Effetti della qualità degli habitat sull'ecologia dei micromammiferi forestali in paesaggi frammentati dell'Italia centrale
Author(s)
Schiavano, Andrea
Date Issued
June 17, 2013
Type
Doctoral Thesis
Abstract
Habitat fragmentation process and their quality surveying are among the most important matters concerning conservation Biology. Landscape fragmentation or “habitat” fragmentation is one of the main causes of biodiversity decline (Fahrig, 2003, Foley et al., 2005).
The habitat fragmentation concerns four different circumstances:
habitat loss;
habitat subdivision;
habitat degradation;
connectivity break.
The first two phenomenons usually coexist in most of real landscapes and are very difficult to recognize because both can be observed contemporaneously in fragmented landscapes. Habitat degradation leads to a gradual worsening of the habitat quality and it can occur independently of the habitat loss and fragmentation, even if it can be eased by these last two circumstances (Mortelliti et al., 2010b). Connectivity refers to the landscape permeability level to the individuals transfers among the different habitat fragments (Taylor et al., 1993).
We can consider the habitat quality to be the environment ability to promote the endurance of the individuals and the population (Hall et al., 1997), anyway there are also some definition based on the effects the habitat quality produces on demographical parameters and the population dynamics (Van Horne, 1983; Pulliam, 1988; Runge et al., 2006). It is very difficult to evaluate and take into consideration habitat quality according to their role in the biology of a species. Different researches show that the habitat quality in a fragment determines the patterns of distribution in spatially structured populations (Mortelliti et al., 2010b; Thomas et al., 2001) and that habitat fragmentation and habitat quality are often collateral concepts (Mortelliti et al., 2010b).
During the processing of the experimental project and the following analisys, it's important to pay attention to the separation of the independent contribute of environmental resources from the fragments spatial configuration role and, then, to the check of the interaction between the fragment resources and geometrical characteristics.
Firstly, by this study, we would like to understand which resources play a key-role in determining demographical parameters strictly connected to the habitat quality such as fecundity
and density, secondly we want to evaluate the role of resources abundance in population dynamics and patterns of distribution of four species of forestry rodents (Muscardinus avellanarius, Apodemus flavicollis, A. sylvaticus e Myodes glareolus) and, in the end, to give the instruments to elaborate the best conservation strategies for this species in agricoltural landscapes, focusing on conservation actions that can optimize the population endurance with lowest costs.
M. avellanarius has been studied in three different landscapes:
the area of Fara Sabina, in Rieti district (11 siti);
the area of Viterbo e Montefiascone, in Viterbo district (9 siti);
the Regional Natural Reserve “Selva del Lamone” in Viterbo district, considered as a check area because it is a not fragmented environmental continuum (10 grids).
Apodemus flavicollis, A. sylvaticus and Myodes glareolus in:
Viterbo landscape (29 patches);
All these landscapes are characterized by a high level of habitat loss (less than 20% of forestry cover) and fragmentation.
In this work we call “landscape” an arbitrarily well defined and delimited area square shaped (16x16 Km), where we have distinguished some fragments in order that for each of them there would be a spatial repetition into the landscape itself and among different landscapes.
Analisys are based on presence/absence data using occupancy models by MacKenzie et al. (2002, 2003) based on progressing survey that allow to understand the capture probability of a species and to focus the false absences, and linear models are fitted to density and fertility data. Patch and landscape scale metrics together with patch internal structure and resource availability were used as predictor variables.
In both cases the models have been chosen following the ”information-theoretic approach” based on AIC evaluation (Akaike Information Criteria).
According to the results, M. avellanarius presence and density are determined by patch maturity level and by the number of fragments connected to each fragment; colonization processes, instead, are regulated by the forestry habitat around the fragments occupied by the species and their spatial configuration, instead the extinction processes depend on patch size and the abundance of bush species producing fruits that are part of rodents' diet; in the end, fecundity is negatively influenced
by the abundance of food resources and the density values of the species itself.
Concerning A. flavicollis, the abundance of resources (acorn production, abundance and variety of bush species) is very important to explain the density values collected, while connectivity becomes important for extinction probabilitiy. Regard to fecundity, the models we have elaborated don't show clearly which variables influence this parameter. Probably, they have a low influence on the reproduction process of this species.
Colonization/extinction dynamics in A. sylvaticus seem to depend, respectively, from connectivity and acorn production. In fact acorns seem to have positive effects on this rodent reproduction, but have negative ones on density values.
Neither of the variables we have considered seem to influence M. glareolus colonization probability, on the contrary, the patch size and the number of hedgerows influence its extinction probability. This condition, together with the reduction of reproduction ability in the areas where A. flavicollis and A. sylvaticus densities are more elevated, presumes the existence of a inter-species competition that penalises the bank vole.
Additional information
Dottorato di ricerca in Ecologia forestale
File(s)![Thumbnail Image]()
Name
aschiavano_tesid.pdf
Size
7.05 MB
Format
Adobe PDF
Checksum (MD5)
918db1c856ffb61d737e23bbb6563749