Japanese Marten

Japanese Marten in the snow

The Japanese marten (Martes melampus) is a forest-dwelling mustelid native to several Japanese islands. A capable climber and adaptable feeder, it hunts animals, eats fruit and carries seeds through woodland. Its seasonal appearance can change from dark summer fur to a yellow winter body and pale head, although individuals and island populations vary.

A Japanese species with different island histories

The Japanese marten, Martes melampus, belongs to the weasel family, Mustelidae, within the mammalian order Carnivora. Its genus, Martes, also includes the sable and European pine marten. Despite that classification, its food is far from exclusively meat: fruit can dominate the recognizable remains in its droppings. It combines the climbing and hunting abilities of a forest carnivore with a substantial role in transporting plant seeds.

Its established native range includes Honshu, Shikoku, Kyushu and the Tsushima Islands. Recent Japanese research distinguishes the Hondo marten, M. m. melampus, on the first three islands from the Tsushima marten, M. m. tsuensis, endemic to Tsushima. These are island forms of one species. An established living population on the Korean mainland remains uncertain.

Japanese martens on Hokkaido and Sado have a different history. Animals from Tohoku were taken to Hokkaido for fur production before the Second World War and subsequently released; introduced populations occur in southern Hokkaido. On Sado, martens were released during 1959–1963 to control Japanese hares regarded as forestry pests. The Sado founders came from Honshu, although their precise origin there is unknown.

Size, shape and the difference between the sexes

An elongated body and sturdy climbing legs allow the Japanese marten to move between ground-level feeding places and trees. Tree climbing gives access to foods beyond the reach of a strictly terrestrial hunter, including insects associated with branches and fruiting vines.

Published reference measurements give a length of 47–54.5 cm for the head and body, with another 17–22.3 cm of tail. Reported adult mass spans 500–1,700 g. The tail measurement is additional to body length, an important distinction when judging the apparent size of a long-bodied animal.

Males are larger than females. In one live-captured sample, nine males averaged 1,563 g and four females averaged 1,011 g: the male mean was roughly one and a half times the female mean. That comparison describes measured animals rather than a fixed weight for each sex. Seasonal coat color is a separate dimension of appearance; a dark animal and a yellow animal need not represent male and female color forms.

Why the coat changes so dramatically

General descriptions of Japanese martens encompass yellowish-brown through dark-brown fur, with a white or cream neck patch. The Hondo form can look especially different between seasons. In summer, its face is black and its body brown; in winter, the face becomes white and the body yellow.

The Tsushima form also changes seasonally. Summer pelage is very dark, whereas winter brings a yellowish-brown body and pale head. Dark fur remains on the feet and around the nose and mouth.

Seasonal change also varies naturally among individuals. Inokashira Park Zoo describes martens called su-sten that retain summer-like coloration through winter. A dark winter coat is consequently not sufficient grounds for rejecting an identification as Japanese marten. Locality, season and several visible features together are more useful than a single expected shade.

Light helps regulate this transformation. A captive experiment with martens from Kagoshima found both a response to light exposure and an internal annual rhythm in molting. Under constant short days, molt cycles became unsynchronized; one male exposed to long days and bright light retained dark fur for three years.

Recognizing a Japanese marten beside a sable

Hokkaido presents a practical identification challenge because introduced Japanese martens occur alongside native sables, Martes zibellina. In the published Hokkaido identification guide, sables generally become darker from the neck toward the tail, usually ending in a black tail tip. Japanese martens can instead have a tip paler than the tail base, sometimes white. That feature varies between individuals and becomes less conspicuous in winter.

Facial markings provide another seasonal clue. Extensive summer black retreats toward a Japanese marten’s nose and eyes during the transition to winter and nearly disappears in midwinter. A bright yellow winter body and white head can be striking, but pale-bodied sables also occur. The direction of color change along the body, facial pattern and tail tip are therefore best considered together.

The European pine marten, Martes martes, offers a different comparison: its head generally matches its brown back, whereas the Japanese marten can have a conspicuously lighter head.

Forest life from burrows to branches

Broadleaved woods are an important habitat, but Japanese martens also use forest valleys, conifer plantations and open fields. Resting shelters include tree dens and ground burrows.

Habitat selection is more specific than this broad list of places used. In Akaya forest in central Honshu, a camera-trap study found Japanese martens associated with broadleaved forest and shorter trees. The authors suggested that complex forest structure supplying cover and fruiting trees could help explain the pattern. This shorter-tree association differed from earlier Tsushima observations favoring mature, taller trees. The contrast suggests that vegetation height alone cannot define suitable habitat across the islands; forest composition and the resources within a stand also matter. Because the Akaya cameras were placed along accessible roads and trails, the height association describes those sampled settings, rather than a universal preference for low trees.

Human environments can adjoin this woodland life. Houses near forests are used, and Tsushima rangers report martens in settlements and town areas as well as forest. Camera records there show individuals standing on their hind legs with the forefeet held before the body while looking around. This upright posture raises the head above its usual position during four-footed movement and offers a distinctive view of a watchful marten.

The species is generally described as nocturnal. On Tsushima, encounters extend to roads, where collisions occur.

A diet that follows place and season

Fruit, invertebrates and mammals are the major recurring components of Japanese marten diets. Their relative importance changes across forest climates. A regional synthesis found mammalian prey more frequently in subalpine and cool-temperate settings, and fruit more frequently in warm-temperate settings. Seasonal changes occurred in all three main food categories, linking the marten’s meals to both its region and the changing supplies within it.

At Otome Highland in central Japan, fleshy fruit was prominent in autumn and winter, insects in summer, and mammals in winter and spring. Fruit occurred in more than 80% of autumn and winter fecal samples; insect occurrence in summer was 40–80%, and mammal occurrence in winter and spring was 60–80%. A dropping can contain several foods, so these are overlapping frequencies of occurrence, not portions of a meal that should add to 100%.

In northern Japan, a separate study found fruit occupying 65% of the fecal area examined, compared with 13% for arthropods and 11% for mammals. There, arthropod representation was higher in spring and mammal representation higher in summer. Fruit increased when measured arthropod and mammal abundance was lower. This differs from Otome’s seasonal pattern and illustrates why a single calendar of food use cannot describe all populations.

Neither the area of recognizable material in droppings nor the frequency of samples containing it directly measures consumed biomass or energy.

Snow can change access to prey even within one region. Over five sampled winters between 2019 and 2024 around Tsuruoka in northeastern Honshu, small mammals occurred more frequently in Japanese marten droppings during low-snowfall years. Red foxes showed the same association, while hare remains and those of other, larger mammals showed no significant snowfall response. The researchers proposed that reduced snow cover made small mammals more accessible: spaces beneath snow can otherwise shelter them from predators. This mechanism concerns the difficulty of reaching prey, not simply how many prey animals inhabit the forest. It helps explain how winter diets can change without a change of season or a move to a different forest zone.

Hunting insects and harvesting unusual fruits

Japanese martens exploit different insect life stages and feeding sites. Food records include ground-dwelling beetles, tree-associated insects, mantis egg masses, and wasp larvae or pupae taken from nests. Consuming an egg mass or nest brood gives access to immature insects gathered in one place, unlike pursuing an individual adult. Records of winged insects do not necessarily mean aerial capture: those animals can also be caught while settled.

At Otome, cicada larvae occurred especially in early summer, while camel crickets appeared in autumn and winter. DNA analysis of feces in southwestern Honshu broadened the picture to include adult beetles, cicadas, sphinx moths, grasshoppers and earthworms, together with seasonal plant foods. Soft-bodied prey can leave few recognizable hard remains, making DNA particularly useful for revealing foods that visual examination may miss.

The same DNA work detected brown-eared bulbul in winter to early spring. Such detections identify food taxa rather than the weight eaten or the circumstances of capture.

Fruit feeding is equally varied. Japanese martens consume small, colorful fleshy fruits as well as larger, duller ones such as Actinidia arguta. Fruiting vines and forest-edge plants are prominent among recorded foods; A. arguta and Vitis coignetiae were frequent in the Otome material. In Hovenia dulcis, the fleshy reward is unusual: the swollen, sweet-smelling fruit stalk is eaten, while the actual fruit is a drupe. What attracts a feeding marten is therefore not restricted to the botanical fruit itself.

What happens to the seeds?

Fruit consumption can move seeds out of the feeding place. In northern Japanese subalpine shrubland, marten droppings contained seeds from woody and herbaceous plants, including a vine associated with lower elevations. Fruits that martens ate tended to have greater size and flesh mass or volume than those they did not eat.

The interval between swallowing and defecation creates an opportunity for transport. In captive Japanese martens at Toyama Municipal Family Park Zoo, mean seed passage took 7.4 hours, with about two-thirds of seeds passed within 12 hours. Passage times varied widely, from 0.6 to 51.7 hours. Seeds swallowed together need not all leave the animal at the same time, so feeding at one plant can lead to deposition at different points along subsequent movements.

A model combined those captive passage measurements with wild ranging data from the northern foot of Mount Fuji. It placed many dispersal events 501–1,000 m from feeding sites and estimated that 20% exceeded 1,000 m. These are modeled distances, not individual seeds followed continuously through a wild marten. They describe how the combination of movement and gut retention can spread seeds beyond the immediate surroundings of a fruiting plant.

Experiments with Actinidia arguta address a different stage: germination. Germination proportions were higher for seeds ingested by captive martens than for seeds in the un-ingested treatments, and removing pulp also improved germination compared with leaving seeds in pulp. Intact seeds survived passage through the gut. Removal of germination-inhibiting pulp offers a mechanism by which eating the fruit can benefit the seed, rather than simply carrying it elsewhere.

The benefit concerned the proportion germinating, not a demonstrated acceleration: ingestion did not significantly shorten germination time. Nor did the tested field light levels significantly alter germination proportion. Transport, germination and seedling establishment are separate steps. A seed deposited intact and able to germinate must still survive after germination before it contributes to a new plant population.

Comparing two seed-carrying mustelids

Japanese martens share eastern Honshu forests with Japanese weasels, Mustela itatsi. The marten is larger and semi-arboreal; the weasel is smaller and terrestrial. Their fruit consumption overlaps, but a western Tokyo comparison found seeds in 81.4% of 381 marten fecal samples and 55.7% of 97 weasel fecal samples. Those percentages compare droppings, not the proportion of individual animals that ate fruit.

Marten droppings represented 28 plant species, compared with 17 in weasel droppings, and contained more seeds overall; almost all recovered seeds were intact. Longer-term work in Ishinomaki also found changes between years in seed occurrence and seeds per marten dropping. Their contribution to dispersal varies with ecological circumstances rather than being a fixed annual service.

Solitary spacing and the anatomy of scent

Japanese martens are usually solitary, with breeding-season associations. They occupy defined territories, and concentrations of droppings near home-range edges are consistent with scent-based spatial communication. Scent can persist at a location after its producer has moved away, offering a possible way for animals that seldom meet to exchange information. The specific messages conveyed by Japanese marten abdominal secretions, however, have not been demonstrated experimentally.

Anatomical work found a specialized abdominal gland region immediately in front of the external genital area in specimens from Tsushima and Tokyo. The skin bore strong-smelling brown secretion, and microscopic examination showed enlarged sebaceous glands together with smaller apocrine glands.

Space use can change around reproduction. At Mount Nyugasa in Nagano, the single adult male continuously followed for seasonal ranging had his largest home range in July and highest activity in July and August. Male reproductive physiology in the same study supported a mainly July mating period. The alignment suggests a link between breeding and this male’s movements, while leaving the typical range size and seasonal behavior of other animals unresolved.

Summer mating and spring births

The Nagano males’ testes enlarged from April and reached their greatest size from late June to mid-July; testosterone was high in May and June. On Tsushima, field observations and reproductive measurements instead support mating from late July to mid-August, with copulations observed in July and August. These local findings place breeding in summer without requiring identical dates throughout the species’ island range.

Tsushima births were recorded from mid-April to early May. All nine litters examined in that account contained two young. The roughly 235–250 days between mating and birth led the researcher to infer prolonged delayed implantation: an interval before the embryo implants, followed by a shorter period of development after implantation. Direct implantation timing was not established, and the suggested 28–30 days after implantation was an estimate.

Four newborn Tsushima martens weighed about 25–30 g and had sparse, fine pale-gray hair, a white throat patch, and closed eyes and ears. They were altricial young dependent on maternal milk and care, very different from a mobile adult able to hunt and climb.

Growing into a climbing hunter

Hand-reared Tsushima juveniles provide a limited view of that progression. Estimated eye-opening ages were 13–22 days, with weaning at 40–50 days. Among the hand-reared juveniles, three had sufficient mobility for arboreal living at estimated ages of 90–100 days. The study also reported prey-capture ability around 100–120 days. These observations distinguish early feeding development from later locomotor and hunting skills; weaning alone does not mean that a juvenile can already perform all the tasks of an adult.

The ages were estimates from hand-reared animals, and natural independence and dispersal ages remain poorly documented. Wild life expectancy is also not established. One Japanese marten survived 12.3 years in captivity, but that record is neither an average wild lifespan nor a confirmed maximum for the species.

Conservation depends on the island

Globally, the Japanese marten is classified as Least Concern, with large stable subpopulations cited in the IUCN justification. The Tsushima subspecies has a different national assessment: Japan identifies it as Near Threatened, and it is also a nationally designated Natural Monument.

Habitat loss and degradation associated with intensive deforestation, plantation establishment and urbanization affect Japanese martens in parts of Gunma, Aichi and Tsushima. The species’ use of some conifer plantations does not mean that replacing other forest with plantations has no ecological cost. Using a habitat and maintaining populations through changes in habitat are different biological outcomes. Traffic mortality adds a direct hazard where Tsushima martens encounter roads.

Introduced populations raise a separate issue. On Sado, martens prey on native Japanese hares, the very animals their historical release was intended to control. On Hokkaido, competition or hybridization with native sables is identified as a potential impact, rather than established displacement across the island. The same Japanese marten is thus a native protected animal on Tsushima and an introduced carnivore on Hokkaido.

Animal Product Reviews

Dive into Animal Streets’ Animal Product Reviews – your guide to the best gear, gifts, books, and gadgets inspired by the animal kingdom. From land-roaming lions to ocean-gliding whales, from household pets to prehistoric giants, we explore top-rated products for enthusiasts, educators, and animal lovers alike. Whether you’re shopping for toys, decor, field gear, or fun collectibles, we’ve sniffed out the best so you can discover it all in one wild place!