Dinâmica populacional e efeitos de variáveis ambientais sobre a fauna de pequenos mamíferos em um fragmento de floresta com araucária no sul do Brasil

Natural populations of small rodents fluctuate continually over time and the population dynamics of any living organism are shaped by both intrinsic and extrinsic factors. Small mammals populations can show a continuum dynamic activity, ranging from stability to regular cycles of variable amplitude. The species dynamics is also known to be influenced by the seasonal structure of the environment. The community of small mammals present in the Araucaria forest could present similar patterns to forests occurring in other temperate regions of the planet and the large seeds of Araucaria angustifolia (dominant tree) are consumed by a large array of vertebrates, mainly rodents and large birds, which act as seed dispersers and predators simultaneously. When species are very similar like the two most abundant of this community (Akodon montensis and Oligoryzomys nigripes) the outcome of competition is often the partitioning of resources such as habitat or food. Plant species are crucial for rodents, because they are the main food source for many species, and this particular relation between rodents and plants is crucial for the process of forest regeneration, since the regeneration development are associated with details on the role of each rodent species as a seed disperser or seed predator. There are a lot of important factors interfering in the regeneration process, and changes in these factors are manifested in variables such as density of tree individuals, basal area, floristic composition, and species richness and diversity. To analyse the populational aspects of the rodents in the area and identify correlations between the Araucaria angustifolia trees with rodent community, to evaluate arthropod-small mammals species association and test the relationship between natural forest regeneration and rodent community, we established a trapping grid of two hectares with 231 trap stations (11 X 21 configuration, 10m spacing). Every trapping series, which consisted of 6 days, traps were placed on the ground and a capture-mark-recapture program was carried out two times per season between November 2008 and August 2009. We followed Cormarck-Jolly-Seber (CJS) method to estimate population parameters and density was estimated for each trapping session. The mean of population density was used as a dependent variable in a linear regression to investigate possible relations between densities and extrinsic factors. The arthropod abundance was quantified with pitfall traps during four of the eight sample periods and we measured regeneration variables that could potentially influence the spatial distribution of the small mammals. The association between rodent and arthropod communities was compared with the similarity matrix based on rodent composition using simple Mantel tests and we used DCA analysis to ordinate sample units based on rodent community. Significance of the associations between regeneration and rodent community was tested by 1000 Monte Carlo permutations based on Pearson's correlation. The populations of the two most abundant rodents in the community (Akodon montensis and Oligoryzomys nigripes) were considered inconstant and their population variations could have indirect effects on other species. Our result suggests that these two species are the main small mammal present in the area and if extrinsic factors do not operate all the times as showed, it strongly suggests that there must be variation in spacing and dispersal behavior of these species or extrinsic factors can also be acting in indirect ways. The presence of the species Mus musculus indicates that this forest area had or still haves anthropic influence. In general, secondary forests provide important habitats and resources to the rodent community and are distinctly associated to the species. Although indirectly, our results indicate that there is a complex combination of regeneration predation/dispersal by rodents and it seems that generalist species might select habitat characteristics primarily at a site level because they are able to use different local factors that exist in a variety of landscapes. Based on our results we can assume that small mammals associated with herbaceous or shrub cover, particularly in riparian areas, will decline when deforestation remove this cover. We also provide initial insight for habitat features that are related to rodent community pointed out the correlations between numbers of individuals and natural regeneration