The water chevrotain: almost like a whale

July 7, 2009 • 7:48 am

In Darwin’s time, the fossil record was far spottier than it is now, and no transitional forms were known. (Darwin does mention the transitional “bird” Archaeopteryx in a later edition of The Origin, but didn’t realize its significance.)  Thus, the evolutionary origin of new “types” of animals and plants was largely a matter of guesswork.

Speculating in The Origin on the evolution of whales, for example, Darwin said this about their possible ancestors: “In North America, the black bear was seen by Hearne swimming for hours with widely open mouth, thus catching, almost like a whale, insects in the water.” Well, Darwin took a lot of flak for this, as the image of a whale evolving from a swimming bear seemed, to many, ludicrous.  Well, the real situation may sound just as ludicrous, but has the merit of being supported by evidence.  Whales evolved from artiodactyls (even-toed animals like deer), in particular small artiodactyls like the recently described Indohyus.  The whole scenario is described in my book, and I won’t recount it here.


Reconstruction of Indohyus by Carl Buell

Indohyus, like some other early artiodactyls, has several features demonstrating an artiodactyl ancestry for the earliest whales.  But now we can see a living form that may be very like Indohyus in size, form, and behavior. This is the water chevrotain. The BBC reports today on a new paper in the journal Mammalian Biology that describes swimming behavior in several species of  water chevrotain (Moschiola spp.).  It is this type of aquatic behavior that could have prompted the transition from dry land to water, as natural selection took a terrestrial animal and made it more and more adapted to aquatic habitats.

If you’ve read WEIT, you’ll know that the chevotain is a small deerlike mammal, with species in both Asia and Africa, that is also called the “mouse deer.”


Water chevrotain (photo by Nick Gordon from ARKive). Cute, no?

The authors observed chevrotains in Borneo.

The modern African species, which is considered evolutionary the most primitive of the Tragulidae (Webb and Taylor 1980), is a species of swamp and riparian habitats. When alarmed they are reported to rush for the nearest river and submerge, swimming upstream and coming to the surface beneath banks or overhanging vegetation (Kingdon 1989). The Asian species, however, are considered dry land animals, and no aquatic behaviour has been recorded (Lekagul and McNeely 1977; Phillips 1980; Payne et al. 1985; de Silva Wijeyeratne, 2008a de Silva Wijeyeratne, G., 2008a. A Photographic Guide to Mammals of Sri Lanka. New Holland, London, Silva Wijeyeratne 2008a). Two recent observations show that at least some of the Asian species of mouse-deer have retained aquatic behaviour, specifically for predator avoidance. We briefly describe these reports, and discuss the possible implications of our finding for understanding the early evolution of mouse-deer and other ungulates.

Aquatic escape behaviour was observed in June 2008 during a biodiversity survey in northern Central Kalimantan Province, Indonesian Borneo. Whilst surveying, the observers, which included the author U, noticed a mouse-deer, which author EM later identified from photographs as greater mouse-deer Tragulus napu, swimming in a forest stream. When the animal noticed the observers it submerged. Over a period of ca. 60 min, the animal came to the surface at least 4 or 5 times, possibly more often. The observers reported that the animal would remain submerged for more than 5 min at the time. Eventually, the animal was caught by hand, without resisting, identified as a pregnant female and subsequently released (Fig. 1).


Tragulus napu caught in a river after having spent 60 min hiding underwater.  (Figure 1 from the original paper)

Aquatic escape behaviour was also recently seen in the mountain mouse-deer (Moschiola spp.) (Groves and Meijaard 2005) of Sri Lanka (de Silva Wijeyeratne 2008b). Three observers, including author DSW, saw a mountain mouse-deer running into a pond and starting to swim. They noticed it was being pursued by a brown mongoose (Herpestes fuscus). The mongoose did not enter the water but at times approached within 2 m of the mouse-deer which responded by flaring its throat and showing the white on its throat. The mountain mouse-deer swam with only the upper half of its head out of the water (Fig. 2), and was completely submerged at times. After 15 min the mongoose left, and the mouse-deer came out of the water, but returned soon with the mongoose in pursuit and once again dived into the pond. Later the mouse-deer was caught. It offered no resistance, and was in an exhausted state. Investigation revealed that similar to the Bornean specimen this was also a pregnant female (de Silva Wijeyeratne, 2008b de Silva Wijeyeratne, G., 2008b. Mountain Mouse-deer photographed in Horton Plains National Park. Sri Lanka Wildlife Newsletter, December 2007–February 2008. left angle bracket angle Silva Wijeyeratne 2008b).

Clearly this animal can stay underwater for many minutes at a time.  And avoidance of predators by jumping into the water and remaining submerged is obviously something that natural selection could favor.  Ergo, hippos, and maybe whales eventually.  The authors also draw a connection to Indohyus, and speculate that this sort of aquatic escape behavior could have been ancestral in the Tragulidae, the family that embraces all nine species of chevrotain.

Our finding confirms that the aquatic escape behaviour is shared between three tragulid species which may have been evolutionary separated for over 35 million years (Hernández Fernández and Vrba 2005). This makes it likely that aquatic escape is a symplesiomorphic trait to all tragulids, suggesting that it could be ancestral to all Tragulidae. The evolutionary position of the Tragulidae as a very early diverged group within the Ruminantia raises the hypothesis whether aquatic habits, including aquatic escape is an ancient behavioural trait of all early ruminants. Although largely speculative, this idea may be supported by the recent discovery of a 48 million year old mouse-deer-like species, Indohyus, which is presumed to be a missing link between whales and ungulates (Thewissen et al. 2007). Based on its morphology, this species, which belongs to the extinct Raoellidae, is thought to have spent much time in and under water. Raoellids, which are included in the Ruminantia, are hypothesized to be the sister group to cetaceans (whales, dolphins and porpoises) (Thewissen et al. 2007).

What is nice about all this is the confluence of behavior, morphology, and the fossil record in providing not only the correct ancestry of modern whales, but also a rare result: a demonstration in real time of how an adaptive transition might have been made.

Here’s a video from YouTube (also mentioned in WEIT) showing the chevrotain’s swimming behavior. I don’t know if the eagle part is real; seems too fortuitious to me.


Aquatic escape behaviour in mouse-deer provides insight into tragulid evolution
Erik Meijaard, Umilaela, and Gerhan de Silva Wijeyatne. Mammlian Biology, in press.

11 thoughts on “The water chevrotain: almost like a whale

  1. Chevrotain hemoglobin must have some interesting oxygen-binding properties.

    It would also be interesting to see the resulting tree based on alignment of the amino acid sequences of chevrotain and whale hemoglobin chains vs. other mammals, too!

  2. When alarmed they are reported to rush for the nearest river and submerge, swimming upstream and coming to the surface beneath banks or overhanging vegetation (Kingdon 1989)

    Why upstream? If they want to get away, downstream would take the further away.

    1. Maybe predators expect them to go downstream & so it’s engrained counterintuition?

      Also, as we were reminded recently, dead fish take that route :-).

      1. Good point and funny reference.

        The chevrotain is awesome. I had no idea that we had something so close to an Indohyus alive today.

    2. About the upstream diving.
      I would guess the animal dives upstream to more easily be hidden under the surface and even reach the bottom of the river. It’s like when you start an airoplane you get better lift if you start opposite the wind direction.

      Watch also as it walks on the river floor how it keeps it’s head down, to make the water flow push it downward as much as possible it seems to.

      Just my humble guess though.

  3. Chevrotain hemoglobin must have some interesting oxygen-binding properties.

    More likely they just have a lot of it.
    Very cool ungulate & bird too. That hyper-Alda narration though…really bad.

Leave a Reply