Resumo em Português:
Resumo Analisam-se padrões geográficos de evolução e diferenciação (por pressão de mimetismo Mülleriano) em 162 espécies de Heliconiini e Ithomiinae na região neotropical (com revisão sistemática e biológica, e descrição de novos taxas), representando 905 entidades diferenciadas, um milhão de dados distribucionais e 2.832 localidades. Isto indica 50 principais centros de evolução e endemismo de floresta tropical (fenômenos biológicos empíricos e dedutivos). São empregados critérios climatológicos, topográficos, pedológicos e botânicos na proposição de 38 correspondentes refúgios florestais (fenômenos históricos indutivos), que atuaram na preservação e diferenciação de populações silvícolas durante o último período longo de clima frio e seco (20.000 — 13.000 anos a.p.). São analisadas as variações nas relações entre biotas em diferentes refúgios, no modo de ação dos refúgios sobre diferentes organismos, nas respostas das populações desses organismos, na conservação atual dos padrões de diferenciação em espécies diferentes, e na integridade passada e presente das áreas refugiais. Conclui-se que (1) a determinação de centros generalizados de evolução de biotas é relativamente fácil, desde que haja número suficiente de dados geográficos e uma sistemática adequada para identificação de linhas monofiléticas nos organismos vicariantes analisados; e (2) a determinação de refúgios florestais é muito mais complicada. Todas as informações disponíveis levam a crer que estes refúgios ocorreram em regiões diferentes durante diversos ciclos climáticos do passado, e que foram fenômenos quantitativos e não qualitativos, onde as condições ecológicas determinaram uma probabilidade maior de preservação e diferenciação do que de extinção da maioria das populações silvícolas. É discutida a estratégia mais indicada para melhor preservação da diversidade genética neotropical, que deverá combinar elementos teóricos e pragmáticos. Dos 38 refúgios propostos, apenas 15 já incluem parques nacionais ou reservas biológicas efetivas, enquanto 22 são muito modificados pela ação do homem, e um (Guaporé) poderá ser totalmente destruído antes de serem minimamente conhecida sua flora e fauna endêmicas.
Resumo em Inglês:
Summary By examination of 162 species of Neotropical Heliconiini and Ithomiinae (Lepidoptera: Nymphalidae), representing 905 differentiated entities, an analysis was developed of their geographical patterns of evolution and differentiation, under selective pressure of Müllerian mimicry. The results of this analysis were used to expand and revise concepts derived from recent work on a variety of Neotropical animals and plants, which have indicated that entire biotas evolved in selected localities of tropical forest which, though contiguous today, were isolated as refuges during past dry climatic cycles. Approximately one million distributional data from 2,832 different localities in ail regions of Central and South America were used to prepare maps of polytypic species and species-groups, which indicated the existence, in these organisms, of 50 principal forest centers of evolution and endemism (which are defined as empirical and deductive, present-day biological phenomena). Climatological, palynological, topographical, pedological (especially useful) and botanical criteria, as well as considerations from theoretical ecology and population genetics (especially characteristics expected from r and K selection on the populations and the reflections of these on the community level) were applied to these centers of evolution. This led to the proposal of corresponding but smaller areas, recognizable as forest refuges (inductive biogeographical phenomena with a climatological, geological, and historical basis). These presumably acted during the last major dry, cold spell in the Quaternary, between 20,000 and 13.000 years b.p. There were discovered 38 such refuges (plus at least 20 additional subrefuges), as well as a medium-elevation area on the eastern slopes of the Andes, which apparently acted in the preservation and differentiation of forest populations during this period of unfavorable climate. Five present-day islands and various remnant forests could also be identified; they showed biotas less differentiated than those of the proposed refuges, suggesting that their evolution is more recent. The remnant forests show the disruption of links between the Atlantic forest island in Pernambuco and the Belém refuge area (probably after the climatic optimum of 6,000-7,000 years b.p.) and between the Bahia, Araguaia and Guaporé/Rondônia refuges (probably in the more distant past). The variations observed in the presumed number, position, and period of action of the refuges, in the relations among species and biotas in the different refuge areas, in the mode of action of the refuges on differentiating organisms with different population characteristics, in the responses of the forest populations of these organisms, in the presently observable patterns of differentiation conserved in species with different competitive and dispersal characteristics, and in the past and present integrity of the refugiai areas, suggest the following conclusions: (1) The determination of centers of evolution for vicariant organisms is relatively easy, provided that there exist sufficient geographical data and an adequate systematic organization to permit identification of monophyletic lines. To the extent that the number of organisms analyzed increases, the number of recognizable centers of evolution also increases, reaching a stable limit only when the areas begin to overlap, and appear as subunits of a single generalizable center. This point was reached in the present work, as the 50 centers are observed in a number of subunits of the two taxonomic groups analyzed. Ecological parameters vary appreciable among the species, as do the genetic architectures of the populations; this results in very diverse present-day differentiation patterns, all however understandable in terms of the centers of evolution proposed. (2) The determination of forest refuges for the same organisms is much more difficult. They are evidently not constant for different past dry cycles; the present patterns of topography, soils, and climate, as well as biotic relationships, show primarily the effects of the most recent forest retraction, but other relationships are indicated which can only be understood by assuming different refuge patterns in previous cycles. The refuges appear to have been quantitative and statistical, not qualitative phenomena. Pedological and botanical studies indicate that they differed from interrefugial regions only in their greater continuity of optimal forest habitat. Thus, refuges were regions where ecological conditions determined that the probability for the conservation and differentiation of the majority of forest populations would be greater than that for their extinction. During dry periods, the mosaic of favorable (closed) and unfavorable (open) forest habitats was biased towards the former in the refuges and towards the latter between them, but smaller and more scattered forest enclaves existed in non-refuge regions just as non-forest habitat existed within refuges. The resulting irregular distribution of forest species, and non-homogeneity of the coevolving forest biotas, led to variable biogeographical patterns observable today, introducing considerable "noise" into the analysis of the differentiation patterns of restricted groups or individual species. For best preservation of the widest range of gene pools which would represent the unique and incomparable diversity of the Neotropical forest flora and fauna developed over many millions of years, both theoretical and pragmatic criteria should be adopted. The data-base for determination of forest refuges is probably still too narrow to permit the exclusive use of the resulting models for assigning of priorities for habitat conservation. Furthermore, many important, restricted endemic species, especially "r-selected" populations with the high productivity usually sought by man, are not forest denizens, but evolved in more variable or marginal habitats. Even in the typical deep-forest biota, it may be shown that many of the rarest and most primitive species occur in marginal areas of heterogeneous and suboptimal habitat, where they can presumably still compete with their younger and more vigorous relatives. Such perirefugial regions of widely variable topography, soils, and plant formations are well known to contain the richest and most diverse flora and fauna in the Neotropics. Thus, until more complete data can be gathered on a wide sample of Neotropical biotas, the best strategy would seem to be preservation of as much as possible of what primary habitats still remain. As intensive scientific studies reveal the characteristics of different regions some could be unconditionally liberated for human occupation, others opened for limited and rational colonization, and others permanently closed to penetration by man, constituting biotic and geological reserves. These reserves should correspond to broadly based models of forest refuges indicated in diverse groups of plants and animals, but should also be confirmed as exceptionally rich areas of high endemism, and include specialized non-forest habitats and complete river headwater basins. And, as even such refuges will not guarantee preservation of all the genetic variability present, a policy of restricted occupation should be enforced in other regions, within the present law which prohibits destruction of headwater basins and cutting on steep slopes, and requires permanent preservation of 50% of the original vegetation in all areas. Of the 38 refuges proposed in this paper, three (Marañón, Ventuari and Araguaia) are already greatly fragmented by the presentday deteriorating climate conditions, and another thirteen show high percentages of suboptimal forest. This indicates that the next dry cycle will probably follow the past pattern in showing refuges in differing positions, determined by different continental climate patterns, in relation to those operative before the present interglacial; this should lead to widespread extinction of extant biotas already greatly reduced today. Effective biological reserves exist in only 15 of the 38 refuge areas, though plans already exist for preservation of some others. Among the possible protected biotas appear only one of the seven known in Colombia, five of the seven in Venezuela, four of the seven in Peru (with two more proposed or underway), and five of the fourteen known in Brazil. Seven refuges are so small in size, that their total destruction could occur very easily; only two of these (Chiriqui, Imataca) have protected areas today, but both these are already heavily cut over. Twenty-two of the refuge areas, including six of the smallest ones, have already been greatly modified by the intervention of man; one of these (Guaporé), in southwestern Brazil, still almost unknown to scientists and including many unique Andean elements in its biota, is in the last stages of total destruction by ranching interests; it should constitute the first large South American biota to disappear before being minimally studied by biologists. A revision of the first four tribes of the Ithomiinae (Tithoreini, Melinaeini, Mechanitini and Napeogenini) was undertaken, to supplement those published without adequate biological data, in 1956-71 by Fox and Fox and Real. All the genera in the subfamily were classified biologically, with habitat and foodplants specified for most of the genera revised. Complete revisions of Roswellia and Tithorea included the correct placement of "Athesis clearista" vitrala Kaye sensu Fox as a subspecies of Roswellia acrisione, with morphological and biological data based on six specimens captured in December 1975 in northwestern Peru (multiplying by four the number known); and ressuscitation of a number of good geographical subspecies of Tithorea tarricina and T. harmonia synonymized by Fox, with description of the new subspecies T.t. franciscoi (Sierra Perijá and San Juan de Colon, northwestern Venezuela), T. h. dorada (Bolivar, southeastern Venezuela), and T h. gilberti (Jaén, on the upper rio Marañón in northwestern Peru). A new subspecies of the Andean Eutrtesis hypereia (E. h. imeriensis) was described from the Pico da Neblina in northwestern Brazil. Complete revisions of Melinaea and Mechanitis, already in publication, reduced the number of species from 19 and 9 to 7 and 5, respectively. Scada echo Fox was synonymized with S. theaphia, and the distribution of S. karschina delicata, known only from the pair of types labelled "Brazil", was fixed in Pernambuco, through the discovery of two more males in public collections. Preliminary revisions of Hypothyris and Napeogenes reduced the number of species in these two genera from 33 and 45 to 16 and 19, respectively, with indication of 25 new subspecies and two new species; one of these (Hypothyris lema, from higher elevations in Bolívar, in extreme southeastern Venezuela) was described.
