About the SA Blog Network

Tetrapod Zoology

Tetrapod Zoology

Amphibians, reptiles, birds and mammals - living and extinct
Tetrapod Zoology Home

Turtles that eat bone, rocks and soil, and turtles that mine

The views expressed are those of the author and are not necessarily those of Scientific American.

Email   PrintPrint

In which Eastie consumes a rat cranium. Image by Mathew Wedel, used with permission.

I have had extensive turtle guilt of late – that is, there just haven’t been enough turtles on Tet Zoo for a while… by which I mean, there haven’t been any. But fortune smiles on the thing that it smiles on, and a few neat photos have recently fallen into my proverbial lap, encouraging me to write the article you’re reading now. My huge friend and colleague Mathew Wedel owns a Box turtle Terrapene carolina. It’s called Eastie… don’t judge; this is because the animal is an Eastern box turtle (or is she? I wonder if Eastie is a Three-toed box turtle). Anyway, Eastie recently found part of a deceased rat’s head while on a backyard jaunt, and proceeded to deliberately snip away at the broken braincase and eat the bone fragments. This bone-eating carried on for about 20 minutes, and Matt thought it interesting enough to take the photos you see here.

Aldabran tortoise chows down on Aldabran tortoise - mm-mm, they's good eatin'! Image by David Attenborough, from Attenborough (1984).

The eating of bones – osteophagy – is well known for turtles, has been recorded in several species, and is observed easily enough in species kept in captivity (like Testudo tortoises). Whenever this subject is mentioned (believe me, it’s always cropping up in conversation), many people recall the photo in David Attenborough’s Life on Earth that shows an Aldabran giant tortoise Aldabrachelys gigantea* scavenging on the carcass of a conspecific (Attenborough 1984). As you can see here, it’s not entirely clear what the tortoise is doing, but it looks like it’s gnawing at dried skin and muscle, not bone. Incidentally, the photo was taken by Attenborough himself. I did used to have a very neat photo showing gnaw marks that a pet tortoise (belonging to my late friend and colleague David Cooper) left on a cow bone – to my frustration, I can no longer locate it.

* Yup. No Dipsochelys dussimieri for Tet Zoo, thank you very much. Did you follow the debate in the ICZN Bulletin?

More Eastie vs rat skull. Image by Mathew Wedel, used with permission.

Why are these turtles eating, and gnawing at, bone? Turtles seem to have high calcium requirements, with some studies indicating that they require unusually high quantities in order for shell and skeletal development to occur normally (Fledelius et al. 2005). In part, this is probably due to the extensive amount of extra bone they grow for their shells, but females also need extra amounts of calcium when forming eggs and growing eggshells: data from female musk turtles at least shows that skeletal bone density drops markedly during the time that eggshells are being formed (Edgren 1960).

The impression you get from what’s been published is that turtles are especially ‘calcium hungry’ relative to other reptiles (errr.. you want this quantified? Hey, I’ll get back to you on that). And it seems that they recognise that bone is a valuable calcium source: they’ll not only eat small bones and nibble on large ones, but will also eat bone fragments out of dried scat belonging to big cats and other carnivorans (Bally 1946, Loveridge & Williams 1957). They can also be persistent and determined to eat the bones that capture their interest, sometimes returning to eat or gnaw at a bone just seconds after the observing humans have moved away (Esque & Peters 1994). In contrast, they tend not to immediately return to the plants they’re eating (Esque & Peters 1994), a fact which suggests that bone is an especially valuable resource that needs to be taken advantage of as quickly as possible.

The turtle we're mostly talking about here: the beautiful, feisty Desert tortoise (this is presumably an Agassiz's desert tortoise). Image by Utahcamera, in public domain.

Furthermore, tortoises are sometimes so keen to ingest calcium that they’ll swallow stones and even take mouthfuls of calcium-rich soils. This has been documented for wild Desert tortoises Gopherus agassizii* in the Mojave Desert (Esque & Peters 1994) and also in captive box turtles (Kramer 1973). Esque & Peters (1994) described 12 observations of soil- or stone-eating in Desert tortoises. This was out of a total of 991 observations of foraging tortoises, so it certainly doesn’t seem that these ingestion events are frequent enough to be easily observed in the wild. The geophagous animals were mostly adult females, mostly indulging in the behaviour during the spring.

* Note that the Desert tortoise has recently been split into two species (with there being suspicions that yet further taxa await recognition within the complex). The animals discussed here almost certainly still belong, however, to G. agassizii. The other one is the Morafka desert tortoise G. morafkai, found east of the Colorado River.

Map showing movement patterns of a Desert tortoise: the individual mostly moved in between the different sleeping burrows shown at the bottom as dark circles, but - on one day - took a special, very deliberate trip to 'M', one of the tortoise mines depicted as triangles. Image from Marlow & Tollestrup (1982).

Despite the rarity of stone-eating observations, about 60% of 185 x-rayed wild Desert tortoises contained stones within their intestines, indicating that stone-eating is normal and common in some species. Esque & Peters (1994) suggested that this behaviour might help maintain a healthy gut pH, help nullify toxic plant compounds, control intestinal parasites, or assist in maintaining beak shape. They also made the interesting point that people should probably avoid removing bones from places inhabited by Desert tortoises – by doing so, we might be altering the pattern of nutrient cycling, and literally depriving tortoises of rare resources that they need.

The ingestion of stones and bones is pretty neat, but perhaps even more interesting is the mining behaviour also practised by Desert tortoises. Marlow & Tollestrup (1982) described how members of some Desert tortoise populations deliberately mine and eat specific layers of soil at key locations that they visit at regular intervals. Their activities mean that these locations are maintained for years as special mines (Marlow & Tollestrup 1982). Yes, mining tortoises.

Maybe this would make a good caption competition. Image by Jason Noble, used with permission.

Finally – in one more effort to partially relieve the turtle guilt, I leave you with these really neat and charismatic photos of a wild Common snapping turtle Chelyda serpentina that Jason Noble took at Catoctin Wildlife Preserve and Zoo, Maryland. I just love the turtle’s expression.

For previous Tet Zoo articles on turtles, see…

UPDATE: Stewart Macdonald kindly supplied the following photos, showing a site in Arizona (a dry creek bed) used as a mine by Desert tortoises. In the upper photo, cavities on the sides of the creek were being used by mining tortoises when the photo was taken (one of these cavities is visible high up on the side of the creek wall). There’s some suggestion that the overhang visible in the photo at the bottom might have been caused by the sediment-eating activities of the visiting tortoises, but this isn’t confirmed. Thanks to Stewart for the permission to use these images.

Refs – -

Attenborough, D. 1984. The Living Planet. Collins, London.

Bally, P. 1946. Tortoises eating bones. East African Uganda Natural History Society 18, 163.

Edgren, R. A. 1960. A seasonal change in bone density in female musk turtles, Sternothaerus odoratus (Latreille). Comparative Biochemistry and Physiology 1, 213-217.

Esque, T. C. & Peters, E. L. 1994. Ingestion of bones, stones, and soil by desert tortoises. Fish and Wildlife Research 13, 105-111.

Fledelius, B., Jørgensen, G. W., Jensen, H. E. & Brimer, L. 2005. Influence of the calcium content of the diet offered to leopard tortoises (Geochelone pardalis). The Veterinary Record 156, 831-835.

Kramer, D. C. 1973. Geophagy in Terrapene omata Agas­siz. Journal of Herpetology 7, 138-139.

Loveridge, A. & Williams, E. 1957. Revision of the African tortoises and turtles of the suborder Cryptodira. Bulletin of the Museum of Comparative Zoology 115, 163-557.

Marlow, R. W. & Tollestrup, K. 1982. Mining and exploitation of natural mineral deposits by the desert tortoise, Gopherus agassizii. Animal Behaviour 30, 475-478.

Darren Naish About the Author: Darren Naish is a science writer, technical editor and palaeozoologist (affiliated with the University of Southampton, UK). He mostly works on Cretaceous dinosaurs and pterosaurs but has an avid interest in all things tetrapod. His publications can be downloaded at He has been blogging at Tetrapod Zoology since 2006. Check out the Tet Zoo podcast at! Follow on Twitter @TetZoo.

The views expressed are those of the author and are not necessarily those of Scientific American.

Rights & Permissions

Comments 35 Comments

Add Comment
  1. 1. Chabier G. 5:59 am 04/11/2014

    Ah, I really like the cute Chelydra serpentina, some years ago we receive one of them at the Rescue Center, It had been caught in the Ebro river shore, and this capture was indeed a cause of concern, as we don’t know how many snapping tortoises could be roaming free out there,and they are even more dangerous for native fauna than the ubiquitous Trachemys scripta.
    But it was a beauty, these long limbs, this long neck always ready to stretch and bite, and the dragonlike back horned tail. In spite of its 9 kg weight, it could climb more than 1 metre on a steel net fence,and it grasped and swallowed big crayfish as if they were popcorns. Cool creature.

    Link to this
  2. 2. Cameron McCormick 6:44 am 04/11/2014

    I have a juvenile Painted Turtle that, in addition to chewing on cattle bones, will eat entire mussel and crab shells. I haven’t been able to observe the behavior in adults, so perhaps they’re not as ‘calcium hungry’.

    Feel free to keep atoning for your turtle guilt!

    Link to this
  3. 3. John Harshman 11:22 am 04/11/2014

    I for one welcome our new archosauromorph overlords.

    Link to this
  4. 4. SciaticPain 1:37 am 04/12/2014

    When I read of bone gnawing turtles I am reminded of the often noted abundant turtle remains from the Mesozoic. I can’t place it but I remember somewhere someone saying that the Cretaceous should be called the age of turtles.

    Link to this
  5. 5. vdinets 3:19 am 04/12/2014

    A long time ago I observed a turtle (Testudo sp., I’ve lost track of recent classification changes) eating small snails off grass blades. It was in Utrish Nature Reserve on Russia’s Black Sea coast. At the time I thought it was after the meat, but now I think it was probably more interested in shells… although there was plenty of empty shells on the ground, so possibly both.

    Link to this
  6. 6. Chelydra 8:41 am 04/12/2014

    Eastie does look like a three-toed box turtle. The patternless tan shell is typical, as are the orange and yellow head markings.

    Link to this
  7. 7. Varanussalvator 9:51 am 04/12/2014

    In recent news, Zootaxa has a paper describing 2 new species of alligator snapping turtles: Macrochelys apalachicolae is from eastern Alabama, western Georgia and northwestern Floria, Macrochelys suwanniensis from the Suwannee River drainage in northern Florida and Georgia, with Macrochelys temminckii now being restricted to the populations in western Alabama, Mississippi and Louisiana.

    Link to this
  8. 8. Heteromeles 10:24 am 04/12/2014

    So what, if anything, does that do to alligator snapping turtle conservation? Are all the new species now potentially threatened by the inherent red-state nature of the region?

    Link to this
  9. 9. Tayo Bethel 11:32 am 04/12/2014

    What are the defining characteristics of these new species of alligator snappers? Can a layman distinguish between the species,or is this another case of differences so subtle that only a trained observer can distinguish species?

    Link to this
  10. 10. Varanussalvator 1:12 pm 04/12/2014

    Heteromeles: The paper does discuss conservation implications. Basically, the population status of each species and state management strategies have to be reviewed, since they would all have previously been lumped under a single species. This is especially important in states like Alabama, Georgia, and Florida, where two or more alligator snapping turtle species occur (though not in sympatry). Movements of alligator snapping turtles across state boundaries and releases mean that there is a risk of hybridisation if individuals from one species of alligator snapping turtle end up in river systems inhabited by another species.

    Link to this
  11. 11. Varanussalvator 1:23 pm 04/12/2014

    Tayo Bethel: Good question. The paper uses the shape of the squamosal and notches in the edge of the carapace to differentiate the species. I don’t think these characters are so readily visible in living specimens, especially since alligator snapping turtles don’t quite lend themselves to being easily examined at close range. I suppose the easiest way is to know the river system and corresponding alligator snapping turtle species that’s supposed to inhabit that system. (Provided you’re talking about finding alligator snapping turtles in their native range, rather than released pets)

    Link to this
  12. 12. Heteromeles 12:03 am 04/13/2014

    Thanks. At this point, it’s probably possible to develop a DNA-based dipstick test where the test material will turn color based on the presence of DNA shed by the various snapping turtles in different rivers. That might be easier for sampling, if someone could figure out how to quantify turtles per m3 of river upstream, based on the intensity of the response.

    Link to this
  13. 13. DavidMarjanovic 9:27 am 04/13/2014

    Did you follow the debate in the ICZN Bulletin?

    No. I can has link plz? :-)


    We might need a celebrity deathmatch here. :-) The latest Gauthier-lab paper hasn’t come out yet, though.

    Link to this
  14. 14. Cameron McCormick 12:24 pm 04/13/2014

    So if there really are three Alligator Snapping Turtle species that last shared a common ancestor in the Miocene, I wonder if there are some soft tissue differences that have gone unnoticed. Some Suwanee River Macrochelys captured in this survey seem to have lots of yellow or orangish coloration, particularly on the upper beak. Of course, given the tendency for these turtles to get coated with all sorts of interesting growths, perhaps coloration wouldn’t be a great field mark.

    Link to this
  15. 15. blakemarkwell 1:19 pm 04/14/2014

    I agree with Chelydra that Eastie appears to be T. carolina triunguis and not T. c. carolina.

    Link to this
  16. 16. WarrenJB 12:05 pm 04/15/2014

    I was on a trip to London, including London Zoo, this weekend, and a walk round the Rainforest Life exhibit brought this post rushing back. Something left a particularly wet splat on a log, and one of the tortoises (red-footed tortoise Chelonoidis carbonaria?) was having a right go at it.

    Apols for the poor quality. (Nokia Lumia 1020 doesn’t live up to all the hype…) I hazard it wasn’t taking in calcium, but I don’t know what it was getting.

    Link to this
  17. 17. irenedelse 3:24 am 04/16/2014

    I wonder about the ecological and geological implications of “minerals-hungry” turtles. Over time, a population of mining, earth-eating turtles would cause a substantial amount of erosion, surely?

    Link to this
  18. 18. Allen Hazen 12:56 am 04/17/2014

    Seriously off-topic (there being no obvious connection between tortoises and hares), but…
    A long time ago, in Ver. 1, the “wrongness” of rabbits was discussed, and mention was made of the question, “Do rabbits exist?” Well, I’ve just found evidence for a NEGATIVE answer!

    Chris Taylor’s “Variety of Life” covered some rabbity (and also some squirrelly) stuff a few weeks ago, and on the phylogeny shown at

    Sylvilagus (the “Cottontails” of Eastern North America) are outside a clade containing Oryctolagus (European rabbits, including domestic forms) and Lepus (hares, including North American “jackrabbits”). Now, at least on the usual American usage, Sylvilagus and Oryctolagus are both “rabbits” and not hares. So, on this usage, “rabbit” designates a paraphyletic assemblage and not a clade. But many proponents of phylogenetic nomenclature make a point of refusing to use “taxon”-names that do not designate clades: to quote G*r*th N*ls*n, “there is no such thing as a reptile” (since any cladeincluding both lizards and turtles will also include birds). So: “there is no such thing as a rabbit.”

    O.k., now that’s off my chest…

    Link to this
  19. 19. Tayo Bethel 4:04 am 04/17/2014

    No such thing as a rabbit?

    I thought rabbit referred to an adaptive form of the family Leporidae, not a specific clade. I had no idea that Sylvillagus and Oryctolaguswere not close relatives, but that’s not altogether surprising–Oryctolagus and Sylvillagus are quite different behaviorally. Oryctolagus and Lepus forming a clade? That was a surprise, though.

    Link to this
  20. 20. vdinets 1:18 pm 04/17/2014

    If we started dropping common names for poly – or paraphyletic assemblages, there would be no whales, dolphins, mice, dogs, wolves, or fish.

    Link to this
  21. 21. irenedelse 2:31 pm 04/17/2014

    But isn’t the problem with popular names like rabbits, or reptiles, more with popularization of science, or basic education? Phylogenetic classification still hasn’t much penetrated the general culture, except maybe for the birds-as-dinosaurs meme…

    Link to this
  22. 22. vdinets 4:00 pm 04/17/2014

    irendelse: I would say it’s a problem with learning science in general, not just popularization. Cladistics-based classification is less intuitive and more difficult to learn/remember than traditional perceived similarity-based classification.

    Link to this
  23. 23. Allen Hazen 1:44 am 04/18/2014

    Tayo Bethel–
    “Adaptive form.” That has possibilities. Do you think we could get people to agree to keep the name “reptile,” as a term for an adaptive form (characterized primarily by featherlessness) of the clade Sauropsida?

    Link to this
  24. 24. DavidMarjanovic 7:39 am 04/18/2014

    Frankly, the less (and the fewer) people think of “reptiles” as having anything exclusive in common, the better.

    Link to this
  25. 25. kattatogaru 8:39 am 04/18/2014

    Darren, do soft-shelled turtles eat bone too? and more importantly, to fulfill a long-standing promise which dates back to your article on Asian soft-shelled turtles, can you please elaborate as promised on the noble tradition of using Oriental ladies as scale bars for assorted large animals?

    Link to this
  26. 26. Tayo Bethel 7:34 pm 04/18/2014

    @Allen Hazen

    If i could rewrite the dictionary, I would eliminate the word reptile from the English language. Far better to find a common name for Archosauria,another one for Sauropterygians, etc. Since that’s not likely to happen, I prefer to avoid using the word reptile unless absolutely necessary.
    As for adaptive form–sorry, but that’s probably not really a possibility. Convergent evolution, y’see, is still a force in the world. There is no reason, for example, why a genus of the lagomorphs that we call rabbits could not adapt to live in the open and evolve toward a Lepus-like condition. That genus of rabbit would then be called a hare, despite the fact that all “true” hares belong to the genus Lepus.

    Link to this
  27. 27. DavidMarjanovic 4:56 pm 04/19/2014

    the noble tradition of using Oriental ladies as scale bars for assorted large animals

    Inaccurate scale bars, notably!

    Link to this
  28. 28. naishd 8:17 pm 04/20/2014

    Thanks for comments, everyone – I’ve been away, hence silence and non-participation. Some new images have just been added to the article showing AN ACTUAL TURTLE MINE.

    Link to this
  29. 29. irenedelse 11:40 am 04/21/2014

    Tayo Bethel:
    I like it when zoos and museums (and other such institutions involved in outreach) use Sauropsids instead of “Reptiles” or “Birds” to classify exhibits. What a change from the zoology class I took in 1991, when we learned that Tetrapods were divided in Amphibians, Reptiles, Birds and Mammals…

    Link to this
  30. 30. Tayo Bethel 1:02 pm 04/21/2014

    The next step is to introduce a simplified way of learning cladistics to elementary science classes.When tha happens we can finally do away with the outdated Linnaean system that has plagued the general public understanding for so long.

    Link to this
  31. 31. DavidMarjanovic 5:40 pm 04/21/2014

    The next step is to introduce a simplified way of learning cladistics to elementary science classes.

    Uh. Dr. Evil? That already exists.

    Link to this
  32. 32. Tayo Bethel 12:31 am 04/22/2014

    DavidMarjanovic :

    :) A bit behind the times–I was taught in a school funded by a church–need I say more? In a country called a Christian nation, where church attendance is the mark of the upright citizen.

    Link to this
  33. 33. Dartian 2:35 am 04/22/2014

    Sylvilagus (the “Cottontails” of Eastern North America) are outside a clade containing Oryctolagus (European rabbits, including domestic forms) and Lepus (hares, including North American “jackrabbits”).

    That dendrogram you’re referring to seems to be based on Averianov (1999) (which, unfortunately, I don’t have access to at the moment). However, more recent studies on lagomorph relationships have arrived at different phylogenetic relationships. Notably, Matthee et al. (2004) and Robinson & Matthee (2005) found that Lepus is sister to a clade that includes both Oryctolagus and Sylvilagus – as well as a few other ‘rabbits’.

    However, as some other ‘rabbits’ are even more distantly related to the above-mentioned lagomorph taxa, it’s still true that ‘rabbits’ are paraphyletic…

    Averianov, A.O. 1999. Phylogeny and classification of Leporidae (Mammalia, Lagomorpha). Vestnik Zoologii 33, 41-48.

    Matthee, C.A., Jansen van Vuuren, B., Bell, D. & Robinson, T.J. 2004. A molecular supermatrix of the rabbits and hares (Leporidae) allows for the identification of five intercontinental exchanges during the Miocene. Systematic Biology 53, 433-447.

    Robinson, T.J. & Matthee, C.A. 2005. Phylogeny and evolutionary origins of the Leporidae: a review of cytogenetics, molecular analyses and a supermatrix analysis. Mammal Review 35, 231–247.

    Link to this
  34. 34. Jerzy v. 3.0. 6:28 am 04/22/2014

    Happy Easter (belated), almost every single common name is paraphyletic.

    BTW, anybody knows why tinamous and some other ratities lay colorful eggs?

    Link to this
  35. 35. DavidMarjanovic 10:31 am 04/24/2014

    Lots of songbirds lay colorful eggs, too.

    Melanin is sometimes used to stiffen the eggshell when there’s a lack of calcium.

    Link to this

Add a Comment
You must sign in or register as a member to submit a comment.

More from Scientific American

Email this Article