Ant community structure in the brazilian tropical dry forest (Cerrado).

Author(s)

Rogerio Silvestre

• Document type: Doctoral Thesis
• Press: Ribeirão Preto.
• Institution: Universidade de São Paulo (USP). Faculdade de Filosofia, Ciências e Letras de Ribeirão Preto (PCARP/BC)
• Defense date: 2000-06-09

Examining board members

Carlos Roberto Ferreira Brandao; Jacques Hubert Charles Delabie; Carlos Alberto Garofalo; Klaus Hartmann Hartfelder; Jader Soares Marinho Filho

• Advisor: Carlos Roberto Ferreira Brandao

Abstract

The present work investigates three aspects of ant community structure in the tropical dry forest of the Brazilian Central Plateau (Cerrado): interespecific interactions, richness and diversity of species and guilds. Firstly, I investigated the influence of aggressive behaviors in interspecific interactions on ant faunistic surveys that employ baits, assessing if the eventual behavioral dominance hierarchy on the food source influences the relative frequencies observed in surveys. I used as model sardine baits that mimic sources of protein, recording the behavioral acts of ants visiting 60 baits, offered for 90 minutes, on the soil of two Cerrado localities. Seventy-two ant species were registered performing 682 interactive behavioral acts, of which 352 were aggressive, resulting in the death of 29 individuals. Of the observed species, most constantly monitor the habitat searching for food, resulting in niche overlapping, and hence interespecific competition was quite frequent. Most baits (85%) were visited in the first five minutes after exposition, and in average 4,8 (1-8) species visited each bait during the 90 min. The most common behavior registered was the opportunistic visit of ants to baits, in which individuals find the bait alone, take some of it and leave the spot. Nevertheless some species were observed obstructing the access of others visiting the same bait, either by releasing irritating sprays, employing massive recruitment, or by agonistic behaviors. My observations indicate that although dominance may occur resulting in the exclusion of subordinate species, the order in dominance hierarchies may change, possibly as a result of the different distances from the nest to the food source, of the species composition interacting at a given bait, and of the different recruiting strategies employed by them in each situation. Secondly, I check the ant fauna through qualitative and quantitative surveys in seven Cerrado localities. For this purpose the samples include all species collected by different techniques. To perform the quantitative surveys, one area (1ha) of Cerrado sensu stricto was chosen before collecting events in each locality, and 25 points were randomly chosen among 121 points in a grid made by 11 transects, separated 10m either, with 11 points, also separated 10m from each other. At each locality sardine baits were offered for 90 minutes, over the soil and on the vegetation, at day and night periods. A total of 4400 baits in all seven localities were offered in order to compare the ant species composition in each surveyed area. I summed to this total 500 baits set in Forests bordering rivers, on transects covering 250m, in five localities. For the qualitative surveys, a set of general procedures was applied including general collecting using forceps and aspirator, excavations on termites nests, the turn of rocks and opening of rotting wood, tearing up logs and stumps, submitting litter samples to Berlese-Tüllgren funnels and Winkler extractors, the setting of pit-fall traps and yellow water trays, and Malaises and light-traps to capture winged forms. Summing all ants collected in qualitative and quantitative surveys I registered 331 species. In this survey I registered the subfamilies: Formicinae, Myrmicinae, Dolichoderinae, Ecitoninae, Ponerinae, Pseudomyrmecinae, and Cerapachyinae. The genera Camponotus, Pheidole, Crematogaster, and Solenopsis were the most rich in number of species. Pseudomyrmex showed a great diversity (29 species), including undescribed ones. Several species had their first record in this latitude as Blepharidatta conops, Gigantiops destructor, Paraponera clavata, Megalomyrmex acauna, and Tingimyrmex mirabilis. Finally, I analyzed the Cerrado ant community trying to classify the species in guilds based on taxonomy, throphic preference, nidification habits, foraging and recruitment strategies, behavioral patterns observed on interespecific interactions, size of workers, “readiness of movement" and estimated mature colony population. I selected for this analysis those taxa associated with field observation data, considering for this purposes a total of 110 species. A cluster analysis, with Euclidean distance, and with complete linkage oriented the formation of 12 guilds: big predators, patrollers, opportunistic, cryptic, leaf cuttings, fungus growers over carcass, generalist myrmicines, aggressive arboreal, nomads, minimal specialist, cephalotines, and nectars feed dolichoderines. After that I compared two Cerrado localities, utilizing the guild model, revealing as well which species substitute others in different localities, and that the communities are functionally similar. (AU)