Bee Ferromagnetism Is a Material Memory Substrate for Local Magnetic Orientation

Sterling Geisel, QBist Lab; Dr. Hideo Tanaka

Abstract

Russo et al. report ferromagnetic hysteresis across most sampled bee species and many non-bee insects, with no phylogenetic signal and with magnetic strength structured by body size, sex, nesting habit, and social behavior. Pudding Theory reads this distribution through Material Memory: repeated exposure to geomagnetic, dietary, and behavioral fields leaves persistent material traces in ferrimagnetic nanoparticles, and those traces bias future orientation probability. The source paper treats saturation magnetization, coercivity, and remanence as evidence for possible magnetoreceptive tissue. The Pudding Theory reading makes a stronger claim. The nanoparticles are not merely sensors added to an organism. They are retained field history. Their remanent order is the physical storage of repeated environmental signal, stabilized in matter and recruited by navigation. The absence of phylogenetic signal is therefore not a failure of evolutionary pattern but the signature of a substrate built by repeated reception. If body-part remanence anisotropy were measured to be indistinguishable from random environmental contamination after matched diet and habitat controls, this Postulate would be falsified.

Source Synopsis

Russo, Allen, Jorgensen, Quigley, Buchanan, Winklhofer, Brady, Packer, Murray, and Gilbert examine whether ferromagnetic material is broadly present in bees and whether its distribution tracks phylogeny, social behavior, nesting habit, sex, or body size. Their study measures room-temperature magnetic hysteresis loops in 185 insect specimens, including 138 bees from 96 species across six families and 47 non-bee insects as outgroups. They classify specimens by the presence of sigmoidal or hysteretic magnetization curves, by saturation magnetization, remanence, coercivity, and by comparison with thresholds derived from bee species where magnetoreceptive behavior has been reported.

The main empirical result is broad presence. Ferromagnetic response appears in 121 of 138 bee specimens and in 36 of 47 outgroup insects. Using saturation magnetization and mass-specific saturation magnetization thresholds, most bees qualify as putatively magnetoreceptive. This breadth includes eusocial, solitary, subsocial, semisocial, cleptoparasitic, ground-nesting, cavity-nesting, and stem-nesting species. The authors find no meaningful phylogenetic signal in coercivity, mass-specific saturation magnetization, or squareness ratio. Blomberg’s K values fall below Brownian expectations, and phylogenetic generalized least-squares models do not change the qualitative results.

The source paper also reports structured variation. Female workers or solitary females show higher coercivity than males. Cavity-nesting bees show higher mass-specific saturation magnetization than ground-, stem-, or open-nesting bees. Eusocial bees show higher coercivity than solitary bees, while solitary bees show higher squareness ratio. Body size correlates positively with saturation magnetization, coercivity, and squareness ratio.

Anatomical measurements complicate a simple receptor-localization model. In 14 dissected female bees, magnetic signal is usually distributed across multiple tagmata. The mesosoma tends to contain the highest percentage of saturation magnetization, while the head has lower coercivity and lower squareness ratio. The source paper argues that contamination is unlikely to explain the pattern because authentic zeros occur, ground-nesting bees are not most magnetic, magnetic strength increases with mass-specific size trends inconsistent with surface contamination, and body parts show distinct magnetic behavior.

The authors conclude that ferromagnetism, and likely magnetoreceptive potential, is common across bees and not restricted to eusocial taxa. They do not provide behavioral proof of magnetoreception, but they argue that the observed magnetic tissues are consistent with magnetite-based mechanisms and deserve direct behavioral testing.

Postulate Lens

This reading applies Material Memory. The source phenomenon already has the structure this Postulate names: matter retains magnetic order after exposure to a field, and that retained order biases later response. In the source paper, remanence is explicitly the magnetization remaining at zero applied field after saturation. Coercivity measures the field required to erase that retained orientation. These are not incidental material descriptors. They are the memory variables.

Material Memory is the right lens because the paper’s strongest pattern is not a clean phylogenetic tree, a single sensory organ, or a single social behavior. It is persistent magnetic trace distributed across bodies and taxa, with trait-dependent strength. That is the expected form of a material memory substrate. The biological body does not only detect the geomagnetic field at the moment of navigation. It stores a history of repeated field exposure in ferrimagnetic grains. The stored trace then changes the probability landscape for orientation.

Pudding Theory Reading

Russo et al. frame ferromagnetism as evidence for putative magnetoreceptive tissue. Pudding Theory accepts the measurement but shifts the ontology. The ferrimagnetic particles are not passive pieces of hardware waiting for a magnetic field to torque them. They are biological records of repeated signal. The bee body is a distributed archive of geomagnetic contact, dietary iron processing, motion through nest geometry, and recurrent foraging trajectories. Its magnetic response is a stored relation between organism and field.

The key observables are saturation magnetization, remanence, coercivity, and squareness ratio. In a standard materials reading, these describe particle abundance, size, domain state, and anisotropy. In the Material Memory reading, they also describe the depth, stability, and directional accessibility of retained signal. Saturation magnetization gives the available magnetic storage capacity. Remanence gives the persistence of prior alignment. Coercivity gives the resistance of that stored alignment to erasure. Squareness ratio gives the degree to which past field exposure remains usable as a directional bias rather than fading into induced response.

This reading explains why the lack of phylogenetic signal matters. If ferrimagnetic magnetoreception were primarily a lineage-specific organ system, magnetic strength should track descent more strongly. Russo et al. find the opposite. Magnetic properties cut across bee families and appear in non-bee insects. Pudding Theory reads that distribution as evidence that the substrate is older and more general than the bee clade, but also more plastic than a fixed organ. The memory-bearing material is repeatedly built, erased, enriched, or reorganized by local life history.

The source paper treats environmental contamination as a possible confound and argues against it. Pudding Theory sharpens the distinction. Random contamination is not memory because it does not carry organism-specific structure. A material memory trace must be patterned by anatomy, diet, behavior, and repeated field coupling. The reported body-part differences are therefore central. A surface contaminant should scale with exposure and surface area. A memory substrate should scale with where the organism repeatedly couples magnetic material to action. The stronger mesosoma and metasoma signals, and the weaker head signal, point away from a single head-based compass and toward a whole-body material history that can still support orientation.

This also changes the interpretation of sociality. The source finds stronger coercivity in eusocial bees but higher squareness ratio in solitary bees. A simple sociality hypothesis cannot absorb both results. Material Memory can. Eusocial nesting supplies repeated shared field geometries, high traffic, and stable return paths, which favor coercive stability. Solitary navigation may require more individually retained directional contrast, giving higher squareness. Sociality is not the cause of magnetism. It is one regime of repeated signal exposure that sculpts the magnetic archive.

The Pudding Theory claim is therefore concrete. Bee magnetism is a retained environmental signal embodied in ferrimagnetic matter. Magnetoreception is not only momentary sensing of Earth’s field. It is the use of stored field history to bias future orientation under ambiguous local cues.

Falsifiable Observable

The distinguishing observable is body-part remanence anisotropy after controlled rearing. Bees reared on matched iron diets in matched geomagnetic environments should still show behavior-linked remanence structure if Material Memory is correct, especially in tagmata coupled to recurrent locomotor and nesting routines. Random contamination or a purely fixed organ model should not preserve such structured remanence across controlled histories. If body-part remanence anisotropy were measured to be indistinguishable from random environmental contamination after matched diet and habitat controls, this Postulate would be falsified.

Editorial Dialogue

Tanaka: The reading risks turning ordinary hysteresis into metaphysics. Remanence is already memory in the narrow materials sense. Magnetite particles retain magnetization. That does not imply that a bee stores navigational history in them. The source paper measures dead specimens, not orientation behavior. It also admits that ferromagnetic signal alone cannot prove magnetoreception.

Geisel: The point is not that hysteresis alone proves behavior. The point is that the measured variables are exactly the variables a material memory account would require. Russo et al. do not find a clean phylogenetic organ. They find broad, distributed, trait-structured ferromagnetism with anatomical heterogeneity. A dead specimen can still carry the stored magnetic state of the living system, just as bone carries load history and otolith chemistry carries habitat history.

Tanaka: But diet, age, and soil exposure could generate the same variation.

Geisel: They could generate variation, but not any variation counts. The reading predicts structured retention after controls: body-part anisotropy, relation to repeated behavioral coupling, and stability under field reversal according to coercivity. Contamination should wash out under matched rearing and should scale with exposure surfaces. The source already reports patterns inconsistent with simple surface contamination. The theoretical claim is that those patterns are not background. They are the phenomenon.

Discussion

This reading buys a different account of why bee ferromagnetism is broad, uneven, and anatomically distributed. The source paper shows that magnetic material is common and may support magnetoreception, but its own results strain a single-organ or sociality-first story. Pudding Theory makes the distribution intelligible as retained signal. The relevant unit is not a bee family or a receptor location. It is the history of repeated coupling between ferrimagnetic matter, geomagnetic field, diet, movement, and nest return.

The limitation is clear. Russo et al. measure magnetic properties, not live navigational decisions. A Material Memory account therefore requires controlled behavioral and magnetic rearing experiments. It would be weakened if controlled bees retained only random or contamination-like magnetic signatures. It would be strengthened if remanence and coercivity predicted orientation under cue conflict better than saturation magnetization alone.

The conclusion changes the research program. Instead of asking only whether bees possess magnetite sensors, the next question is whether their magnetic tissue stores a usable history of field interaction. That is a sharper biological claim and a sharper physical one.

References

1. Russo, L., Allen, C., Jorgensen, C. S., Quigley, L., Buchanan, C. C., Winklhofer, M., Brady, S. G., Packer, L., Murray, A., & Gilbert, D. A. “Broad presence of ferromagnetism in bees and relationship to phylogeny, natural history, and sociality.” arXiv:2603.20312. DOI: doi:10.48550/arxiv.2603.20312.

2. Geisel, S. “Pudding Theory: A Topological Theory of Information Fields.” QBist Lab Working Papers, September 10, 2025.

3. Gould, J. L., Kirschvink, J. L., & Deffeyes, K. S. “Bees Have Magnetic Remanence.” Science 201, 1026–1028 (1978).

4. Hsu, C.-Y., & Li, C.-W. “Magnetoreception in Honeybees.” Science 265, 95–97 (1994).

5. Kirschvink, J. L., Walker, M. M., & Diebel, C. E. “Magnetite-based magnetoreception.” Current Opinion in Neurobiology 11, 462–467 (2001).

6. Wajnberg, E., Acosta-Avalos, D., Alves, O. C., de Oliveira, J. F., Srygley, R. B., & Esquivel, D. M. S. “Magnetoreception in eusocial insects: an update.” Journal of The Royal Society Interface 7, S207–S225 (2010).

7. Winklhofer, M., & Kirschvink, J. L. “A quantitative assessment of torque-transducer models for magnetoreception.” Journal of The Royal Society Interface 7, S273–S289 (2010).

8. Shaw, J., Boyd, A., House, M., Woodward, R., Mathes, F., Cowin, G., Saunders, M., & Baer, B. “Magnetic particle-mediated magnetoreception.” Journal of The Royal Society Interface 12, 20150499 (2015).