Jump to content
Forum for Epiphytic Myrmecophytes

Melastomataceae myrmecophytes

Recommended Posts


Pachycentria Carl (Karl) Ludwig von Blume published in Flora 14: 519, 1831.

This genus, now including the monospecific Pogonanthera (Blume) form part of the Dissochaeteae an Old World tribe of the Melastomataceae, a species-rich, pan tropical family distributed in Burma, Thailand, Peninsula Malaysia, Sumatra, Java, Borneo, Philippines, Sulawesi and New Guinea.

  Ecology. Pachycentria have certain anatomical characters presumed to predispose their members toward interactions with ants.  Indeed some terrestrial members of the family provide ants with readymade domatia in hollow stems or leaves.  (Clausing 1998 & 2000.)  However, of the eight Pachycentria species, most are terrestrial shrubs reaching about 8 m tall (Clausing 1998 & 2000,) hence are beyond the span of these notes.  Our interest is in a few of the smaller species that are primarily or totally epiphytic and have evolved intimate relationship with ants.

Beccari (1884-6) and Huxley (1980) both suspected that certain species of Pachycentria occasionally housed ants in their hollow swollen roots and Janzen 1974 cited by Clausing (1998.) reported that in Bako National Park, Sarawak, Malaysia; P. tuberosa (now P. glauca) sometimes grew epiphytically on Hydnophytum formicarum tubers inhabited by ants although the Pachycentria themselves were not ant occupied.  (Both Medinilla maingayi Clarke (sic) and Pachycentria constricta (Blume) Blume were also reported to grow epiphytically on ant-plants or in ant-carton nests without themselves being inhabited by ants.  (Kiew & Anthonysamy 1987 cited by Clausing 1998.)  M. maingayi is now correctly P. glauca subsp. maingayi. (Clausing's 2000 Pachycentria revision.)

  Pachycentria tubers have the anatomical structure of storage roots and are not ant inhabited unless as sometimes happens the epidermis of some tubers crack and they become hollow, a process that takes several days.  This seems not to harm the plant and one of the numerous nearby ant colonies will quickly spread into vacant tubers.  (Clausing 1998.)  Seed of both Pachycentria constricta and P. glauca is attractive to ants, which are induced to take them to their arboreal nests. (Clausing 1998 & 2000.)  However, any perhaps olfactory ant attractant has not yet been identified but an incentive for myrmecochory may be provided by the presence of lipids in seed embryos. (Clausing 1998.)  Seed of both these species is also larger (2.0- 2.5mm long) than others in the genus (which only reach 1 mm long.)  They are dispersed over longer distances by small to middle-sized birds that eat the fleshy berries.  However, ants possibly reclaim some seed from bird droppings and probably glean both food and seed directly from the berries themselves.  Certainly, seedlings of both species have been observed growing directly on ant-house plants and in ant gardens. (Clausing 1998.) (Janzen 1974, Kiew & Anthonysamy 1987, cited by Clausing 2000.)

  Both Pachycentria constricta and P. glauca provide pearl bodies, tiny lipid-rich, hence ant-nutritious foods on their leaves, stems, inflorescences and even flowers, which when eaten contribute to the energy budget of resident ant colonies” (Clausing 1998.)  Hence they provide a positive input to entire ant-house guilds if only indirectly.  It is therefore, not at all surprising, that in the stunted, very nutrient-poor Kerangas forests of Bako National Park, Sarawak; P. glauca grows extensively if not exclusively on the tubers of Hydnophytum formicarum (Clausing 1998.)  Also both P. constricta and P. glauca are often found in arboreal ant gardens along with Hoya spp. (Clausing 1998.)  Therefore, arboreal Pachycentria species may merely exploit myrmecophytic guilds.  Clausing (1998) considers the association with ants of arboreal Pachycentria species is facultative not obligate.  Yet I would suggest that in seasonally harsh and very nutrient-poor environments such as at Bako National Park that an association with ants is important perhaps even obligate for survival there, even if it must be mediated by the presence of ant-house plants such as Dischidia major, Hoya species, Hydnophytum formicarum, Lecanopteris sinuosa, or Myrmecodia species.

  “As the principle purpose of Pachycentria root swellings is probably water storage,” we have here plants certainly adapted to occupy the same exposed and periodically arid niches so often inhabited by arboreal ant-plants, especially in nutrient poor habitats where symbiosis with ants considerably assists survival.  However, if one presumes that such environmental niches were already supplied with well-evolved ant-house plants, the need to supply further ant domatia is greatly diminished.  (Clausing 1998.)

  “Without good fossil records it seems our understanding of how some interactions evolved is extremely problematic.  Furthermore, comparisons between New and Old World ant-plant evolutions must surely be complicated by the abundance of arboreal ant-plant species throughout South East Asia.” (Clausing 1998.)

  Flowers are normally hermaphroditic with out-crossing encouraged by a separation of anther tips and stigma.  (Clausing 2000.)  Pollination is performed by bees that harvest pollen using wing vibrations (buzz pollination); this being the only food reward 'offered' by these flowers.

  P. hanseniana, P. pulverulenta and P. varingiifolia are also epiphytic so may be of myrmecophyte interest but both of the following species are accepted as true ant garden epiphytes by Orivel & Leroy (2011.)

  Pachycentria glauca subsp. glauca Triana. Not listed in the Tropicos database.

This small 20-60 cm (8- 24”) tall species is of special interest because it is found (as subspecies glauca) in fascinating Bako National Park, Sarawak, on Borneo island and it also occurs in Brunei, Sabah and in Indonesia’s Kalimantan Province to the west.  (Clausing 1998 & 2000.)  It is therefore, a species that enthusiasts may easily see in habitat if only in the many Internet posted photographs from this popular park.

  P. glauca can sometimes superficially resemble Hydnophytum juveniles having comparable pendent or creeping to erect stems and an outwardly similar leaf structure.  Here however, the comparison ends because its adventitious roots form not a single larger tuber as in Hydnophytum but many small, primarily globular swellings, each reaching only 2-3 cm in diameter.  (Clausing 1998, 2000.) Its roots enter the tubers of Hydnophytum formicarum in Bako National Park but they do not penetrate tuber tissue hence are not parasitic like mistletoe but are parasitic on the myrmecophyte -Philidris mutualism.  Indeed, P. glauca subsp glauca and I quote, “almost exclusively grows with epiphytic ant-house species such as Dischidia R. Br. (Asclepiadaceae) sic., (Kiew & Anthonysamy 1987) or Hydnophytum (Janzen 1974)” “where its seeds must have been deposited by ants and germinated in the nutrient-rich debris in or near the nest.”  (Clausing 1998.)  It also grows frequently on other epiphytic ant-plants such as Myrmecodia and Lecanopteris.

  Habitats vary from sea level to 1100 m. (3609 ft.) in Kerangas and other heath forests.  (Clausing 2000.)  These are often nutrient poor habitats and it is particularly noteworthy that Philidris is the ant genus associated with P. glaucaPhilidris are both ant-house mutualists and makers of ant carton.

  One to three rather insignificant flowers open each day and flowering may extend for some weeks to months.

P. glauca subsp maingayi (Clarke) Clausing, occurs only in Thailand, (Wai 2009) Peninsular Malaysia and on nearby Sumatra island in a small area just across the narrow Malacca Strait. (Clausing 2000.)

P. constricta (Blume) Blume, synonyms P. formicaria, P. macrorhiza, P. tuberculata.

  Description: an epiphytic to rarely terrestrial shrub reaching about 1-2 5 m. (40-100 inches) tall. Orange to brown tuberous swellings on adventitious roots can individually reach 20cm (8") long and 4 cm (1.6") wide, hence they are far more sausage-shaped than those of P. glauca. Its thin leaves are very varied in both form and size and petioles (leaf stalks) can vary from being subsessile (almost stalk less) to being 1.5 cm long.  (Clausing 1998 & 2000.)

  Habitats: Are highly varied from wet or dry, nutrient poor or rich, primary, disturbed or logged lowland forests, or lower montane rain forests, peat-swamp and other marshy forests, heath forests, kerangas forests, riverine forests and even riverbank trees.  Altitudes vary from sea level to 2,000 m. (6,562 ft.)  The species are commonly low-growing epiphytes perched at heights of 1-10 m. (3-33 ft.) but it has been recorded as high as 30 m. (98 ft.)  (Clausing 1998 & 2000.)

  Range. More widespread than P. glauca being distributed from Burma, Thailand, and Peninsular Malaysia to the islands of Sumatra, Java, Borneo, Sulawesi and Philippines (Mindanao and Zamboanga islands.)  (Clausing 1998 & 2000.)

Ecology: P. constricta is primarily an ant-garden mutualist and when growing in carton ant-nests, its tangled roots contribute to the overall stability of the nest.

Cultivation conditions will influence leaf size and thickness.

Medinilla crassifolia, Blume.  Not in the Tropicos database but Orivel & Leroy (2011) list this species as a true ant garden inhabitant in peninsular Malaysia.


A number of Medinilla species both epiphytic and terrestrial are popular in horticulture due to their attractive flowers; however, perhaps of special interest here are a number of Madagascan species reputed to have Pachycentria type swollen roots.

Link to comment
Share on other sites

Hi Derrick,


I got several Melastomes (Melastomataceae), especially Medinillas, not really for their interaction with ants, but just for their ornamental value. It is a very interesting family, that indeed produces tubers some times. Here are some pictures of what I think is Medinilla basaltarum from Madagascar and the Comoron Isles. Its roots are not big enough for ant interactions, and I don't see any holes in my plants, but some kind of spidermites really loves this plant. So far I didn't see big enough tubers with any holes on Pachycentria glauca or the Medinilla specie from Gregor Hambali either. Both of them come from the Philippines, so far as I know.











Link to comment
Share on other sites

  • 2 months later...

Thank you Niek, I have only just found this. Very interesting. I wonder if ant gardens exist in Madagascar?  Edit.  It seems they do but there is little information on the WWW. However, Piere Jolivet mentions Madagascan ant gardens in his 1996 book Ants and plants: an example of coevolution

And there is one mention of this Medinilla species in Annales des sciences naturelles: Botanique et biologie végétale.  Is it known to be epiphytic?  If so it may very well be myrmecophytic if not an ant-house species.

Link to comment
Share on other sites


This topic is now archived and is closed to further replies.

  • Create New...