Showing papers by "Edward O. Wilson published in 1976"

Behavioral discretization and the number of castes in an ant species

Abstract: 1. Two extreme possibilities in the evolution of temporal castes can be envisaged. First, workers can undergo changes in responsiveness to various kinds of stimuli in a strongly discordant manner as they grow older, so that each task is addressed by a distinctly different frequency distribution of workers belonging to different age groups. Because these age-frequency distributions change almost gradually from one task to another in covering many such tasks, the resulting temporal caste system is referred to as continuous. At the opposite extreme, the aging worker can undergo changes in responsiveness to different stimuli in a highly concordant manner, so that all of the tasks are attended by one or relatively few frequency distributions of workers belonging to different age groups. The resulting temporal caste system is referred to as discrete, and the evolutionary process leading to it is called behavioral discretization (Fig. 1). 2. The temporal system of the minor worker caste of Pheidole dentata proves to be much closer to the discrete state, although it is not extreme in form (Figs. 3, 4). On the basis of ethograms constructed of stressed and unstressed colonies in which the approximate ages of the minor workers were known, it is possible to recognize five discrete female castes: the queen, a single temporal subcaste of the major worker, and three temporal subcastes of the minor worker. These are the elements which can now be employed in ergonomic analyses of the species' caste system.

The organization of colony defense in the ant Pheidole dentata mayr (Hymenoptera: Formicidae)

Abstract: 1. Colonies of Pheidole dentata employ a complex strategy of colony defense against invading fire ants. Their responses can be conveniently divided into the following three phases: (1) at low stimulation, the minor workers recruit nestmates over considerable distances, after which the recruited major workers (“soldiers”) take over the main role of destroying the intruders; (2) when the fire ants invade in larger numbers, fewer trails are laid, and the Pheidole fight closer to the nest along a shorter perimeter; (3) when the invasion becomes still more intense, the Pheidole abscond with their brood and scatter outward in all directions (Figs. 1, 4). 2. Recruitment is achieved by a trail pheromone emitted from the poison gland of the sting. Majors can distinguish trail-laying minors that have just contacted fire ants, apparently by transfer of the body odor, and they respond by following the trails with more looping, aggressive runs than is the case in recruitment to sugar water. Majors are superior in fighting to the minors and remain on the battleground longer. 3. The first phase of defense, involving alarm-recruitment, is evoked most strongly by fire ants and other members of the genus Solenopsis; the presence of a single fire ant worker is often sufficient to produce a massive, prolonged response (Figs. 2, 5, 6). In tests with Solenopsis geminata, it was found that the Pheidole react both to the odor of the body surface and to the venom, provided either of these chemical cues are combined with movement. Fire ants, especially S. geminata, are among the major natural enemies of the Pheidole, and it is of advantage for the Pheidole colonies to strike hard and decisively when the first fire ant scouts are detected. Other ants of a wide array of species tested were mostly neutral or required a large number of workers to induce the response. The alarm-recruitment response is not used when foragers are disturbed by human hands or inanimate objects. When such intrusion results in a direct mechanical disturbance of the nest, simulating the attack of a vertebrate, both minor and major workers swarm out and attack without intervening recruitment.

The First Workerless Parasite in the Ant Genus Formica (Hymenoptera: Formicidae)

Abstract: Formica talbot.ae Wilson, a member of the microgyna group and the first proven workerless parasite in Formica, is described here. The species is known from Michigan, Iowa, and North Dakota. The discovery of talbotae com­ pletes the inferred evolutionary progression within Formica from independent existence through temporary parasitism to permanent, workerless parasitism. During her exhaustive survey of the ant fauna of the Edwin S. George Reserve of Michigan, an effort previously unparalleled in North America, Mary Talbot has uncovered a surprising num­ ber of rare and undescribed species. One of the most significant is the species to be described below, a member of the Formica microgyna group which to the best of my knowledge is the first adequately documented example of a workerless parasite in this large Holarctic genus. Formica talbotae Wilson, new species Diagnosis (queen). A small species even for the microgyna group, characterized further by the following combination of traits: subquadrate head; smoothly rounded anterior clypeal border; thick petiolar node with relatively thick, rounded crest; short (0.05-0.08 mm), dense standing pilosity over all of body and appendages, including scape; many of the hairs on the tho­ racic dorsum, propodeum, petiole, and fore coxae spatulate. So far as known, talbotae is exclusively a workerless parasite of Formica obscuripes Fore!. Relationships. During the study I examined specimens of all of the microgyna group species for which sexual forms are known. Most of those known from workers solely were also examined, but are in any case considered probably distinct on the basis of the possession of a worker caste alone. The closest species is F. dirksi Wing, which differs in the queen caste by its slightly lar­ ger size; much longer, less frequently spatulate pilosity; and more rounded head shape. F. spatulata Buren is also close but its queen

Sergei Kovalev: A Colleague in Trouble

Abstract: predict the approximate, potency of a carcinogen with some degree of probability, this would be extremely useful. One would like to be able to do this in drug development and in evaluation of the hazard of complex mixtures (such as water effluents, air pollution samples, and so forth), in which animal cancer tests are impractical. It is clear that by using a simplified system such as a rat liver homogenate and bacteria, one would not expect to be able to precisely predict carcinogenic potency in a rat (or a human). If one could predict it with a high probability within + an order of magnitude this would be extremely useful, considering the range of carcinogenic potency. We believe the test may well be able to do this. Sivak chooses his carcinogenic potency examples from a much too narrow range where one could not see any correlation that existed. Russell and Meselson' (4) at Harvard are actively pursuing the area of the degree of quantitative correlation between a chemical's carcinogenic potency in animals and mutagenic potency in the Salmonella test, and following their lead we are doing the same. There are some animal carcinogenicity data from feeding experiments of appropriate quality for calculating carcinogenic potency and also some data on humans that meet the requirements. 2) Sivak says that we selected our strains to detect carcinogens and therefore the fact that they detect carcinogens is \"self-fulfilling and not a true test.\" We selected our strains primarily on the basis of maximizing the detection of known mutagens (we did not think about carcinogenicity until much later), and fewer than 10 percent of the 175 carcinogens we actually tested in the validation of the method were used in the development of the strains. [In addition, the test has been independently validated (90 percent correlation) in a blind study of 120 chemicals (5).] Very few chemicals in general are mutagens or carcinogens, and the finding that more than 90 percent of carcinogens tested have been detected as mutagens (and that almost every mutagen that has been given an adequate cancer test is a carcinogen) may actually mean something. The chemicals known to be carcinogenic in humans represent an unselected sample, and the test detects almost all of them as mutagens (2). 3) Sivak questions the \"equivalency\" (we would not use that word) of mutation in bacterial DNA with \"the multistep, multifactorial process of carcinogenesis in eukaryotic organisms.\" We have briefly discussed the idea of DNA damage (somatic mutation) as the initiator of most chemical and radiation carcinogen-