Comprehensive Description
provided by Smithsonian Contributions to Botany
Acacia cornigera (L.) Willdenow
Acacia cornigera (L.) Willdenow, 1806:1080.
Mimosa cornigera L., 1753:520.
Acacia cornigera var. americana Dc., 1825:460.
Acacia spadicigera Schlechtendal and Chamisso, 1830:594.
Acacia cubensis Schenck, 1913:360.
Acacia nicoyensis Schenck, 1913:360.
Acacia hernandezii Safford, 1914:358.
Acacia furcella Safford, 1914:359.
Tauroceras spadicigerum (Schlechtendal and Chamisso) Britton and Rose, 1928:85.
Tauroceras cornigerum (L.) Britton and Rose, 1928:86 [excluding synonymy].
Acacia interjecta Schenck, 1913:361.
Acacia rossiana Schenck, 1913:361.
Acacia campecheana Schenck, 1913:361.
Acacia turgida Safford in Wheeler, 142:plate 45.
Acacia cornigera is the best known of the swollen-thorn acacias. It is readily distinguished from all other swollen-thorn acacias by the unique combination of elongated “canoe-shaped” petiolar nectaries (Figure 8B), very elongated inflorescences (Figure 12A), thorns usually round in cross section, and completely indehiscent seed pods.
Linnaeus’ type specimen of A. cornigera is a sterile branch representative of the eastern Mexican portion of the total population; it was collected from a cultivated plant in Holland, presumably grown from Mexican seed (Rudd, 1964). Rudd has compared pinnules of the type with those of A. sphaerocephala and found that the pinnules of the type have the readily visible secondary venation lacking on A. sphaerocephala
leaflets. Owing to the peculiar thorns of A. mayana, there is no doubt that the type does not belong to that species.
The type specimen of A. spadicigera (site 1) is easily identified as A. cornigera because of its elongate inflorescences and thorns round in cross section. Its straight short thorns are representative of A. cornigera from lowland wet sites in Veracruz. Safford’s type specimen of A. hernandezii is unmistakably from the northern wet lowlands portion of the A. cornigera population. A. cornigera is the only species of swollen-thorn acacia in the Rascon area (site 2). Safford’s leafless type specimen of A. furcella is representative of the A. cornigera population in the Lake Catemaco region (site 3). The type has the very dark brown thorns and widely spread type B thorns characteristic of high-elevation, eastern Mexican A. cornigera. The inflorescences of the types of both A. hernandezii and A. furcella are representative of A. cornigera inflorescences.
Schenck’s type of A. rossiana (site 4) is representative of that portion of the A. cornigera population inland from Minatitlan, Veracruz, Mexico. I have been unable to locate his type specimen for A. campecheana; possibly it was destroyed at Berlin. He states that its leaves are like those of A. spadicigera and A. rossiana, and it was therefore most likely collected from that part of the A. cornigera population that extends along the coastal lowlands from Veracruz to Yucatan. Its name suggests that it was collected in Campeche, and there appear to be no swollen-thorn acacias in that area that could be confused with A. cornigera. Schenck described A. interjecta from material growing in the Singapore and Kew botanic gardens; the description agrees with that of A. cornigera and the seed probably came from Mexico, since that appears to have been the source of most introduced A. cornigera seed. Schenck described A. nicoyensis from material representative of the southern end of the geographic range of A. cornigera (site 77). His type specimen has the white and slightly flattened type A thorn found on Costa Rican A. cornigera growing in very dry and insolated sites. The inflorescences and seed pods of the type of A. nicoyensis are unmistakably A. cornigera.
Schenck’s type specimen of A. cubensis was taken from a cultivated plant (site 82); the white and partially recurved type B thorns indicate that the seed came from the A. cornigera population some where north of Guatemala. A. cornigera apparently occurs only in gardens in Cuba and its presence unlikely is due to natural introduction by birds or other means. The absence of ant entrance holes in the thorns of the type specimen of A. cubensis indicates that the ants have not yet been introduced to Cuba.
Since A. turgida is supported only by a photograph (Wheeler, 1942), it probably represents a name that Safford intended to propose. I cannot find the locality (site 5) for the specimen, but it appears to have been collected from that rather distinctive portion of the A. cornigera population growing where the Pan-American highway crosses the Guatemala-Mexico border.
KEY COLLECTION LOCALITIES*
Mexico:
1. Type of A. spadicigera: La Laguna Verde, Veracruz, Mexico. March 1820, Schiede, no. 685.
2. Type of A. hernandezii: Rascon, S.L.P., Mexico. VII–19–22–1905, E. Palmer, no. 669.
3. Type of A. furcella: Catamaco, Veracruz, Mexico (300 m). IV–26–1894, E.W. Nelson, no. 427.
4. Type of A. rossiana: Santa Lucrezia, Isthmus of Tehuantepec, Veracruz, Mexico. X–8–1906, H. Ross, no. 918.
5. “Type” of A. turgida: Herba Santa, Chiapas, Mexico (see plate 45 in Wheeler, Bull. Mus. Comp. Zool. Harvard, 90:1–262).
6. Guatemala-Mexico border (near Ciudad Cuahtemoc) to 19.0 mi N on Hgy. 190, Chiapas (720–600–780 m). VIII–7–1966, DHJ, no. 499; VIII–16–1967, DHJ, nc (common to rare).
7. 2.6 mi NE to 0.9 mi W Chiapa de Corzo on Hgy. 190, Chiapas (580–475 m). VIII–17–1967, DHJ, nc (common to rare).
8. 9.7 mi W Tuxtla Gutierrez to 8 mi NE Cintalapa on Hgy. Chiapas 190 (860–1090–570 m). VIII–18–1967, DHJ, nc (common to rare to occasional).
9. 13.5 to 9.0 mi E of intersection of Hgy. 190 and 185 (La Ventosa), Oaxaca (80 m). VIII–18–1967, DHJ, nc (com mon to rare).
10. 9.9 to 23.4 mi N of intersection of Hgy. 190 and 185 (La Ventosa), Oaxaca (210–290–250 m). VIII–18–1967, DHJ, nc (common to occasional).
11. 33.3 mi N of intersection of Hgy. 190 and 185 (La Ventosa) to Acayucan, Oaxaca-Veracruz (230–40 m). VIII–18–1967, DHJ, nc (common).
12. Near Salina Cruz, Oaxaca (sea level). X–3–1965, DHJ, nc (occasional).
13. Coatzacoalcos to Veracruz on Hgy. 180, Veracruz (0–350–0 m). III–21–1964, DHJ, no. 1935; VI–15–1966, DHJ, nc (occasional to common).
14. 8.3 mi SE bridge over Rio Coatzacoalcos, Veracruz, to Cardenas, Tabasco, on Hgy. 180 (0–40 m). VI–15–1966, DHJ, nc (very rare, about 1–2 plants per 2 miles).
15. Cardenas to 37.1 mi S, Tabasco (40–100 m). VI–17–1966, DHJ, nc (rare to common).
16. Cardenas to Paraiso, Tabasco (40–10 m). VI–17–1966, DHJ, nc (rare).
17. Cardenas to Ciudad del Carmen, Tabasco-Campeche (40–0 m). VI–22–1966, DHJ, no. 494 (rare, and only on well drained sites), fl.
18. Campeche to Ruinas Edzna (29 mi E on Hgy. 180), Campeche (0–50 m). VI–22–1966, DHJ, nc (very rare to occasional).
19. Hopelchen to Ticul on Hgy. 180, Campeche-Yucatan 50–80). VI–28–1966, DHJ, nc (very rare to occasional).
20. 12.2 mi S Peto (Santa Rosa) on Hgy. to Cd. Chetumal, Yucatan-Quintana Roo (50–80 m). VI–29–1966, DHJ, nc (common), fl.
21. Cd. Chetumal to 69.6 mi NW on Hgy. to Peto, Quintana Roo (0–20 m). VI–30–1966, DHJ, nc (common), fl.
22. 30 mi S to Campeche on Hgy. 180, Campeche (0–80–0 m). VI–25–1966, DHJ, no. 491; VII–8–1966, DHJ, nc (rare to common).
23. 3.2 mi E Cruz Grande to 19.1 mi NW Puerto Escondido on Hgy. 200, Guerrero-Oaxaca (50–250–70 m). VI–18–1967, DHJ, nc (rare to common to occasional).
24. 9.3 mi N Playa Pie de la Cuesta to 16.5 mi SE Papanoa on Hgy. 200, Guerrero (0–20 m). VI–17–1967, DHJ, nc (very common in swamps, absent on dry hills).
25. Alvarado to Tlacotalpan to Cd. Aleman to Las Tinajas to Paso del Toro, Veracruz (0–100–0 m). Throughout the year, DHJ, no. 1939 (common), flower February-May.
26. Tierra Blanca to La Granja to Temascal, Veracruz-Oaxaca (20–70 m). Throughout the year, DHJ, nos. 1966, 1967, 1968, 1969, 1970, 1971, 1972, 1973, 1974, 1975, 1976, 1977, 1978, 1979, 1980, 1981 (very common).
27. Shores of the lake behind Presa Miguel Aleman, Temascal, Oaxaca (70 m). X–20–1963, DHJ, no. 1937; XI–3–1963, DHJ, nc (occasional to very common).
28. Cd. Aleman to Loma Bonita to Rio Playa Vicente, Veracruz-Oaxaca-Veracruz (50–100 m). XI–4–1963, DHJ, no. 1938 (common to very rare on red clay hills).
29. Cd. Aleman, Veracruz to Valle Nacional, Oaxaca, on Hgy. to Oaxaca, XI–6–1963, DHJ, nc (common to occasional).
30. 22.8 to 27.5 mi SW Conejos (turn off of Hgy. 140 to Huatusco) on Hgy. to Huatusco to Cordoba, Veracruz. XI–10–1963, DHJ, nc (common to rare).
31. 4.3 mi E Cordoba, Veracruz on Hgy. 150, Veracruz (800–0 m). X–16–1963, DHJ, nc (rare to very common).
32. Hgy. 150 SE to San Ignacio de la Llave, Veracruz (60–10 m). X–16–1963, DHJ, nc (common).
33. Tampico to Tuxpam on Hgy. 105, Veracruz (0–100–0 m). 1–2–1964, DHJ, nos.1942, 1943; VI–13–1966, DHJ, nc (common to rare).
34. Tuxpan to 1.2 mi S Villa A. Camacho on Hgy. 130, Vera cruz. 1–2–1964, DHJ, no. 1940; VI–13–1966, DHJ, nc (occasional to rare).
35. Poza Rica to Nautla on Hgy. 180, Veracruz (50–0 m), 1–20–1964, DHJ, nc (occasional to rare).
36. 6 mi SW Tamazunchale to Cd. Valles on Hgy. 85, San Luis Potosi (500–100 m). 1–1–1964, DHJ, no. 1936; XII–31–1963, DHJ, nc (rare to common).
37. 36.6 mi E Rio Verde to Cd. Valles on Hgy. 86, San Luis Potosi (1140–100 m). VIII–24–1967, DHJ, nc (rare to common).
38. Cd. Valles to 6.4 mi N Antiguo Morelos on Hgy. 85, San Luis Potosi-Tamaulipas (100–250 m). VIII–24–1967, DHJ, nc (rare).
39. Antiguo Morelos to 20 mi NW Cd. del Maiz, Tamaulipas-San Luis Potosi (250–? m). XII–31–1963, DHJ, nc (common on wetter hillsides).
40. Cd. Valles to Tampico on Hgy. 70, San Luis Potosi-Veracruz (100–10 m). VI–11–1966, DHJ, nos. 496, 497; VIII–24–1967, DHJ, nc (common to occasional).
41. Pt. Talisman (Guatemala-Mexico border) to 3 mi N Arriaga on Hgy. 200, Chiapas (50–150 m). 1–11–1965, DHJ, no. 1047 (common).
42. Presa Benito Juarez, near Jalapa de Marquez on Hgy. 190, Oaxaca (200 m). IX–7–1968, DHJ, nc (1 plant).
43. Pichucalco, Chiapas. 1–1907, G.N. Collins and C.B. Doyle, no. 260 (US 692169) mp.
44. 3 mi S Pinola Las Rosas along road to Soyatitan, Venustiano Carranza (W of Comitan), Chiapas (1400 m). VII–27–1965, D.E. Breedlove, no. 11367 (US 2484541), fl.
45. Zacualpan, Veracruz (1000 m). VII–1917, C.A. Purpus, no. 7748 (US 884536), fl., and IX–1908, H. Schenck, no. 836 (US 764665).
46. Fortuno, Rio Coatzacoalcos. III–1937, L. Williams (F 897476), fl.
British Honduras:
47. Sta. Elena to Belize on main Hgy. (5–0 m). VI–30–1966, DHJ, nc (rare to occasional), fl.
48. El Cayo, El Cayo Prov. V–5–1931, H.H. Bartlett, no. 13007 (US 1493273), mp.
Guatemala:
49. Escuintla to Taxisco along Hgy. CA–7, Depto. Escuintla-Santa Rosa (330–100 m). 1–13–1965, DHJ, nc (occasional).
50. Escuintla, Dcpto. Escuintla, to Puente Talisman, Depto. San Marcos (Guatamela-Mexico border) on Hgy. CA–7 (330–50 m). VII–16–1966, DHJ, no. 466; 1–11–1965, DHJ, nos. 1045, 1046 (common to occasional), fl.
51. Retalhuleu SW to Champerico, Depto. Retalhuleu (200–0 m). VIII–1–1966, DHJ, nc (occasional).
52. Retaljuleu to 14.0 mi NE on Hgy. S, Depto. Retalhuleu (200–900 m). VII–28–1966, DHJ, no. 464 (common to rare).
53. Escuintla to 7.4 mi NE on Hgy. 3, Depto. Escuintla (330–880 m). VIII–13–1967, DHJ, nc (common to occasional).
54. 0.3 mi SW to 4.1 mi NE Los Amates on Hgy. CA–9 (150–110 m). VIII–9–1967, DHJ, nc (common).
55. Finca San Joaquin, Slopes above Rio Negro, near San Cristobal de Verapaz, Depto. Alta Verapaz (1420–1080 m). VIII–14–1967, DHJ, nc (rare to common).
56. 14.9 mi SE to Guatemala-Mexico border (near Ciudad Cuauhtemoc), Depto. Huehuetenango (900–1000–720 m). VIII–6–1967, DHJ, nc (common).
57. Patalul, Depto. Solola. 11–14–1906, W.A. Kellerman, no. 5915 (US 578841), fl.
58. Finca Sepacuite, Depto. Alto Verapaz. IV–1905, O.F. Cook and R.F. Griggs, no. 8 (US 590344), mp.
59. 1–2 mi N Ocos, Depto. San Marcos (sea level). III–15–1940. J.A. Steyermark, no. 37867 (F 1035247).
Honduras:
60. 12.5 to 19.5 mi NW Siguatepec on Tegucigalpa-Puerto Cortes Hgy., Depto. Comayagua (910–540 m). VII–31–1967, DHJ, nc (common to very rare).
61. 0.7 mi N Rio Undo to 18.6 mi S Puerto Cortes on Tegucigalpa-Puerto Cortes Hgy., Depto. Cortes (90–10 m). VIII–2–1967, DHJ, nc (common to occasional).
62. San Pedro Sula to 8.1 mi SW on Hgy. to Santa Rosa, Depto. Cortes (90–130 m). VIII–2–1967, DHJ, nc (occasional along streams).
63.20.6 mi NE to 28.4 mi SW Sula on San Pedro Sula-Santa Rosa Hgy., Depto. Santa Barbara (260–780 m). VIII–3–1967, DHJ, nc (common to rare), fl.
64. Vegas del Rio Agua, 3 km from Yoro, Depto. Yoro (1000 m). V–8–1956, A. Molina, no. 6807 (US 2412019), fl.
EI Salvador:
65. (Depto. Cuscatlan) San Salvador on San Salvador-Nueva Ocotopeque Hgy. (420–590 m). VIII–4–1967, DHJ, nc (common to rare).
66. 6 mi E San Salvador on Interamerican Hgy., Depto. San Salvador (650 m). 1–14–1965, DHJ, nc (occasional), fl.
67. 29.7 to 31.6 mi E La Libertad on La Libertad-La Union Hgy., Depto. La Paz (50–80 m). VIII–5–1967, DHJ, nc (rare).
68. Intersection of Hgy. from La Herradura with La Libertad-La Union Hgy., Depto. La Paz (20 m). VIII–5–1967, DHJ, nc (1 plant).
69. 15.2 to 10.2 mi SE Santa Ana on Interamerican Hgy., Depto. Santa Ana (440–700 m). VIII–6–1967, DHJ, nc (common).
70. San Vicente, Depto. San Vicente (350–500 m). III–2–1922, P.C. Standley, no. 21687 (US 137448, US 1137096), fl.
71. Ateos, Depto. La Libertad. IV–17–1922, P.C Standley, no. 23360 (US 1139047), fl.
72. San Martin to Laguna de Ilopango, Depto. San Salvador. IV–1–1922, P.C. Standley, no. 22580 (US 1138317), fl.
Nicaragua:
73. Penas Blancas (Depto. Rivas) to 7.2 mi NW on Hgy. 2 (50 m). VII–25–1967, DHJ, nc (common along rivers).
74. 3.4 mi SE Leon on Hgy. 12, Depto. Leon (20 m). VI–27–1967, DHJ, nc (30 acre patch), fl.
75. 1.2 mi W Matagalpa to 3.6 mi N on Hgy. 3, Depto. Matagalpa (650–880 m). VII–28–1967, DHJ, nc (common), fl.
76. 8 mi S Esteli on Hgy. 1 (940 m). VII–28–1967, DHJ, nc (I plant).
Costa Rica:
77. Type of A. nicoyensis: Shores of the Gulf of Nicoya, Guanacaste, Costa Rica (sea level). February 1900, Tonduz, no. 13538.
78. 20 mi NW turnoff of Interamerican Hgy. to Puntarenas, to Penas Blancas (border of Costa Rica and Nicaragua) on Interamerican Hgy., Guanacaste Prov. (10–100 m). January-September, periodically in 1963–1967, DHJ, nc (rare to common, patchy distribution and associated with wet spots), fl., February-April.
79. Liberia to Playa Coco along Hgy. 21, Guanacaste Prov. (100–0 m). January-September, periodically in 1963–1967, DHJ, nc (rare to very common, primarily in wet spots), fl. and mp., in February.
80. Hda. Palo Verde, COMELCO, Bagaces, Guanacaste Prov. (sea level). IV–2–1966, DHJ, nc (occasional).
81. Filadelphia, Guanacaste Prov. III–2–1946, Echeverria, no. 279 (F 1442853), mp.
Cuba:
82. Type of A. cubensis: North coast of Cuba. IV–21–1963, C. Wright, no. 2402.
NATURAL HISTORY
Acacia cornigera (Figure 70) and A. collinsii have the greatest geographic and ecological range of all the swollen-thorn acacias. The riparian and swamp habitats that were occupied by A. cornigera previous to human interference are in lowland areas that are ideal for slash-burn milpa agriculture. A. cornigera has probably been common in lowland agriculture fallow fields as long as Indians have lived in the Central American lowlands. Since occupied A. cornigera may reach reproductive maturity in 3 to 4 years, and since birds, people, and cattle distribute its seeds widely in second growth vegetation, A. cornigera populations have expanded widely into pastures, roadsides, and short-fallow fields generated by contemporary agricultural practices. At many locations, A. cornigera is found in all types of disturbance sites, from creek bottoms to hilltops. Only A. sphaerocephala and A. collinsii are on occasion found in drier habitats than A. cornigera. The natural history of A. cornigera in eastern Mexico has been described in detail (Janzen, 1967b), and the comments in the remainder of this section are designed to highlight the way it differs in other parts of its range from that portion of the population.
Local short-distance dispersals of A. cornigera by
humans make it difficult to understand its pre-Colombian range. For example, at site 42 in western Mexico, a single A. cornigera was found growing by a new hydroelectric dam. The closest native population of A. cornigera is several hundred miles away in the Isthmus of Tehuantepec and along the west coast in the Acapulco region (sites 23, 24) (also an introduction?). This plant probably arrived by either road-building machinery or through humans; farmers and Indians often eat the sweet pulp around the seeds and ingest many seeds simultaneously (Janzen, 1967b). The work crews on public construction in Mexico often come from several hundred miles away. If more than one plant was introduced in the area of the dam, it is likely that viable seeds will be produced. The bruchids that normally kill most A. cornigera seeds will be absent, at least at first, and this swollen-thorn acacia will likely spread down the watercourse away from the dam. It will probably not spread around the lake since the general area of the lake is extremely dry and water levels will fluctuate greatly. The single plant was occupied by Pseudomyrmex ferruginea. The queen probably came from the population of A. collinsii about 25 miles to the south or the population of A. hindsii 50 miles to the northwest. As a second example, in the central Yucatan Peninsula, A. cornigera is only common around small settlements, cattle corrals, and old Indian ruins (sites 18 to 21). These small sites are also usually the wetter habitats in the region (the outlying areas are heavily occupied by A. collinsii), and it is impossible to know if the presence of A. cornigera is due to recent and frequent introduction by humans and cattle, or by natural dispersal by birds.
Deliberate long-distance introduction of A. cornigera may have some interesting repercussions in the future. Introduced to the southern tip of Florida as a garden plant, A. cornigera has occasionally escaped to form feral populations (e.g., Oneco and Chapman Field, US 1433083 and US 764685 in the U.S. National Herbarium). These plants are severely damaged by the frosts that occasionally damage orange orchards, but they do flower and set fertile seeds if there are several together. Obligate acacia-ants have not been introduced, but it would be interesting to examine the interaction of occupied swollen-thorn acacias with the subtropical communities of southern Florida. Reproductively isolated from the Mexican parent genotype, A. cornigera might even develop a frost-resistant population in Florida. A. cornigera has also escaped on Martinique, Guadeloupe, and Cuba in the Caribbean (US 2229384, US 1714928, US 764657). Here, with a climate not unlike that of some parts of eastern lowland Mexico, the populations (again without their ants) have a chance to expand into the native vegetation just as many other species of introduced trees have done on Caribbean islands. The central question here is whether the herbivore pressure in the island communities is strong enough to require occupation by obligate acacia-ants for the survival of a large breeding population of A. cornigera. All introductions to the Old World tropics appear to have been to botanical gardens, and I have no information on their escape into natural habitats.
Acacia cornigera and A. collinsii occupy adjacent habitats over much of their range. Always A. cornigera is in the wetter creek bottoms and swamp edges while A. collinsii is on the drier hillsides (e.g., Guanacaste Province, Costa Rica). In the wetter parts of the range, A. hindsii and A. cornigera often occur together (e.g., Pacific foothills of Guatemala) and here, A. cornigera is the drier hillside plant with A. hindsii along the watercourses. The distinctiveness of these separations by habitat was probably once very clear. Current agricultural practices have rendered wetter habitats drier and more inso-lated by clearing the vegetation, with the consequence that the formerly separated swollen-thorn acacias now grow side by side.
In the southern part of its range, the A. cornigera population is much more broken up into sub populations than in eastern Mexico where it is found in all sorts of disturbance and climatic regimes from sea level to 1200 m, and from areas with a 7–month dry season to areas with almost no obvious dry season. In habitats where one expects to find A. cornigera in the northern part of the range, one instead encounters A. collinsii and A. hindsii. To determine the role of competitive exclusion in this system, we need much more information on the parasites and predators (specifically, bruchid beetles) that these species have in common, and the ability of each to grow and maintain the largest possible obligate acacia-ant colony in the shortest period of time.
Why A. cornigera has stopped its southward movement in Guanacaste Province in Costa Rica is very enigmatic. A. collinsii continues south to Colombia with a nearly continuous population that occupies many habitats similar to those occupied by A. cornigera further north. It may well be that it has not stopped at all, but rather that it is in the process of slowly moving southward. The only possible ecological explanation that comes readily to mind is that the Guanacaste plants of A. cornigera are specialists at surviving the severe dry season there and therefore cannot advance slowly through the wet-forest area immediately down the coast from Puntarenas. What is needed is a longdistance dispersal of more than one seed simultaneously from Guanacaste to the dry Pacific coast of Panama.
The dependency of A. cornigera on a colony of obligate acacia-ants for survival to reproductive maturity has been well documented for lowland eastern Mexican habitats (Janzen, 1967a). Judging from the badly damaged condition of naturally unoccupied A. cornigera throughout its range, this dependency is the general case. It is clear, however, that the intensity of damage to unoccupied A. cornigera varies from one habitat to another, just as it did from one season to another in the lowlands of eastern Mexico. High-elevation sites subjected to wet, cool, and windy weather (e.g., site 75) produce the healthiest appearing unoccupied A. cornigera, but even here it is doubtful if an unoccupied tree can survive until reproductive maturity unless growing in a cluster of acacias in a large and well-grazed cattle pasture. In central highland Guatemala (e.g., site 53) and on the highland Atlantic side of Honduras (site 60), A. cornigera is occasionally “occupied” by a large colony of Carnponotus planatus which appear to compete directly for thorns with the obligate acacia-ants colony. The C. planatus patrol the acacia quite effectively and the plants are healthy in appearance. How long an acacia can be occupied by C. planatus and survive is not clear. On some occasions, a colony of C. planatus and one of obligate acacia-ants occupy the same acacia, but usually in different branches. C. planatus occurs throughout much of the range of swollen-thorn acacias but in other areas does not appear to offer any protection to the acacia. It also lives in hollow twigs in many other species of tree.
Acacia cornigera has gained the same “freedom” from herbivorous insects gained by A. hindsii in areas of intense agriculture (e.g., site 74 in El Salvador); however, it is much rarer than A. hindsii in these sites and will probably become extinct in the near future. Seedlings are almost never encountered. Without the ants to keep the vines off, the stiff and upright young shrub of A. cornigera
makes an excellent vine standard leading to its death by shading. The young plants of A. hindsi are more shade tolerant and less rigid than A cornigera of the same age.
As with A. hindsii and A. collinsii, the seedlings of A. cornigera are scattered throughout the habitat occupied by the parents. As stressed earlier (Janzen 1967b), they attain the height of 10 to 50 cm and die after 1 to 2 years if not occupied by a colon) of obligate acacia-ants. When in heavily shaded sites, the seedlings tend to produce few swollen thorns (Figure 71A), while in fully insolated sites thorn production is much greater (Figure 71B) Once occupied, a variety of different shrub life forms may develop. In dry areas, about as much wood goes into the central trunk as into branches (e.g., Figure 72A), and the shrub ends up being very bushy in appearance (Figures 73A, B). Only very rarely does A. cornigera become a large tree 10 to 15 m in height; however, such trees are normall) found in natural disturbance sites along rivers and
may at one time have made up a much larger proportion of the reproducing population than at present.
Over most of the distribution of A. cornigera, sexual branchlets and their products are rather uniform; however, wetness of the habitat does appear to influence this. In the Pacific foothills of Guatemala (site 50) the sexual branchlets may be 20 to 50 cm long (Figure 74A). In contrast to the situation at this wet site, most populations of A. cornigera produce sexual branchlets 1 to 10 cm long (Figure 74C). The inflorescence itself is not influenced by the amount of moisture at the site, except that at some of the driest sites, some of the shortest inflorescences were found (again suggesting that A. sphaerocephala is a dry-land offshoot of A. cornigera). The number of florets per inflorescence is poorly correlated with the length of the inflorescence (Figure 75) at any one site but over a variety of sites shows a weak correlation (Figure 76).
The indehiscent pods of A. cornigera are split open by birds (Figure 75B) and the seeds dispersed by birds that frequent second-growth vegetation (Janzen, 1967b). Over most of the range of A. cornigera, the pod walls are thick and hard to break between ones fingers. They tend to have sharp and stiff long points on the end (Figures 74B, C); however, in riparian habitats at site 26, the pod walls are very thin and even split of their own accord when drying. The plants on the adjacent dry hills have the usual thick-walled pods. This difference is genetic as shown by a lack of intermediates
and occasional plants bearing the “wrong” type of pods for the habitat where the plant is encountered. It is interesting that this site is only about 80 miles from that portion of the A. sphaerocephala population below the city of Veracruz that also has very thin pod walls. The number of seeds in A. cornigera pods is representative for dry-land, swollen-thorn acacias (Figure 77) but nearly all the seeds are killed by bruchid beetles. The very high predation on A. cornigera seeds by these beetles may be due to the high concentration of pod crops to be found in fields and roadsides; when the plants were widely spaced in natural disturbance sites, the bruchids might have a much more difficult time locating the pod crops.
The thorns of A. cornigera display every color found in swollen-thorn acacias, but there is little intertree or intratree variation at any one site. In general the thorns in very hot and dry lowlands are white, ivory, or pale gray. At high elevations they are often very dark brown or nearly black. Though the correlation is not perfect, it is tempting to interpret these differences as being adaptive in respect to the ant colony. In lowland areas, temperatures in fully insolated thorns can be high enough to drive out the ants, and light-colored thorns may reflect radiation and slightly lower the temperature inside the thorn. It is perhaps significant in this connection that A. collinsii also has very light-colored thorns in dry and hot areas and dark thorns in high and wet sites. The dark thorns of A. cornigera could aid in warming up the ant colony on cool days and thereby increase the degree of patrolling by ants. It is significant in this respect that A. sphaerocephala has very dark thorns at the one high (cool) and dry site where it occurs, in strong contrast to the light-colored thorns it bears in dry lowland areas. All of the wet-forest swollen-thorn
acacias have very dark brown thorns, and this may be important in warming the thorns by occasional direct insolation. Bicolored thorns (Figure 78C) are somewhat puzzling in this respect. They are most prominent in the small and insolated population of A. cornigera on the west coast of Mexico; here A. cornigera grows only in swamps, with A. collinsii on the extremely dry and hot adjacent hillsides.
The spectacular “bull’s horns” swollen-thorns of A. cornigera are extremely variable in shape, but in a consistent and often clinal manner. Many sites are characterized by a particular thorn type, and many of these thorn types resemble thorns found on other species of swollen-thorn acacia (though often not sympatrically). Type A thorns of A. cornigera (Figures 78, 79) are usually round or nearly so in cross section and form a shallow to steep “V” with the branch wearing them. When flattened so that they superficially resemble A. hindsii thorns (Figure 80D), they are still easily distinguished by having very thick and rough walls. Type B thorns (Figures 80, 81) are highly variable in form, but all are much larger in volume than type A thorns. There is no part of the range of A. cornigera where it is consistently occupied by multiple-queen species of obligate acacia-ants, so
selection for large type B thorns is omnipresent. Type B thorns of A. cornigera are the hardest to break into of any of the swollen-thorn acacias. Several are often twisted together, making it difficult to even pull one off the acacia.
- bibliographic citation
- Janzen, Daniel H. 1974. "Swollen-Thorn Acacias of Central America." Smithsonian Contributions to Botany. 1-131. https://doi.org/10.5479/si.0081024X.13
