002, Maxfield & Cole 2019, and Langston 2023), namely that all individuals possess a gonad with a section with ovarian tissue potential and a section with testicular tissue potential separated by a section of tissue that usually is not involved in gametogenesis, but can become gametogenic later in life. However, there are multiple other factors that do not add up. One is that the social structures of hagfish are different. Many vertebrates that change sex bidirectionally seem to live in pairs and have a high mortality rate due to both short lifespan and high predation risk. It is hypothesized that these pressures makes changing sex an advantageous ability, because it increases every individual’s chances of being able to reproduce with any conspecific they might meet in their fleeting lifetime. (Pla & Piferrer 2021) Hagfish seem to occur in high densities, and most species are hypothesized to have lifespans better measured in decades rather than in days. Furthermore, hagfish are currently believed to reproduce relatively infrequently and none are known to produce a large amount of gametes per reproductive cycle, meaning that taking the time to switch off gamete production in one section of their gonad and switch it on in the other might end up decreasing the overall amount of reproductive chances they have in their lifetime rather than increasing it. While the structure of the gonad could facilitate some degree of species-wide sexual fluidity, it could also facilitate a primarily gonochoric population containing a percentage of intersex individuals with a fully developed gonad. For these reasons, more evidence would be necessary to make a claim that hagfish sexual differentiation involves changing sex, as a one-time event or serially.

I think it’s possible that part of the confusion around hagfish reproduction is due to a larger issue surrounding understanding hagfish in general, which is that people tend to think of hagfish as ‘primitive’ even though their lineage has been around and evolving just as long as everything else alive today. While they did diverge from the rest of the vertebrates a very long time ago and can therefore provide valuable insight into the timeline of the development of various traits in early vertebrate evolution, they aren’t actually frozen in time. They (and the other surviving agnathans, lampreys (also a very cool group of animals, with significantly less mysterious reproductive cycles)) have survived hundreds of millions of years of sharing environments with the proliferating jawed vertebrates. Please consider the advantages of jaws. Contemplate the majesty of the noble hagfish, which not only gets by without but occupies a massively ecologically valuable niche in a challenging environment. Consider the suite of adaptations necessary to enable this. Yet there is a history of people automatically assigning hagfish traits that are assumed to accompany a ‘primitive’, basal, or less sophisticated state, and sexual differentiation strategies outside of gonochorism have historically been one of those traits. Less so these days, but some people are still citing those older sources when they talk about hagfish.

This is particularly frustrating because it’s not untrue that we can use hagfish as a reference point when trying to understand the history of vertebrate evolution. However, it’s not because they’re a fixed window into the past. It’s because we have areas of study like ‘evolutionary developmental biology’ and ‘comparative genomics.’ Understanding hagfish sexual differentiation could tell us more about the history of vertebrate sexual differentiation in general because similarities and differences from other vertebrates may indicate information about our last common ancestor, which was a very long time ago.

So they don’t change sex?

I would say ‘probably not,’ but I would say it with caveats. One, we know very little about alive hagfish, full stop. A lot of what we ‘know’ about hagfish is guesswork, and much of that guesswork is proven wrong when people find ways to actually check. Hagfish are assumed not to move much, but we aren’t really doing catch and release with them and tracking their movements, we are mostly fishing them up and dissecting them and making conjectures based on where we catch a lot of them (it’s difficult to keep a tracking collar on an animal that regularly ties itself in knots.) Hagfish are primarily thought of as scavengers, but have been observed actively and successfully hunting apparently healthy prey while ignoring accessible carcasses. (Zintzen et al. 2011) Hagfish are assumed to live in a dull and empty sensory world, but they have a unique body-wide chemoreceptive system that we know very little about. Two, you can make conjectures about what an organism seems likely to be doing based on other facts about it, but you don’t actually know for sure until you test your hypothesis directly, because nature and evolution are not strictly logical. I can say that it seems like an inefficient allocation of resources to turn different parts of the gonad on and off throughout every individual hagfish’s life based on the observed population density of hagfish and what we’ve observed of their reproductive cycles as compared to the circumstances of animals that are confirmed to possess the capacity to change sex, but there are many traits and behaviors I think are an inefficient allocation of resources that are scientifically validated to occur in living things. Evolution didn’t ask me my opinion on such matters.

What follows is idle personal conjecture and not to be taken as solid information. If hagfish do have a flexible sex determination system, I think it most likely that the flexibility is exclusive to the juvenile stage rather than a permanent ability in adult animals. It’s possible that Gorbman and Powell et al. were correct that all juvenile hagfish go through a period of ovarian tissue development, though if you look at the recorded size ranges in Martini & Beullig and Fleury et al. and take size as an indicator of age (which works best in juvenile hagfish and starts to become problematic when they’re mature, but we don’t currently have a better way to estimate hagfish age), it doesn’t look like every hagfish develops a fully mature ovary and then some later go on to develop a fully mature teste as in true protogyny. The size ranges of hagfish with ovarian tissue and hagfish with testicular tissue are fairly similar, and if all hagfish developed a mature ovary before developing a mature teste, you would expect to see a lot more small mature females and a higher minimum size for mature males. But it’s difficult to observe a dynamic sexual system in organisms that are dead. These samples provide snapshots of a single point in a hagfish’s life, leaving the stages of development before they were caught mysterious and terminally closing the possibility of future development. So even though true protogyny seems unlikely, there remains a possibility that some amount of ovarian development happens in all hagfish.

This next idea is based on very little, but it also seems possible to me that sexual differentiation in some hagfish may not be controlled genetically, but environmentally. This happens in a lot of different animals. If that is the case, it could be very difficult to figure out, because environmental factors that influence sex are varied and we don’t know what factors hagfish may be sensitive to. My pet hypothesis based on nothing is that many species of hagfish appear to be colony animals, so I wonder if juveniles could potentially be responsive to the sex balance of the local population. Because they likely rely heavily on their chemoreceptive abilities to understand and navigate their environment, I believe it makes sense to assume that hagfish receive and respond to chemical information about their local conspecifics in addition to information about nearby predators and prey. They may be able to detect population-wide sex balances, and that information may affect their sexual differentiation.

(As an aside, another piece of idle personal conjecture about the way chemoreception may be a part of the mysteries of hagfish reproduction is that I think they may be able to determine information about the fertility status of specific other individual hagfish. This would facilitate reproduction in the absence of synchronous reproductive cycles. Hagfish don’t seem to produce a large amount of sperm or eggs compared to many other oceanic creatures, so it would be problematic for their individual and species-wide fertility if they didn’t have other means of heightening the likelihood of successful fertilization. Chemical signaling is a very widespread strategy for communicating reproductive information, so it seems like a reasonable possibility that chemoreception plays a part here. This is not a certainty. It would be difficult to confirm. There are many barriers to studying the sex lives of deep sea animals.)

HOWEVER. We don’t know where hagfish lay their eggs or how they fertilize them, we don’t know how the juveniles might differ in their habits and preferred environment from adults, we don’t know what factors control or influence their sexual differentiation to what degree. As much as I personally enjoy learning about different strategies of sexual differentiation, we must be careful not to form hypotheses based on what we personally think is cool. So this is not me saying that hagfish definitely work this way, this is idle speculation. The only thing I am confident saying about this is that I think we are approaching the limits of what dead hagfish can tell us about alive hagfish, and fundamentally this question is not going to be fully answerable without observations of living animals over time.

What would it take to settle this?

If I were to run an experiment to test whether hagfish change sex or if their sexual development is responsive to environmental conditions, I would need multiple difficult things. One, I would need a protocol for effectively keeping hagfish healthy in captivity to the point that they could endure regular biopsies, which we don’t really have, currently. Their average lifespan in captivity is much lower than their assumed average lifespan in the wild. Part of that is that it’s difficult to keep deep-sea animals on the surface. The conditions are very different, we’re only capable of replicating some of them, and we don’t always know which conditions are necessary for animals to thrive, let alone thrive to the point of reproduction. However, another factor is that many people possess outdated beliefs about hagfish biology and do not provide for several identifiable needs and natural behaviors, such as not giving them substrate to burrow in and housing them in empty tanks, which is likely stress-inducing due to the constant feeling of exposure. This is probably related to the bias that suggests the less an organism resembles ‘complex’ and ‘highly evolved’ creatures such as humans (a common species of highly derived lobe-finned fish), the less meaningful its needs are. I am not surprised at the mortality rate. (If you contact me I will advise on hagfish-keeping for free and that is a real offer.)

Two, I would need years. Ideally I would watch hagfish develop from hatching to several years into sexual maturity. Current estimates put hagfish sexual maturity at around 4 years, but these estimates are mostly come from growth rates based on plotting the sizes of dead hagfish or, rarely, measuring growth over time in laboratory conditions with husbandry issues that the researcher running the study admitted could have affected the outcome. (Yamagutchi 2025) And, again. 80 currently identified species of hagfish with a divide between the two major genera dating back to before the K.T. event. They probably don’t all mature on the same timeline. So we don’t know exactly how long this will take. I don’t think there are many institutions willing to provide that kind of funding on that kind of fluid time scale for an organism that doesn’t exactly have charismatic megafauna status. (Do you think more people would like hagfish if they were bigger? Should I open a crowdfunding project to engineer a 10 meter long species of hagfish? Let me know.)

Three, I would need a lot of hagfish, and a lot of lab space. The conditions required for hagfish to mature and the conditions that might affect sex differentiation are unknown, so it would be best to have a multitude of tanks with varied qualities. Furthermore, while I have thoughts on how to improve hagfish husbandry, realistically this experiment is still likely to have a significant mortality rate due to the unknowns in raising hagfish to maturity. The fact is that it is not currently possible to replicate every single feature of the entire deep sea on land, so the setup is certain to be found wanting even if there was an infinite hagfish research budget. However, as there are many complications, risks, and current technological impossibilities in tracking wild animals through the deep sea and repeatedly biopsying them, maintaining hagfish in laboratory conditions that would allow the identification, prolonged observation, and repeat sampling of specific individual hagfish still seems like the easiest way to obtain meaningful data.

I do see why no one has run this experiment yet, though I think it would be valuable to do so. Given the importance of hagfish to the ecology of nearly every ocean on the planet (see this post for a little more on that), the fact that there is an ongoing commercial demand for their flesh, and the fact that many quirks of their biology may render them vulnerable to population collapse in unique ways, learning more about hagfish sex isn’t something that we should give up on.

Sources:

Brownstein, C. D., & Near, T. J. (2024). Colonization of the ocean floor by jawless vertebrates across three mass extinctions. BMC ecology and evolution, 24(1), 79. https://doi.org/10.1186/s12862-024-02253-y

Cole, K. Gonad morphology, sexual development, and colony composition in the obligate coral-dwelling damselfish Dascyllus aruanus. Marine Biology 140, 151–163 (2002). https://doi.org/10.1007/s002270100681

Davis, J., Meservey, S., Agulay, A., Wishinski, J., & Macnevin, L. (2001). Sexuality And Embryogenesis Of The Atlantic Hagfish," Myxine Glutinosa: SEAH". https://repository.library.noaa.gov/view/noaa/46195/noaa_46195_DS1.pdf

Fleury, A. G., MacLennan, E. M., Command, R. J., & Juanes, F. (2021). Reproductive biology and ecology of Pacific hagfish (Eptatretus stoutii) and black hagfish (Eptatretus deani). Journal of fish biology, 99(2), 596-606. https://doi.org/10.1111/jfb.14748

Gorbman, A. (1990). Sex differentiation in the hagfish Eptatretus stouti. General and comparative endocrinology, 77(2), 309-323. https://doi.org/10.1016/0016-6480(90)90315-D

Kavanaugh, S. I., Powell, M. L., & Sower, S. A. (2005). Seasonal changes of gonadotropin-releasing hormone in the Atlantic hagfish Myxine glutinosa. General and comparative endocrinology, 140(2), 136-143. https://doi.org/10.1016/j.ygcen.2004.10.015

Kuraku, S., & Kuratani, S. (2006). Time scale for cyclostome evolution inferred with a phylogenetic diagnosis of hagfish and lamprey cDNA sequences. Zoological science, 23(12), 1053-1064. https://doi.org/10.2108/zsj.23.1053

Langston, R. (2023). Histological evidence of sequential hermaphroditism in Hawaiian sandburrowers Crystallodytes cookei and Limnichthys nitidus. Environmental Biology of Fishes, 106(1), 61-78. https://doi.org/10.1007/s10641-022-01373-y

Marlétaz, F., Timoshevskaya, N., Timoshevskiy, V. A., Parey, E., Simakov, O., Gavriouchkina, D., Suzuki, M., Kubokawa, K., Brenner, S., Smith, J. J., & Rokhsar, D. S. (2024). The hagfish genome and the evolution of vertebrates. Nature, 627(8005), 811–820. https://doi.org/10.1038/s41586-024-07070-3

Martini, F. H., & Beulig, A. (2013). Morphometics and gonadal development of the hagfish Eptatretus cirrhatus in New Zealand. PLoS One, 8(11), e78740. https://doi.org/10.1371/journal.pone.0078740

Maxfield, J. M., & Cole, K. S. (2019). Structural changes in the ovotestis of the bidirectional hermaphrodite, the blue-banded goby (Lythrypnus dalli), during transition from ova production to sperm production. Environmental biology of fishes, 102(11), 1393-1404. https://doi.org/10.1007/s10641-019-00914-2

Muramatsu, B., Suzuki, D. G., Suzuki, M., & Higashiyama, H. (2024). Gross anatomy of the Pacific hagfish, Eptatretus burgeri, with special reference to the coelomic viscera. The Anatomical Record, 307(1), 155-171. https://doi.org/10.1002/ar.25208

Nozaki, M., Ichikawa, T., Tsuneki, K., & Kobayashi, H. (2000). Seasonal development of gonads of the hagfish, Eptatretus burgeri, correlated with their seasonal migration. Zoological Science, 17(2), 225-232. https://doi.org/10.2108/zsj.17.225

Ota, K. G., & Kuratani, S. (2006). The history of scientific endeavors towards understanding hagfish embryology. Zoological Science, 23(5), 403-418. https://doi.org/10.2108/zsj.23.403

Ota, K. G., Kuraku, S., & Kuratani, S. (2007). Hagfish embryology with reference to the evolution of the neural crest. Nature, 446(7136), 672–675. https://doi.org/10.1038/nature05633

Pla, S., Maynou, F. & Piferrer, F. Hermaphroditism in fish: incidence, distribution and associations with abiotic environmental factors. Rev Fish Biol Fisheries 31, 935–955 (2021). https://doi.org/10.1007/s11160-021-09681-9

Powell, M. L., Kavanaugh, S. I., & Sower, S. A. (2004). Seasonal concentrations of reproductive steroids in the gonads of the Atlantic hagfish, Myxine glutinosa. Journal of Experimental Zoology Part A: Comparative Experimental Biology, 301(4), 352-360. https://doi.org/10.1002/jez.a.20043

Sower, S. A., & Baron, M. P. (2011). The interrelationship of estrogen receptor and GnRH in a Basal vertebrate, the sea lamprey. Frontiers in endocrinology, 2, 58. https://doi.org/10.3389/fendo.2011.00058

Weinrauch, A. M., Edwards, S. L., & Goss, G. G. (2015). Anatomy of the Pacific hagfish (Eptatretus stoutii). Hagfish Biology; CRC Press: Boca Raton, FL, USA, 1-39. https://www.researchgate.net/profile/Greg-Goss/publication/281845044_Anatomy_of_the_Pacific_Hagfish_Epatatretus_stoutii/links/611b04d10c2bfa282a4d8d94/Anatomy-of-the-Pacific-Hagfish-Epatatretus-stoutii.pdf (this is a direct pdf download)

Yamaguchi, Y. (2025). Growth, Feeding, and Age of the Inshore Hagfish, Eptatretus burgeri. Zoological science, 42(3). https://doi.org/10.2108/zs240097