Azteca

The paper "Azteca ants in Cecropia trees: taxonomy, colony structure, and behaviour" (JT Longino, 1991) discusses what is known about the different species of Aztea ants that inhabit Cecropia trees.

The dolichoderine genus Azteca is a strictly neotropical group of arboreal ants (Emery 1893, Forel 1928). They are abundant in lowland habitats from Mexico to Argentina, occurring as both generalized foragers and as specialized inhabitants of myrmecophytic plants. Azteca species exhibit a variety of nesting habits, including the construction of carton nests, the occupation of live and dead plant stems (Forel 1899, Ule 1901, Emery 1913, Davidson 1988, Ayala et al. 1996), and the formation of ant gardens. Ant gardens are arboreal ant nests which sprout epiphytes from carton nest material (Ule 1901, Wheeler 1921, Longino 1986, Davidson 1988, Corbara et al. 1999, Kaufmann & Maschwitz 2006). Striking cases of symbiosis occur between Azteca and highly specialized myrmecophytic plants, the most notable case being the relationship between Azteca and Cecropia ( Müller 1876, 1880-1881, Bequaert 1922, Wheeler 1942, Benson 1985, Longino 1991a, b). Also, Azteca ants have developed complex trophic relationships with many species of coccoid Hemiptera (Wheeler 1942, Johnson et al. 2001, Davidson et al. 2003). Azteca workers are often found tending mealy bugs (Pseudococcidae) and soft scales (Coccidae). For Azteca species that nest in live stems, the interior walls of the nest are often encrusted with mealy bugs and scales. Species building carton nests and ant gardens maintain dense populations of mealybugs and scales under the carton of the main nest or under small carton "pavilions " scattered over the vegetation. Very little attention has been paid to the taxonomic diversity of Coccoidea associated with Azteca , and usually only cursory observations of their presence are made during field collections. Because of the richness of the ecological interactions among Azteca , plants, and hemipteran symbionts, Azteca species have been and will continue to be subjects in the study of adaptation and coevolution, and therefore taxonomic work on the genus is particularly important.

The taxonomic bounds of the genus have not changed since its inception (Forel 1878, Shattuck 1992). Members of the genus can be recognized by the combination of (1) a thin, somewhat flexible cuticle, (2) anterolateral margins of clypeus extending anterior to mediolateral regions (with the exception of the aurita group, as reported here), (3) mandible with 7-9 teeth, (4) at least larger workers with cordate head shape, with margin of vertex concave, (5) surface sculpture (other than on mandibles) smooth, micropunctate, microalveolate, or combinations of these, (6) the total absence of coarse surface elements such as spines, tubercles, carinae, rugae, striations, or large puncta, (7) a distinctive petiole which is strongly sloping anteriorly and has a rounded posteroventral lobe, and (8) worker caste polymorphism.

The Asian genus Philidris (former Iridomyrmex cordatus group) is highly convergent with Azteca . In contrast to Azteca , the anterolateral margins of the clypeus are posterior to the mediolateral portions, and the mandible has 10-12 teeth. Male characters (Shattuck 1992) and recent molecular evidence (P. S. Ward, pers. com.) ally Philidris with other Asian dolichoderines and confirm that the similarity is due to convergence.

The relative clarity of the generic status of Azteca is not mirrored in species-level taxonomy. Several factors contribute to taxonomic confusion in Azteca , some historical, some biological. The only revision of the genus Azteca is that of Emery (1893). Over 140 species-group names were subsequently published by Forel, Wheeler, and others, with no attempts at revision. Many species were described from workers only, with no biological data. Since it is often particularly difficult to separate Azteca species with workers only (Longino 1991a, b, 1996), many named Azteca species are difficult to circumscribe.

Wheeler and Bequaert (1929) belatedly stated "Apparently the females [i.e., queens] furnish more reliable characters for identification than the workers in the genus Azteca ." An analogy can be drawn between the taxonomy of Azteca and the taxonomy of many plants. Botanists typically shun sterile material because it is often more plastic within species and less differentiated between species than reproductive material. Such is the case in Azteca . Workers are polymorphic within colonies, and colonies exhibit prolonged ontogenetic changes in worker morphology (pers. obs.). In contrast, queens are much less variable morphologically and exhibit strong interspecific differences. Within a single locality, species with strongly differentiated queens may have workers that are barely distinguishable.

Correlated with sharp differences in queen morphology are distinctive nesting habits. Nesting habits show great interspecific variation and little intraspecific variation. For example, queens of Cecropia-inhabiting species colonize very young Cecropia saplings. These queens are often very abundant in the environment, colonizing saplings and apparently competing for domination of saplings (Longino 1989b). I have made extensive collections of neotropical arboreal ants by breaking live and dead branches, searching for carton nests and ant gardens, and dissecting other myrmecophytes such as Cordia , Acacia , Triplaris , Tococa , and Ocotea . The Azteca species which dominate Cecropia trees are found only in Cecropia trees. In spite of high queen density and competition for saplings, I have never encountered one of these Azteca species, either colonies or founding queens, in any plant cavity other than that of a Cecropia . Thus, when only workers are available, biological data on nest site can be of critical diagnostic importance.

Because of the unreliability of worker morphology, many names in Azteca may remain in nomenclatural limbo indefinitely. Identities of species based solely on a type series of workers, with no data on queen morphology or nesting behavior, will only be resolved by a thorough understanding of the subtle differences between workers of all the species at the type locality. In the mean time, it is important to have species descriptions and a nomenclature for this important genus of neotropical ants.

Key to queens (Costa Rica)

The queen of A. chartifex is unknown, but it will probably key to A. tonduzi .

1. Outer margin of hind tibia with abundant erect setae, MTSC usually> 20, if MTSC in the 10-20 range, CI> 95..............................................................................................................................................................2

- Outer margin of hind tibia with erect setae reduced to absent, MTSC <20, if MTSC in the 10-20 range, CI <80.......................................................................................................................................................11

2. Head subrectangular (Fig. 2, coeruleipennis , xanthochroa ), CI <90, obligate Cecropia inhabitant..........3

- Head with sides more rounded (e.g., Fig. 2, constructor ), CI> 90, nesting habits various........................4

3. Color dark brown to black, HW <1.55mm; head shape as in Fig. 2 ( coeruleipennis ); palpal formula 6,4.. ................................................................................................................................................ coeruleipennis

- Color orange, HW> 1.80mm; head shape as in Fig. 2 ( xanthochroa ); palpal formula 5,3...... xanthochroa

4. Color solid black throughout; mesosomal dorsum with dense brush of setae over entire surface; many setae on margin of vertex as long as distance between lateral ocelli (Fig. 2, constructor ); HW 1.4-1.8mm (Fig. 4A); palpal formula 5,3; obligate Cecropia inhabitant....................................................... constructor

- Color rarely solid black, usually red brown, or black with lighter yellow brown on variable extent of anterior and lateral face; mesosomal dorsum often with abundant setae but never a dense brush; setae on margin of vertex shorter than distance between lateral ocelli; HW various; palpal formula 6,4; not obligate Cecropia inhabitants....................................................................................................................................5

5. HW> 2.3mm; face usually uniformly colored light red brown to dark brown...........................................6

- HW <2.2mm; face with variable combination of black to lighter yellow brown (varies from almost entirely black with narrow region of lighter color at anterior margin of head capsule, to almost entirely yellow brown with patch of infuscation on posteromedian vertex)............................................................7

6. Antennae relatively long (SI> 58); ocelli large (OCW> 0.20mm); posterior sternal lobe of petiole shallow and evenly sloping to apex of posterior tergal lobe (Fig. 1E, 5); apex of petiolar node in lateral view flattened and scale-like (Fig. 5) ....................................................................................................... instabilis

- Antennae relatively shorter (SI <55); ocelli small (OCW <0.15mm); posterior sternal lobe of petiole more convex and shorter, meeting posterior tergal lobe before apex (Fig. 1D, 5); apex of petiolar node acute but not as strongly flattened, more bluntly rounded (Fig. 5) ....................................................... gnava

7. HW> 1.9mm (Fig. 4A); face extensively yellow brown with darker brown infuscation on vertex (Fig. 3) ....................................................................................................................................................... sericeasur

- HW <1.9mm; face coloration as above or more extensively dark brown to black .................................... 8

8. HW <1.58mm (Fig. 6A); head almost entirely black, with small band of lighter coloration on anterior head capsule (Fig. 3) ..................................................................................................................... flavigaster

- HW> 1.58mm; head coloration as above or with more extensive yellow coloration extending up sides of head and into antennal fossae ...................................................................................................................... 9

9. Sternal lobe of petiole with posterior margin forming a separate convexity that extends as far posteriorly as posterior tergal lobe, with a small notch or concavity between the sternal convexity and the tergal lobe (Fig. 1D, 5); head relatively short (Fig. 6A); forms ant gardens ........................................................... nigra

- Sternal lobe of petiole evenly curved to tergosternal suture, not forming separate convexity or notch, posterior tergal lobe extending further posteriorly than sternal lobe (Fig. 1E); head relatively longer (Fig. 6A: velox , quadraticeps ); nests in live or dead stems with variable construction of small carton shelters, but not forming large conspicuous ant gardens ............................................................................................... 10

10. Head relatively shorter and more cordate, sides more strongly narrowing anteriorly (Fig. 3, 6A,B) .. velox - Head relatively longer and more quadrate, sides less strongly narrowing anteriorly (Fig. 3, 6A,B) ............ .................................................................................................................................................. quadraticeps

11. Meso- and metatibial spurs absent; anterior margin of clypeus strongly convex, median lobe extending anterad to lateral lobes (HLB/HLA> 1.04); entire body smooth and shiny, highly polished ( aurita group) ................................................................................................................................................................... 12

- Meso- and metatibia with distinct pectinate spurs; anterior margin of clypeus weakly convex, subparallel with lateral lobes (HLB/HLA <1.01); body surface duller, not shiny ....................................................... 14

12. Color yellow-orange; posterolateral margins of vertex forming pronounced triangular lobes (Fig. 2); CI> 85 (Fig. 4A) ...................................................................................................................................... pilosula

- Color brown; posterolateral margins of vertex not forming pronounced triangular lobes; CI <80 .......... 13

13. HW <0.65mm (Fig. 4A); sides of head relatively more rounded and converging toward posterolateral lobes of vertex (Fig. 9); outer surface of metatibia with abundant suberect pubescence and numerous, regularly spaced, short, erect setae ...................................................................................................... nanogyna

- HW> 1.00mm (Fig. 4A); sides of head subparallel, not converging toward posterolateral lobes of vertex (Fig. 2); outer surface of metatibia with pubescence relatively short, sparse, appressed, and inconspicuous; outer surface of metatibia lacking erect setae or with a few, irregularly spaced, short setae near the base ................................................................................................................................................. schimperi

14. Head relatively short and broad (CI> 78) ................................................................................................. 15

- Head relatively long and narrow, subrectangular (CI <78) ...................................................................... 18

15. Petiolar node strongly flattened, scale-like (Fig. 5); head short and broad (CI> 100, Fig. 4A) ........ tonduzi

- Petiolar node not strongly flattened, apex usually bluntly rounded; head relatively longer and thinner (CI <90) ........................................................................................................................................................... 16

16. HW> 1.53mm (Fig. 4A); dorsal surface of mandible evenly covered with abundant large piligerous puncta (Fig. 1G), setae arising from puncta erect, about as long as width of mandibular teeth; builds carton galleries on tree trunks, not a Cecropia specialist .......................................................................... forelii

- HW <1.45; much of dorsal surface of mandible with sparse, small puncta bearing setae that are reduced to short remnants, no longer than width of puncta (Fig. 1F), puncta with longer setae restricted to apex and near masticatory margin; Cecropia specialists..........................................................................................17

17. Fourth abdominal tergum with <6 erect setae, exclusive of posterior row; dorsal surface of head, when viewed in profile, with setae occurring in three clusters separated by distinct gaps, one cluster on and just above the clypeus, one around the ocelli (these may be entirely absent), and one on the upper vertex; scape relatively short (SI 45-49, Fig. 6C); color usually black............................................................ alfari

- Fourth abdominal tergum with> 10 erect setae (rarely fewer), exclusive of posterior row; dorsal surface of head, when viewed in profile, often with setae bridging the gap between the ocellar region and the upper vertex, and often with setae extending up from the clypeus almost to the ocellar region; scape relatively longer (SI 49-54, Fig. 6C); color black to lighter red brown............................................... ovaticeps

18. Color largely orange; HW> 1.2mm (Fig. 4A)...................................................................................... beltii

- Color largely or entirely black; HW <1.2mm...........................................................................................19

19. Mandible with even covering of coarse, piligerous puncta; mandible surface appearing bristly..............20

- Mandible with row of piligerous puncta along masticatory margin, but large puncta sparse to absent on mandible surface proximal to this row, and with at most four puncta bearing setae.................................21

20. Scape relatively short (SI 39-43).......................................................................................................... brevis

- Scape longer (SI 50-52)................................................................................................................. nigricans

21. Head strongly rectangular, with flat sides and posterolateral lobes of vertex relatively angular; head relatively longer and narrower (SI <63, Fig. 4B)...........................................................................................22

- Head less rectangular, with sides slightly convex, and lateral margin of vertex more broadly rounded; head relatively shorter and wider (SI> 63, Fig. 4B).......................................................................... pittieri

22. Petiolar node low, anterior face of petiole flat (Fig. 5); metatibia with few erect setae (MTSC <5); propodeum with sparse short setae concentrated posterior to spiracle; mandible lacking large puncta proximal to masticatory margin; size relatively large (HW> 0.92mm, Fig. 4B)........................................... oecocordia

- Petiolar node higher, anterior face somewhat concave (Fig. 5); metatibia with relatively more setae (MTSC 10-20); propodeum with setae sparse or abundant; mandible with about 5 large puncta proximal to masticatory margin; size relatively smaller (HW <0.93mm, Fig. 4B)...................................... longiceps

Key to workers (Costa Rica)

Identifying Azteca species from the morphological traits of individual workers is difficult. Colonies show strong size variation among workers (Wheeler 1986), and the size of the largest workers increases as colonies mature. The larger a worker is, the more queen-like it is and the more differentiable from other species. The following key is most likely to work when a series from a mature colony is available, so that the largest workers can be selected for examination. When size ranges are given in the key, they refer to these larger workers. The workers of A. nanogyna and A. quadraticeps are unknown.

1. Middle and hind tibia lacking apical spur; anterior margin of clypeus strongly convex, median lobe protruding further than lateral lobes (HLB/HLA> 1.04), Fig. 3: pilosula and schimperi ); palpal formula 4,3 ( aurita group)...............................................................................................................................................2

- Middle and hind tibia with distinct, pectinate apical spur; median lobe of clypeus not protruding, subparallel with lateral clypeal lobes (HLB/HLA <1.01); palpal formula 5,3 or 6,4...........................................3

2. Dorsal surface of mandible smooth, not striate; posterolateral margin of vertex evenly rounded (Fig. 3); sternal lobe of petiole with sharp, longitudinal carina.................................................................... schimperi

- Dorsal surface of mandible striate; posterolateral margin of vertex subangular (Fig. 3); sternal lobe of petiole longitudinally tectiform but not carinate................................................................................... pilosula

3. Outer surface of hind tibia completely devoid of erect setae, rarely with 1 or 2 short setae on largest workers; palpal formula 5,3 ................................................................................................................................. 4

- Outer surface of hind tibia with 5 or more erect setae (these may be very short, less than a quarter of tibial width, and difficult to see); palpal formula 5,3 or 6,4.. ............................................................................... 8

4. Promesonotum strongly produced, bulging, dropping steeply to flat dorsal face of propodeum (Fig. 7) ..... ......................................................................................................................................................... chartifex

- Promesonotum less strongly produced, posterior mesonotum more shallowly sloping and meeting dorsal face of propodeum at more obtuse angle ..................................................................................................... 5

5. Dorsal surface of mandible largely smooth with small non-setose puncta, setigerous puncta restricted to masticatory margin (Fig. 1F) ....................................................................................................................... 6

- Dorsal surface of mandible evenly covered with large setigerous puncta (Fig. 1G) ................................... 7

6. Dorsum of mesosoma with a relatively "clean" look, with relatively few erect setae, these of relatively uniform length (MNSC 2-17, median 8, Fig. 7) .................................................................................. alfari

- Dorsum of mesosoma more "scruffy," with more setae, and these of irregular length (MNSC> 10, median about 20, Fig. 7) .............................................................................................................................. ovaticeps

7. Dorsal face of propodeum with moderately abundant erect setae; SI> 50; HW of larger workers often> 1.1mm; basal half or more of mandible microalveolate, dull ............................................................... forelii

- Dorsal face of propodeum devoid of erect setae; SI <50; HW of larger workers usually <1.0mm; entire mandible generally smooth and shiny on interspaces between puncta ................................................. brevis

8. Palpal formula 6,4; color yellow; dorsal face of propodeum and metanotal groove together form a single flat "shelf" that abruptly meets rising posterior mesonotum (Fig. 7); obligate Cecropia inhabitant ............. ................................................................................................................................................ coeruleipennis

- Palpal formula 5,3 or 6,4; color various; dorsal face of propodeum and mesonotum forming two convexities that meet at impressed metanotal groove (e.g., Fig. 7, sericeasur worker); nesting habits various ..... 9

9. Palpal formula 5,3; mostly species (with the exception of A. tonduzi ) nesting in myrmecophytes ( Cecropia , Cordia , Triplaris ) or narrow-gauge live stems of various plant species, with foraging restricted to the hostplant stem interiors or surfaces ........................................................................................................... 10

- Palpal formula 6,4; species nesting in large plant cavities or ant gardens, with conspicuous and generalized surface foragers .................................................................................................................................. 17

10. Head relatively broad, CI> 86 (Fig. 6D); obligate Cecropia inhabitants or nesting in dead stems .......... 11

- Head relatively narrow, CI <90 (Fig. 6D); inhabitants of live stems but not obligate Cecropia specialists ................................................................................................................................................................... 13

11. Setae on metatibia sparse and short, MTSC 10-20, length of setae one fourth to one third maximum width of tibia; HW of larger workers <0.91mm (Fig. 6D); nests in dead stems often augmented with carton nests .................................................................................................................................................... tonduzi

- Setae on metatibia abundant and longer, MTSC> 20, length of setae one half or more maximum width of tibia; HW of larger workers> 1.10mm (Fig. 6D); obligate inhabitants of Cecropia ................................ 12

12. Petiole in profile with node more massive than sternal lobe, perpendicular distance from tergosternal suture to apex of node greater than or equal to distance to ventral margin of sternal lobe (Fig. 7); face with mottled coloration, light brown to orange with variable extent of medial infuscation (Fig. 3).... xanthochroa

- Petiole in profile with node less massive than sternal lobe, perpendicular distance from tergosternal suture to apex of node less than distance to ventral margin of sternal lobe (Fig. 7); face uniformly brown (Fig. 3) ..................................................................................................................................................... constructor

13. Most of dorsal surface of mandible covered with large piligerous puncta, mandibles appearing bristly (Fig. 1G); nesting in live stems of multiple plant species, not obligate inhabitant of Cordia or Triplaris ... ........................................................................................................................................................ nigricans

- Dorsal surface of mandible with large piligerous puncta, if present, restricted to masticatory margin and apex, mandibles not appearing bristly (Fig. 1F); nesting habits various, including obligate inhabitants of Cordia and Triplaris ..................................................................................................................................14

14. Setae on metatibia few and short, MTSC usually about 5, setae about as long as one quarter width of tibia; propodeum with cluster of about 4 setae where dorsal face rounds into posterior face, 0-4 setae anterior of this cluster on dorsal face, these setae short, 1-2 times width of propodeal spiracle; largest workers with HW> 1.0mm (Fig. 6D).............................................................................................................................15

- Setae on metatibia more abundant and longer, MTSC 8-17, setae about as long as half width of tibia; dorsal face of propodeum with 4 or more setae, these setae long,> 4 times width of propodeal spiracle; largest workers with HW <1.0mm (Fig. 6D)..................................................................................................16

15. Face of largest workers yellow; dorsal surface of mandible with interspaces between puncta largely smooth and shining, at most basal third of mandible shagreened; posterior margin of sternal lobe of petiole with a layer of dense, short, pubescence-like pilosity and 2-4 longer setae; opportunistic inhabitant of multiple ant plants and live stems of non-myrmecophytes................................................................... beltii

- Face of largest workers brown; dorsal surface of mandible with interspaces between puncta shagreened on basal half or more; posterior margin of sternal lobe of petiole with cluster of erect setae of even length; obligate inhabitant of Cordia alliodora ....................................................................................... oecocordia

16. Head relatively broader (CI 78-91, Fig. 6D); obligate inhabitant of Cordia alliodora (or understory Lauraceae, see under A. pittieri )............................................................................................................... pittieri

- Head relatively narrower (CI 73-81, Fig. 6D); obligate inhabitant of Triplaris melaenodendron ............... ........................................................................................................................................................ longiceps

17. Mandibles opaque, densely microalveolate/punctate, substriate; posteroventral margin of petiole with abundant long coarse setae, shorter pubescence layer much less conspicuous than long setae (Fig. 7); majors very large (HW up to 2.2mm); nests in hollow trunks of large, live trees, with nest entrance a fissure at the base of the tree ................................................................................................................ instabilis

- Mandibles shiny to weakly sculptured; largest workers typically smaller, HW <1.8mm; nesting habits various (the workers of the following species are highly variable and I have not been able to discover diagnostic characters that always differentiate them; the following key couplets reflect average differences among species, but I have examined many worker series in this complex that I could not place).... 18

18. Gastral dorsum bright yellow, sharply contrasting with dark brown petiole and mesosoma; measurements as in Figs. 6E,F.............................................................................................................................. flavigaster

- Gastral dorsum dark brown, or, if light yellow brown, petiole and mesosoma are lighter brown as well, gastral color not sharply contrasting with mesosoma color.......................................................................19

19. Head relatively broad (Fig. 3, 6E, CI 101-109); scape relatively short (Fig. 6F, SI 70-85); face usually uniformly dark brown (lighter orange brown in one series); setae on posterior margin of vertex usually long, subequal in length to eye length; forms ant gardens................................................................... gnava

- Head relatively narrower; scape relatively longer; face color various but often a mixture of light and dark brown; setae on vertex margin usually shorter than eye length; nesting habits various............................20

20. Largest workers with HW usually> 1.2mm; general body coloration often light orange brown; head shape often strongly cordate (Fig. 3)....................................................................................................... sericeasur

- Largest workers with HW usually <1.2mm; general coloration often dark brown with light orange brown restricted to anterior and lateral portions of head; head shape often less strongly cordate (Fig. 3: velox and nigra ).........................................................................................................................................................21

21. Scape relatively short (Fig. 6F, SI 78-89); nests in plant cavities......................................................... velox

- Scape relatively longer (Fig. 6F, SI 84-105); nests in ant gardens....................................................... nigra

Alto Paraná , Amambay, Canindeyú , Central, Ñeembucú , Pte. Hayes (ALWC, IFML, INBP, LACM, MZSP, NHMB).

Azteca , Forel, Bull. Soc. Vaud. Sci. Nat. (2) xv. p. 384 (1878); Dalla Torre, Cat. Hymen. vii. p. 163.

L'examen d'un grand nombre de [[ male ]] et de [[ queen ]] d' Azteca me permet de declarer constant et definitif le type de leur aile que j'ai indique sur celui d'un hermaphrodite de l'A. muelleri . Elles n'ont qu'une cellule cubitale. La nervure transverse s'unit a la nervure cubitale a son point de partage. La cellule radiale est fermee. Cela distingue ce genre des Liometopum et de presque tous les Dolichoderinae .

Il est fort singulier que jusqu'a ces dernieres annees, specialement jusqu'a la belle monographie d' Emery [Mem. Accad. Sci. Bologna, (5) iii. pp. 119 - 152 (1893), Monogr. Gen. Azteca , Forel], ce genre si considerable et si abondant dans les forets tropicales de l'Amerique n'ait ete connu que par deux ou trois especes, dont on ignorait meme les [[ male ]]. Fritz Mueller a rendu ces fourmis celebres par sa decouverte de la symbiose de l' Azteca muelleri, Emery , avec le Cecropia peltata a Blumenau, arbre qui donne a

l' Azteca sa nourriture, tandis que celle-ci le defend contre la destruction de ses feuilles par les Atta .

En Colombie j'ai eu l'occasion d'observer beaucoup d' Azteca , car la foret en est remplie. Leurs moeurs - etudiees malheureusement seulement a la course, pendant un court voyage - offrent beaucoup de points communs:

Les Azteca sont (sauf une seule espece parmi celles que j'ai observees) toutes tres guerrieres, vivent exclusivement sur les arbres, ou peu s'en faut, marchent en relevant l'abdomen qu'elles font pirouetter en tout sens, comme les Tapinoma , ce qui correspond a la forme de cet organe et du pedicule. Sans exception, elles repandent toutes une forte odeur de Tapinoma des qu'on les inquiete, et c'est cette secretion de leurs glandes anales qui les rend si redoutables. Je les ai vu mettre ainsi en desarroi et en fuite une armee d' Eciton hamatum bien plus gros qu'elles et arme d'un aiguillon.

Azteca font leur nid soit dans des arbres creux, soit en carton, sur les troncs ou les branches d'arbres. Leurs nids de carton sont tres elegants, suspendus autour des derniers rameaux, parmi les feuilles (voir Tab. II. fig. 3) ( Azteca chartifex ), ou a des branches plus fortes en forme de cone renverse ou de stalactite, ou encore adosses a un tronc ou a une branche en forme d'outre, avec le bas renfle et le haut plus ou moins raminci. Ces nids, surtout les grands, adosses aux troncs ou a de grosses branches, ont une surface curieusement sculptee, c'est a dire que le carton y forme des bas reliefs en forme de larmes geantes et aplaties. Les nids sont en general elargis et arrondis en bas, attenues en haut. On distingue par cet aspect les nids d' Azteca des nids de termites dans la foret. Souvent ils sont a une grande hauteur qui empeche de les atteindre. Pour m'en procurer un pareil, j'ai du faire abattre au machete un grand Cecropia, par notre guide. Quelques-uns etaient assez bas pour etre atteints. Le carton de ces nids est delicat et friable, varie du reste selon les especes. Tandis que l' A. chartifex, r. multinida , fait des nids gros comme le poing, d'autres especes en forment de fort grands. J'en ai mesure un ( Azteca lacrymosa ) de 70 centimetres de hauteur sur 40 cm. de largeur et 20 cm. d'epaisseur, adosse au tronc d'un arbre.

Les Azteca forment souvent des colonies sur le meme arbre, c'est a dire que la meme fourmiliere y construit plusieurs nids qui demeurent en relations constantes d'amitie les uns avec les autres. Des observateurs superficiels ont pris ces nids pour des nids de termites envahis par des fourmis. Il n'est cependant pas difficile de les distinguer de ceux des termites arboricoles qui sont beaucoup plus durs et ont une autre architecture interne et externe. L'enveloppe lacrymiforme des nids d' Azteca cartonnieres est formee d'une mince couche de carton qui menage de nombreuses ouvertures cachees comme des meurtrieres sous les " larmes " citees plus haut, de sorte qu'on ne les voit qu'en regardant obliquement. Les cases et galeries periferiques du nid sont plus ou moins aplaties et cette partie est tres fragile. Au centre par contre, les cases sont plus arrondies et le carton est plus consistant.

Un grand nombre d'autres Azteca vivent dans les arbres creux. Il m'a ete en somme impossible, faute de temps et d'instruments, de constater si ces dernieres font oui ou non un nid en carton dans l'interieur de l'arbre, comme le Lasius fuliginosus et le Liometopum microcephalum d'Europe. Cela ne parait pas etre le cas de Y Azteca muelleri , observee par Fritz Mueller, mais l' Azteca constructor , Emery, construit un nid de carton dans les cavites de son Cecropia. Je considere la chose comme probable pour les especes qui n'offrent pas de symbiose ou d'adaptation speciale. Dans un arbre creux a S. Antonio, j'ai pu sortir un peu de carton du nid d'une Azteca . C'est la seule fois que cela m'a ete possible. J'ai enfin observe deux especes d' Azteca dont l'une habite sur les rochers de la foret dans des galeries de carton qui serpentent sur ces rochers comme celles du Cremastogaster stollii sur les arbres, tandis que l'autre ( A. hypophylla , n. sp. ) vit sous les feuilles d'une plante grimpante. Les feuilles de cette plante s'appliquent tres exactement, plus encore que celles du lierre, contre l' ecorce de l'arbre. L' Azteca se borne a coller entierement le bord de la feuille a l' ecorce avec son carton et vit avec sa famille sous les feuilles ainsi transformees en loges tres aplaties. Le corps de cette Azteca est aussi tres aplati. D'autres especes sont adaptees a certaines plantes d'une facon ou de l'autre.

cette derniere espece, toutes les Azteca que j'ai vues, tant les cartonnieres que celles qui vivent dans le creux des arbres, sortent en masse et avec fureur, des qu'on approche de leur arbre. Lorsqu'on frappe l'arbre avec force, c'est une legion de ces petits defenseurs qui sort de partout et se jette avec rage sur l'agresseur. Il n'est pas commode de prendre un nid d' Azteca en carton. En un instant on est couvert de fourmis de la tete aux pieds. N'ayant pas d'aiguillon, elles ne peuvent cependant que mordre, chatouiller et repandre la secretion odorante de leurs glandes anales.

Autant que j'ai pu en juger, les Azteca paraissent etre surtout carnivores et vivre de rapines. Elles considerent leur arbre comme leur domaine qu'elles se disputent avec les Pseudomyrma . Mais il est probable qu'il y en a d'autres qui se nourrissent plutot de substances vegetales, comme l' A. muelleri . Ce sont surtout les Azteca qui sont les reines des arbres de la foret vierge americaine, et je commence a croire que les epines et les poils laineux des Atta leur servent surtout a se proteger contre les Azteca . Quant on ne visite que les ports et les lieux cultives, on ne se doute pas de la presence des Azteca dans la faune americaine. Il faut entrer dans la veritable foret vierge ou au moins sur ses confins pour les trouver et observer leur role.

La symbiose de Y Azteca muelleri avec le Cecropia peltata est un cas special qu'il faut se garder de generaliser pour le genre Azteca . Wasmann a fait observer que le mimetisme ne se developpe que lorsque la forme du corps, la couleur ou les m oe urs d'une espece lui avaient deja prepare le terrain, c'est a dire lorsque des analogies fortuites preexistantes avaient prepare un avantage a ceux des individus dont l'analogie s'accentue tout specialement. Eh bien! il me parait en etre de meme avec la symbiose. - Natura non fecit saltum. Nous avons vu l'instinct du jardinage des champignons se developper peu a peu chez les Attii. De meme je crois que la symbiose de l' Azteca muelleri s'est developpee en partant de la base biologique generale des Azteca , dont le domaine sont les arbres, qui vivent dans leurs cavites et qui defendent ce domaine avec acharnement contre les ravages des Atta phyllotomes. De ce fait general a l'adaptation speciale de telle ou telle espece d' Azteca a un arbre special et surtout a un Cecropia, dont le tronc est creux, il n'y a qu'un pas bien facile a faire du cote de la fourmi. Ce qui est plus difficile a expliquer est la part active que prend d'apres Mueller le Cecropia peltata a la nutrition de l' Azteca muelleri par la production des corpuscules de Mueller.

Au point de vue biologique, on peut donc en somme diviser les Azteca en trois groupes assez naturels: -

°. Truncicoles (Truncicola). - Ce sont les especes qui habitent indifferemment l'interieur des troncs d'un arbre quelconque plus ou moins excave, mort ou pourri, en y construisant ou non du carton: instabilis , velox , delpini , etc., en general poilues.

2 °. Cartonnieres (Chartifices). - Celles qui font des nids de carton a l'air libre, a la surface des troncs ou des branches d'arbres (peut-etre aussi sur les rochers): chartifex , festai , lacrymosa , lallemandi .

3 °. Adaptees (Adaptatae). - Celles qui sont adaptees (avec symbiose ou non) a un arbre (Cecropia) ou une plante speciale: muelleri , constructor , alfaroi , coeruleipennis , angusticeps , depilis , schumanni (adaptee aux vesicules des feuilles d'un Chrysobalanea), virens , tonduzi , hypophylla , xanthochroa .

Azteca is a strictly Neotropical genus of ants in the subfamily Dolichoderinae. The genus is very diverse and contains around 84 extant species and two fossil species. They are essentially arboreal and many species have mutualistic associations with particular plant species, where the genus Cecropia presents the most conspicuous association. In the Brazilian Amazonia, Azteca species are associated with species of Codonanthopsis.