Vachellia ruddiae (D. H. Janzen) Seigler & Ebinger

Acacia ruddiae

Arbol usque ad 20 m alta; ramuli novelli puberuli, glabrescentes; stipulae spinescentes interdum magnae, teretisculae, usque ad 10 cm longae, 9 mm latae; folia circiter 15−40 cm longa; nectaria sessilia cupulata ad juncturas pinnarum omnium et 0−2 in petiolis; pinnae (12−) 25−40-jugae, 4−5 cm longae; pinnulae 65−75-jugae, lineariae, 2−4 mm longae, 0.5−1.2 mm latae, acutae vel obtusae,

puberulae; flores aurei in capitulo globoso; involucellum annulare in medio pedunculo; legumina (immatura) atrobrunnea, puberula, granulata glabrescentes, oblonga, 5 cm longa, 7 mm lata, complanata, obtusa, stipite 5 mm longo; legumina matura probabiliter usque ad 6–8 cm longa.

TYPE.—D. H. Janzen 828, Costa Rica, Limon Province, ferry landing, Moin River, Moin, 13 February 1965 (holotype US 2639554; isotypes CAS, F, GH, MEXU, MICH, MO, NY, UC).

Type tree 20 m tall with 35 cm diameter trunk 1 m above ground; gray-green bipinnate leaves mostly 15 to 40 cm long with 50 to 80 pinnae per leaf, pinnae 40 to 50 mm long with 130 to 150 pinnules per pinnae, pinnules 2.5 to 3.0 mm long and 0.75 to 1.0 mm wide (up to 9,000 pinnules per leaf); occasional swollen-thorns on less than 10 percent of the nodes and very frequently only one side of the thorn swollen, thorns not occupied by obligate acacia-ants but by a miscellany of other ant species; no food bodies (Beltian bodies) on pinnules; single nectary at junction of each pinna with rachis, 0 to 2 nectaries on petiole, nectaries cylindrical up to 1 mm tall; yellow inflorescences spherical and up to 6 mm diameter, peduncle up to 10 mm long and 0.50 to 0.75 mm thick; immature fruits flat and slightly curved, probably 6 to 8 cm long at maturity.

Acacia ruddiae is distinctive by never being occupied by obligate acacia-ants, lacking swollen thorns on most branches except the bases of some lateral branches, having better than 5,000 very small leaflets per leaf, growing in rain forest, becoming a very large tree, and by its range (Figure 23). It also differs from other Central American members of the genus Acacia by the same traits; it is undoubtedly distinguishable from the remainder of the genus by fine details of flower, fruit, and leaf structure, but it is impossible to know these differences at the present state of Acacia taxonomy.

The tree from which the type specimens (13 sheets) were collected was at least 30 years old, and perhaps much older, and was still standing in 1967. It was about 80 m south (west side of the road) of the south beach of the ferry landing on the Moin River and 200 m from the ocean beach, at Moin, Costa Rica (site 1). The branches were cut from high in the tree and were representative of the 1-2-year-old branches in the canopy. Although the specimens were collected during the dry season, the area was still subjected to frequent rains and the tree was only partly deciduous. It was at the end of its flowering season, and the full sized but immature green pods were from the same flower crop, rather than from the previous year’s flowers as is usual with acacias of drier habitats. As is usual with the canopy branches from A. ruddiae this size, there are few swollen thorns on the branches. The tree from which the type specimens were collected was exceptional in two respects: First, most of the A. ruddiae population in this area has long since been destroyed by housing development. Second, the tree grew in a relatively open site and thus had strong

development of lateral branches; most A. ruddiae are involved in intense intertree competition and thus are tall and slender.

This acacia is named in honor of Dr. Velva E. Rudd whose unfailing assistance and encouragement has greatly aided in bringing this study to completion, and whose taxonomic work with Leguminosae will always be a tremendous assistance to field biologists.

Acacia ruddiae is included with the swollen-thorn acacias, despite the fact that it is not occupied by obligate acacia-ants, because it appears to represent an intermediate step in one of several possible evolutionary pathways to the evolution of swollen-thorn acacias. Its ecology is therefore of special significance in this study, A. ruddiae is the only Central American acacia, aside from hybrids of swollen-thorn acacias with nonant-acacias, that appears to represent an intermediate step in the evolution of a New World swollen-thorn acacias.

KEY COLLECTION LOCALITIES

1. TYPE.—Ferry landing, Moin River, Moin, Limon Prov., Costa Rica. II–13–1965, DHJ, no. 828, fl. gp.

2. Pandora, Limon Prov., Costa Rica. VIII–26–1963, C.D. Michner, nc.

3. 16 mi N Vara Blanca on road from Heredia to Puerto Viejo (Sarapiqui Dist.), Heredia Prov., Costa Rica (320 m). V–1–1965, DHJ, no. 819, fl.

4. 1 km W La Virgin, Sarapiqui Dist., Heredia Prov., Costa Rica (80 m). IX–5–1971, DHJ, nc.

5. 1 km W La Lola (on Limon-Siquirres railroad), Limon Prov., Costa Rica. VIII–19–1963, DHJ, nos. 1868, 1872, 1873, 1874, 1875, 1876, 1877, 1878, 1881.

6. 11 mi N Florencia, San Carlo Dist., Alajuela Prov., Costa Rica. VII–14–1963, DHJ, no. 1883, 1885, 1866.

7. Along Changuinola River, Changuinola Valley, Bocas del Toro Prov., Panama. II–29–1924, V. C. Dunlap, no. 462 (F 709739, US 1405662) (listed incorrectly as “A. cookii” in Woodson and Schery), fl.

8. Lancetilla valley near Tela, Depto. Atlantida, Honduras. VI–6–1927 to III–20–1928, P. C. Standley, no. 54627 (US 1408382).

9. 1.9 mi W Puerto Matias de Galvez (on road to Escobas), Izabal, Guatemala. VIII–10–1967, DHJ, no. 1582.

10. 27.4 mi S Puerto Barrios, Izabal, Guatemala. VIII–8–1967, DHJ, nc.

NATURAL HISTORY

Acacia ruddiae is one of the most peculiar Central American members of the genus. It lives in very wet forest habitats, which is distinctly unlike other acacias, yet does not have the obligate acacia-ants which generally allow survival of swollen-thorn acacias in wet habitats (Janzen, 1966a).

In Costa Rica, where the only large populations of A. ruddiae have been encountered, the natural habitat of this tree is in primary succession along lowland rivers through rain forest (sites 4, 5), swamp edges (site 5), and in large, old landslide scars on the sides of very steep gorges (site 3). These areas receive 4 to 5 m of rain a year and have a poorly developed dry season of 1 to 2 months with occasional rains. Some of the sites occupied by A. melanoceras and A. allenii receive as much rainfall, but have a more noticeable dry season. Judging from the paucity of collections, A. ruddiae apparently also exists as a very rare tree in the rain forest that extends in a narrow strip up along the Caribbean coast to the Puerto Barrios area of Guatemala (sites 8 and 9), but there is no reason to believe that it does not become locally common in these forests just as it does in Costa Rica.

Acacia ruddiae is by no means generally distributed throughout the rain forest of the Costa Rican Caribbean lowlands. At site 4 it was common in 1- to 30-year-old primary succession on gravel river banks and grew to a 25-m-tall tree with 30 cm DBH; it had not, however, extended its population out into the surrounding old cacao groves and pastures. At site 5 there was a large population along about a mile of old river bank, and occasional large saplings in the immediately adjacent parts of a banana plantation and on swamp edges in primary forest. Many of the swamp trees were old and scattered widely, in contrast with the many adjacent and even-aged A. ruddiae along the river. Extensive search throughout many other swamps and disturbed forest habitats within ten miles located no other A. ruddiae. At site 3, A. ruddiae occurs only

in a small portion of the total landslide area in this strongly dissected topography. It has not spread along the road into other disturbed habitats from this site. While detailed weather data are not available for site 3, it may be slightly drier—hence better insolated—than other disturbed sites slightly above or below it in elevation. This is indicated by the high success of grasses in early stages of succession here, in contrast with dicot regeneration at casually cleared sites a few miles farther up or down the road.

Seedlings growing in deep shade resemble those of A. melanoceras and A. allenii in that the leaves are arranged in a whorl so that in vertical view, they do not overlap. When growing on bare gravel bars, the seedlings are more bushy and asymmetrical in appearance (Figure 24B). The nodules on the rootlets of A. ruddiae seedlings are more noticeable than on any other swollen-thorn acacia (Figure 24A).

The seedlings differ from those of all other swollen-thorn acacias in lacking swollen thorns. The first swollen thorns do not appear until the plant is 1 to 3 m tall and has several hundred leaves. The nectaries (Figure 8C), however, are produced as soon as on any swollen-thorn acacia. The first ones appear on the rachis just below the point where the pinnae join the rachis, and petiolar nectaries begin to appear on the 25th to about the 50th leaf. There are no Beltian bodies on A. ruddiae, and thus the only contact that the seedlings may have with ants is when occasional worker ants visit the nectaries and forage on the leaf surfaces. In this sense the young A. ruddiae are no different from nonant-acacias, except that the nectaries are larger than on nonant-acacias and may therefore produce more nectar.

On saplings, the first swollen thorns are scattered along the main trunk and lateral branches. They are very thin-walled, hollow, easily crushed, and

occasionally only half of the stipule is enlarged. They often split at the base upon drying (Figure 25B), and ants, especially Crematogaster, often use this as an entrance hole. As the tree grows larger, some of the lateral branches have no thorns at all (Figure 26A), some branches have large swollen thorns at the base and tufts of leaves at the ends (Figures 27A, 27B), and some branches are very short and bear thorns best classified as type B (Figures 27A, C). The A. ruddiae in the Puerto Barrios area (sites 8 and 9) had most of its swollen thorns on branches such as those in Figures 26B and 27, while in Costa Rica the branches were usually very short with many thorns (as in Figure 27A) or very long with just a few scattered thorns (Figure 28A) The lower trunk of a large, old A. ruddiae is covered with short branches with many thorns, making the tree very difficult to climb.

The crowns of A. ruddiae are more diffuse than any other swollen-thorn acacia and therefore look more like most mimosaceous legumes than do swollen-thorn acacias. Looking up through a crown (e.g., Figure 28B), one can often see large amounts of sky; only the most densely leafed branches (e.g., Figure 51A) look like typical swollen-thorn acacia branches. This is the case despite the fact that the crowns of A. ruddiae are usually fully insolated; while the young seedlings appear to grow fairly well in heavy shade, the mature trees almost always have canopy level or emergent crowns, suggesting that A. ruddiae probably cannot mature in heavy or broken shade.

The trees of A. ruddiae do not bear flowers until they are in the 8 to 15 m height and 15 to 30 cm DBH class. Small clusters of yellow and bee-pollinated inflorescences (Figure 25A) are scattered through the crown and not concentrated on main sexual branches as is common for swollen-thorn acacias. The seed pods appear to mature during the 2 to 4 months immediately following flowering, as with A. melanoceras and A. allenii, rather than waiting until the following dry season as with dryland, swollen-thorn acacias.

Adults of A. ruddiae show no more obvious insect damage than do other acacias not occupied by obligate acacia-ants. Since some have no ants living on them at all—though some worker ants forage in the foliage for nectar and probably insects—this immunity is likely due to the usual kinds of chemical defenses encountered in ordinary acacias. Shoot tips of A. ruddiae are slow growing in comparison with more usual swollen-thorn acacias. Associated with this they are very tough, compact, bitter in taste, and covered with dense hair (Figure 5A). They strongly resemble those of A. macracantha and A. farnesiana.

By virtue of the very well developed foliar nectaries and enlarged thorns, A. ruddiae has far more ants foraging on its leaf surfaces than ordinary nonant-acacias. None of these ants are obligate acacia-ants, but rather those species—Pseudomyrmex, Crematogaster, Camponotus—that live throughout the general forest canopy and that forage for dead, moribund, or live insects on foliage in general. While queens of obligate acacia-ants may occasionally be dispersed into habitats with A. ruddiae (definitely the case at sites 9 and 10 in Guatemala), I have never seen one of the characteristic type of entrance holes made by these queens in about 2,500 swollen-thorns examined carefully on 16 individual trees of A. ruddiae.

To examine the use made of A. ruddiae thorns by ants, a representative 3 m tall, fully insolated sapling growing in river-edge vegetation at site 4 was completely dissected. The acacia had 798 leaves on 51 branches; far less than 20 percent of the large leaves produced by the acacia were subtended by swollen thorns. Of the 371 thorns, 302 were, or had been, tenanted by some insect (on a “typical” swollen-thorn acacia, all thorns would have been). Nineteen percent of the thorns had split naturally at the base, and all of these thorns had been entered by insects. Of all the thorns that had been entered by insects, by any means, 20 percent had parts of ant colonies in them or queens attempting to start a colony. Of the remaining thorns that had been entered, 95 percent contained microlepidopterous larvae, weevil larvae, and Orthoptera eggs. When these insects emerge, the acacia thorns are very likely to be occupied by ants. Spiders and other insects tenanted the remaining 5 percent of the thorns lacking ants. The ants represented at least 7 species in 5 genera—Solenopsis, Crematogaster, Camponotus, Pseudomyrmex, Paracryptocerus. While a detailed census is not available, cursory examination of A. ruddiae thorns in large old trees shows almost every thorn to have a part of an ant colony in it. It is extremely likely that the continual foraging of these ant species on the foliage occasionally results in less herbivore damage to the tree than if the ants were not present.

If a mutant strain of A. ruddiae provided any type of solid food to the ants, the probability is very high that it would quickly lead to the evolution of a strongly mutualistic interaction. It is impossible to know if A. ruddiae represents the gradual degeneration of a once intense mutualism or is on its way to more mutualistic interaction. The latter is more likely, since the former would require a strong relaxation of selection pressure by herbivores, such as might occur if a swollen-thorn acacia were to immigrate to a small oceanic island. The evolution of a type of Beltian body may be starting to occur with A. ruddiae; some of the plants at site 4 have clear tips on the pinnules, extending into the leaf about 0.25 to 0.5 mm. If the ants were to start to chew on this for moisture or nutrients during the time of year when insect prey is scarcest, and therefore maintain larger colony sizes than on plants lacking this modified pinnule tip, the evolution of a food body might occur. This is likely to be the way that typical Beltian bodies first appeared on acacias. Plants with a larger year-round population of ants would probably be less damaged by the periodic small outbreaks of host-specific insects that probably occur on many tropical trees, and certainly occur with nonant-acacias. Interestingly, the species of ant that began this behavior would be likely to concentrate more of its foraging on the acacia and less off, and be likely to tenant an increasingly greater proportion of the thorns on the acacia. This could occur either through faster colony growth than could occur with ant species totally dependent on insect prey for growth, or through the ants moving into thorns left untenanted during periodic population lows in the other species of ants. Once the entire ant colony becomes concentrated in the acacia, it has a vested interest in the health of the acacia and coevolution of the interaction should occur at an ever increasing rate (Janzen, 1966a). As will be discussed further in respect to hybrids, this acacia appears to be a plant that would have greatly increased fitness by the gaining of the genetic information necessary for the production of Beltian bodies. This could well happen through introgression with a wet-forest acacia such as A. melanoceras at the southern end of the range of A. ruddiae, or A. gentlei or A. cookii, and the northern end of the range of A. ruddiae.