Macaranga

AntWiki: The Ants --- Online

The following introduction is based on the Maschwitz et al. 2004.

The most prominent ant-plant system of perhumid South-East Asia consists of the pioneer tree genus Macaranga and its manifold associations with ants. The genus Macaranga (Euphorbiaceae) comprises species which are not ant-inhabited and/or facultatively ant-associated as well as obligate ant-plants (review in Fiala 1996, Fiala et al. 1999). While food bodies and/or extrafloral nectar produced by non-myrmecophytic species attract various opportunistic ants for occasional visits, obligate myrmecophytes also offer nesting space for their specific ant partners. In obligate myrmecophytes, the ants also harvest honeydew produced by specific scale insect partners maintained inside the internodes (Heckroth et al. 1998). In exchange for the provision of food and nesting sites, the ant partners effectively protect their host plants against herbivore damage, competition by climbers (e.g., Fiala et al. 1989, Heil et al. 2001) and fungal infection (Heil et al. 1999). For further details on ecology and biology of ant-associated Macaranga we refer to the extensive work of our group in the last 15 years (for overview see, e.g., Fiala & Maschwitz 1990, 1992; Fiala 1996, Fiala et al. 1999, Blattner et al. 2001, Federle et al. 2001, Feldhaar et al. 2003a, and references therein).

Macaranga comprises about 280 species with a range stretching from West Africa throughout the Oriental region eastward to the Fiji Islands (Whitmore 1973). It is the only plant genus in the palaeotropics exhibiting such a substantial radiation of myrmecophytes (about 29 known myrmecophytic species in the Malayan Archipelago, see recent revision of Davies 2001). Most myrmecophytic species are found in two sections (Pachystemon and Pruinosae) each of which, although closely related, represents separate evolutionary acquisitions of myrmecophytism (Blattner et al. 2001, Davies et al. 2001). The majority of partner ants belong to the myrmicine genus Crematogaster (subgenus Decacrema). Additionally, two Macaranga species (M. puncticulata and M. lamellata) were found to be specifically inhabited by species of the genus Camponotus, subfamily Formicinae (Maschwitz et al. 1996; Federle et al. 1998a, b; Fiala et al. 1999).

The center of diversity of this group of three-partner associations (plants, ants and scale insects) is Borneo, spreading westward to Sumatra and northward to the Malay Peninsula (Fiala et al. 1999), with Borneo hosting many more endemic host plant species (see revision of Davies 2001) and more ant lineages as well (Fiala et al. 1999, Feldhaar et al. 2003b). So far no information exists on the age and geographic origin of the Macaranga-ant system. The first (nonmyrmecophytic) Macaranga species – thought to be bushlike pioneers of open places – was hypothesized to originate in the Oligocene or early Miocene (38 to 15 mya) when the climate was predominantly seasonal and dry (Slik & van Welzen 2001). Rainforest, the habitat of all myrmecophytic Macaranga species, colonized the region in the mid-Miocene when the climate changed to perhumid (Morley 2000) and also when the greatest northward extension of rainforest occurred, so we assume a development of the first close Macaranga-ant associations not before this period. For any interpretation of the evolutionary development of the Macaranga-ant associations, information on their biogeography is required. The section Pachystemon, which comprises most myrmecophytic species, is restricted to the moist tropics within the floristic boundaries of West Malesia (sensu van Steenis 1950) and does not extend into the more seasonal monsoon forest regions (e.g., Whitmore 1975, Fiala et al. 1999, Davies 2001). We do not know what limits the distribution of the myrmecophytic associations. Climatic factors certainly play an important role since low temperature and increased dry periods do not allow continuous food production of the plants for their partner ants. Inhabited myrmecophytic trees in the Malay Archipelago do not occur above 1250–1400 m (Fiala et al. 1999), although ants not associated with myrmecophytes were still found foraging on vegetation up to 2300 m in the Mt. Kinabalu area, Sabah (Kern 1996). Since ants and plants as well as associated coccids are usually not able to survive without each other (e.g., Fiala 1996, Heckroth et al. 1998), and ants were never found outside their host plants, it is difficult to tell whether Decacrema ants alone could survive in other climatic zones. The interdependency of all partners is so strong that obviously their distribution range is always correlated.

Macaranga bancana

Figure 1. Upper row: Habitat of Macaranga bancana trees (white arrow, with trilobed leaves), middle row left: stem full with large larvae and pupae and alate females of Crematogaster captiosa, right: opened stem of Macaranga winkleri, alate female of C. sp. 8. (photo E. Linsenmair); lower row (left to right): colonisation of Macaranga bancana by C. captiosa at nighttime; colonising queen of C. sp. 8 on Macaranga winkleri at daytime (photo E. Linsenmair); colonising queen of Crematogaster linsenmairi on Macaranga pearsonii, plant pith from the interior which is later used to seal the entrance hole is visible on the stem surface.

Fialfa et al. (2017) - The pioneer tree species Macaranga bancana, sect. Pachystemon, (formerly known as M. triloba (Thunb.) Müll.Arg., 1866 (Davies 2001)) occurs in in lowland areas of dipterocarp forests as a comparatively small subcanopy species up to 23m height. Like most myrmecophytic Macaranga trees, M. bancana has episodic annual reproduction (Davies and Ashton 1999; Moog et al. 2002, Fiala et al. 2011), and the peak availability of young domatia-bearing saplings is mainly following this reproductive cycle. However, seedlings can also arise from seed banks after forest disturbance (Howett & Davidson 2003). The M. bancana population in our main study site, Ulu Gombak, flowers mainly from January until April, with varying intensity over the years, and a few plants were also seen flowering in June, August and in November. Fruit set occurred usually three to four months after flowering. In open areas, often dozens of seedlings could be found directly underneath mother trees, but mortality was high. See Crematogaster captiosa for more about this plant and its ant inhabitants.

Macaranga lamellata

Figure 2. Colony-founding queen of Colobopsis macarangae with a first batch of eggs inside a hollow internode of a Macaranga lamellata sapling.

Maschwitz et al. (1996) – The size, spatial distribution, light environment and growth rates of M. lamellata trees were studied in two 10-ha primary forest plots at Lambir to investigate whether there were ecological differences between Colobopsis- and Crematogaster-inhabited plants. In a sample of 72 plants, 48 (67%) were inhabited by Colobopsis macarangae, 21 (29 %) by Crematogaster species, and 3 (4%) planes had no ant-occupant. The size-structure of the Colobopsis inhabited plants did not differ from Crematogaster-inhabited plants (P = 0.56; Kolmogorov-Smirnov two-sample test), and all plants above 2.6 cm diameter at breast height were ant-inhabited. Nine other myrmecophytic Macaranga species occurred in these plots, namely M. beccariana, M. kingii var. platyphyla, M. havilandii, M. hosei, M. hullettii, M. hypolenca, M. trachyyphylla, M. triloba and M. winkleri. In both plots, M. lamellata, tended to be more evenly distributed than the other Macaranga species; this is due to its greater shade-tolerance (Davies et al., in prep.). Although we did not identify the ants in all M. lamellata trees in the two plots, there appears to be no obvious spatial patterning of trees with different ant-colonists. Crematogaster colonies were intermingled with Colobopsis-inhabited plants. The higher density of M. lamellata in Plot 1 is at least partly due to variation in soils and drainage within the plots (Davies 1996). There was also no significant difference in the estimated light environment between the trees with different ant-colonists (Davies et al., in prep.). The genus of ant-colonist was not significantly correlated with diameter growth rates over two and a half years (P = 0.42; Mann-Whitney U-test), though unoccupied plants grew somewhat slower than ant-inhabited plants.

Macaranga griffithiana

Maschwitz et al. (2004) - Macaranga griffithiana in Peninsula Malaysia is a very common lowland ant-plant found mainly in open disturbed habitats with moist to swampy soils, i.e. on plains which can be flooded during wet times, on banks of streams, etc.. Roadside drains are frequently occupied by M. griffithiana. It was usually associated with Crematogaster (Decacrema) msp. 1 and with a specific coccid partner (Coccus caviramicolus) although a few other scale species also occurred. In south-east Thailand 160 km south-east of Bangkok, about 80 km from the Cambodian border, we found M. griffithiana, associated, however, with a different ant species: Colobopsis markli from the subfamiliy Formicinae.

Plant characters. M. griffithiana grew up to 15 m within forested habitats but was most common as a bushy treelet or broadly ramified shrub at open sites and forest edges. Thin young stem parts and branches, as well the undersides of the rather weakly trilobed leaves, are waxy and whitish in color. The stem becomes hollow by pith degeneration thus forming intermodal domatia. Seedlings can begin to develop domatia at a height of less than 0.5 m; below that height stems often still have a rather small diameter (about 5 mm only).

The paired stipules are rather short lived and persist only in the two or three most apical leaves of a shoot. Food bodies are found beneath the stipules and (as seen on saplings kept in the greenhouse) in large numbers on both sides of young leaves, preferably close to the veins.

Macaranga puncticulata

Federle et al. 1998 - Mated queens of Colobopsis sp. 1 are capable of locating young M. puncticulata plants and of chewing entrance holes into the domatia. In contrast to all the other obligate plant-ants associated with Macaranga, Colobopsis sp. 1 never cultivates any trophobionts, neither inside nor outside the domatia. Instead of taking up honeydew from coccids, the ants obtain carbohydrates from extrafloral nectaries along the leaf margins of M. puncticulata. This food source is an uncommon trait within Macaranga myrmecophytes. This is the first record of a two-partner ant-plant system without endophytic coccids in the genus Macaranga.


Several characters of the association partners are characteristic of close myrmecophytic relationships:

– maintenance of the full set of ant-plant characters in M. griffithiana, i.e., formation of thin-walled, fragile domatia, and a high rate of food body production,

– high specificity of associations at several separated subpopulations in open and forested habitats, monopolization of trees by colonizing ants although other arboreal ant species were present,

– host plant localization, and domatium opening behavior (chewing of entrance holes into living plant tissue) of the colony-founding queen,

– frequent food body consumption by the partner ant,

– aggressiveness and probable pruning behavior (this was also reported from Colobopsis colonizing M. puncticulata (Federle et al. 1998),

– presence of a single coccid partner species. This homopteran tending as well as food body harvesting can make the ants independent of food searching away from the plant.

All these characters taken together speak for a close association of all three partners. It obviously represents another example of the independent origin of a myrmecophytic association within the diverse genus Macaranga. The two other associations of Macaranga plants with Colobopsis species differ in many aspects to the one reported in this study. The Macaranga species involved are not closely related, M. griffithiana (West Malaysia, Thailand) and Macaranga lamellata (Borneo) belong to different clades within the section Pachystemon, and Macaranga puncticulata (West Malaysia) is even more distant, forming its own clade (Blattner et al. 2001, Davies 2001). On the ant side as well, initial phylogenetic analyses based on mitochondrial DNA sequence data of all Camponotus species involved revealed that C. markli is not closely related to either of the other Macaranga-colonizing Colobopsis species (J. Gadau et al., unpubl. results), i.e., it has obviously evolved independently as a plant-ant.

Related Pages

References