Effects of habitat fragmentation on non-volant small mammals in Macacu River watershed, RJ, Brazil: from individuals to communities.
Habitat fragmentation is a process with detectable effects at individual, population, and community levels of ecological organization. Its effects at the individual level can be detected by indirect measures of fitness and its determinants, such as body condition and use of space, respectively. At the landscape scale, the structural characteristics of the fragments and the ongoing processes of fragmentation directly affect the survival of small mammal populations and communities. The results are changes in the composition, richness, and relative abundance of species constituting the communities. The main objective of this thesis was to evaluate the effects of habitat fragmentation on individual and community levels of organization for non-volant small mammals in a landscape of fragmented Atlantic Forest in Macacu River watershed, RJ, Brazil. Chapter 1 evaluated the effects of forest fragments (size, isolation, and amount of edge) and matrix (property ownership and economic use) characteristics, habitat structure, climatic seasons, and slope orientation on community parameters (species richness, abundance, species composition, and beta-diversity). Small mammals were surveyed in 27 fragments and three continuous forest sites, from 1999 to 2009, with a constant effort of 800 trap-nights per site. The effects of fragment and matrix characteristics on community parameters were evaluated by a model selection approach. Habitat structure was described by nine variables, associated with species composition and abundance through Multidimensional Scaling and Multiple Regression. The effects of climatic seasons and slope orientation on community parameters were evaluated by t-tests. A total of 1023 individuals of eight marsupial and 12 rodent species were captured. Didelphis aurita Wied-Neuwied, 1826, Philander frenatus (Olfers, 1818), and Akodon cursor (Winge, 1887) were the most abundant species accounted for 75.7% of individuals captured. The variation in community parameters was best explained by different factors: species richness by isolation, beta-diversity by property ownership and amount of edge, and species composition by isolation and amount of edge. As expected, more isolated fragments were poor in species. For beta-diversity, property ownership effects were attributed to different use of the fragments by the human population, resulting in different habitat heterogeneity and disturbance intensities. Canopy and understory openness, as well as presence of fallen logs, ordered communities by species composition and abundance. Didelphis aurita and Marmosa paraguayana (Tate, 1931) were considered habitat generalists. Akodon cursor was more abundant in habitats with more open canopy and understory, and smaller presence of fallen logs. The inverse pattern was found for P. frenatus. Community parameters did not differ between climatic seasons and slope orientation. Chapter 2 evaluated the effects of habitat fragmentation on body condition of the marsupials D. aurita and P. frenatus. Body condition was measured by least square means of body mass adjusted for a fixed body length in a generalized linear model design. The effects of fragment (size, amount of edge, and abundance of Astrocaryum aculeatissimum) and matrix characteristics (property ownership and economic use), age of individuals, and climatic season on body condition of D. aurita and P. frenatus were evaluated using a model selection approach. A total of 267 individuals of D. aurita and 85 of P. frenatus were captured in three continuous forest sites, and 87 of D. aurita and 185 of P. frenatus in 23 forest fragments. Individuals of both species had better body condition in continuous forest sites than in forest fragments, demonstrating that their fitness was negatively affected by habitat fragmentation in the Macacu river watershed. Although the populations of both species persist in fragments, they may do so in poorer condition. In fragments, age was the main determinant of body condition of D. aurita, with adults in better body condition than sub-adults. This pattern may be related to the trade-off between energy allocation for growth versus reproduction. Body condition of P. frenatus was mainly determined by property ownership. Individuals of P. frenatus were in better condition when the fragment was surrounded by large rural properties, which are less disturbed by livestock and human activities such as timber and hunting. Chapter 3 described and evaluated the use of space by P. frenatus in two small forest fragments, and compared it with the use of space in a continuous forest site. From 2007 to 2009, a capture-mark-recapture study was carried out in two forest fragments. Captured adults of P. frenatus were released with a spool-and-line device that allowed a detailed tracking of the animal path. Use of space was evaluated by daily home range (DHR) using the Minimum Convex Polygon method, path tortuosity using Fractal D, and percentage of vertical use. Use of the matrix and shelter was described as well. In continuous forest, located in the Serra dos Orgaos National Park, data were obtained using a similar protocol, from 1998 to 2005, as part of a previous capture-recapture program of the Laboratorio de Vertebrados, UFRJ. The effects of site (fragments vs. continuous forest site), sex, climatic season (dry vs. wet), and amount of thread tracked were evaluated by model selection approach. Frequencies of diameters and slopes of supports used in the vertical strata were compared between sites, sex, and frequencies of availability in the environment using G-tests. Eighty paths in fragments and 39 in continuous forest were tracked. The less tortuous paths in fragments can be related to more open vegetation in the understory, representing fewer obstacles to movements on the ground. A tendency for smaller DHRs in fragments compared to continuous forest and for females were found, and are probably related to high population density of P. frenatus and small size of the fragments (<10ha). Also, smaller DHRs in fragments were compensated by a greater vertical use. In the rainy season, a period of increased resource availability, the DHRs were smaller because individuals could meet their energy requirements covering smaller areas. Differences in frequency of use of diameters and slopes of supports between sites and sexes can be related to differences in body size. Individuals rarely used the matrix. Negative effects of habitat fragmentation on non-volant small mammals in the Macacu river watershed were detected at different levels of ecological organization, from individuals to communities. The evaluation of community structure, combined with measures of body condition, demonstrates that the two marsupial species more abundant in forest fragments, P. frenatus and D. aurita, persist in fragments in worse condition than in continuous forest sites. The effects of habitat fragmentation varied depending on the level of ecological organization: community structure was affected mainly by fragment characteristics whereas body condition mainly by matrix characteristics. More precise evaluations of the effects of habitat fragmentation should consider its effects on community, population, and individual scales.
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||RESUMENES DE TESIS|
|Date:||Jun 1, 2012|
|Previous Article:||Murcielagos cavernicolas de la provincia de Cordoba, Argentina.|
|Next Article:||Analisis demografico de la respuesta poblacional de Akodon azarae al parasitismo de la mosca Rogenhofera bonaerensis (Diptera, Cuterebridae).|