acesap

Activity 4: Modelling

A4-wp1 - Climatic Modelling

Participants: Piero Cau, Emanuele Eccel

The production of a high-resolution climate characterization is the focus of the investigation on the existing time series of temperature, precipitation and derived climatic indicators. Particularly for the past 20 - 30 years, possible links will be sought between the assessed climatic shifts and the observable changes in Trentino's biota. In order to define in a high spatial detail the state of the physical environment and its evolution, first of all the present conditions will be assessed. A4-WP2 will collaborate to produce a spatial interpolation on the main climatic quantities: temperature, precipitation, direct indices (such as absolute maximum, mean, extreme values) and derived indices (e.g., length of vegetative season, monthly rain days, etc...). According to the different use of processed data, the integration time will range from day (for derivation of indices) to ten-day or monthly periods (for average values). Aiming to a snapshot of the present conditions, it would be advisable to relate the investigation to the most recent data, rather than to the standard canonical climatic period 1961-1990, at least for some species (namely animals). Therefore the majority of the available historical series dating back to about 30 years will be considered - several dozens in Trentino - having care to sample the little-anthropised territory as well as possible. More detail will be sought in "target areas", significant for sampling, as Mt. Baldo. Also series from neighbouring regions (South Tyrol, Lombardy, Veneto) will be considered, in order to improve spatial coverage at border areas. For areas not contiguous with the province of Trento, corresponding climatic series will be provided. Particular requirements, such as the climatic characterization of radiation indices, air humidity, soil temperature, snow cover, microclimatic features, will be taken into account where possible. The selected instrumental series will undergo validation, aiming to excluding clearly or probably erroneous data from the analysis, data gap filling, and homogenisation, to tackle major measure shifts due to station displacements, instrument changes, data collection protocols, or systematic erroneous periods.

By using the same climatic quantities, attention will be drawn on the most recent change, happened in the last 20 - 30 years and whose consequences may be already observable. Secondly, climatic evolution scenarios will be considered, suitably downscaling data to reach a sound quantitative expression of the climatic signal for the future (about 50 years); the above-mentioned instrumental series will be used for this purpose. Output from general circulation, coupled atmosphere - ocean models (AO-GCM) will be the base for climatic projections.

An increase in time resolution, useful for the calculation of some important climatic indices, will be carried out by producing simulated series with weather generating algorithms, yielding daily resolution from the monthly model output. These algorithms will be calibrated with homogenised station daily series. Climatic evolution will be represented by expressing the relevant climatic features in "standard" time-slices: 1961-90, 2021-50, and, if useful, 2071-99.

A4-wp2 - GIS modelling - landscape genetics

Participants: Duccio Rocchini, Cristina Castellani, Markus Neteler

Genetic diversity is important for the maintenance of the viability and the evolutionary and adaptive potential of populations and species.

Two previously separate research areas, genetics and landscape ecology have been integrated into the new discipline "landscape genetics" (see Holderegger and Wagner, Bioscience, 2008). This combined approach merges population genetics, landscape ecology and spatial statistics, typically performed in a GIS (Geographical Information System) environment. The combination of genetic markers with related spatio-environmental data is used to examine population demographics and evolutionary processes. For this purpose, genetic characteristics of a species are mapped across a landscape or differences between population are modeled by using landscape compositional and structural metrics. Neutral markers such as mitochondrial and Y chromosome haplotypes, microsatellite frequencies, single-nucleotide polymorphisms as well as genetic markers like the Major Histocompatibility Complex (MHC) allele frequencies are used.

Genetic diversity can be adaptive or neutral. Selectively neutral genetic variation is generally believed to not affect the fitness of the species. From patterns of neutral genetic diversity and differentiation demographic and evolutionary events like bottleneck, expansion, isolation, gene flow, divergence can be inferred. Adaptive markers, in contrast, are subject to selection constraints and are, therefore, better suited for studying the response to environmental changes. The opportunity to analyse neutral and adaptive variation in a GIS framework makes it possible to start evaluating i) the putative role of biotic and abiotic factors in shaping genetic diversity and ii) the differentiation in a determined area.

For instance, the genetic and landscape ecological data can be analysed for identifying barriers, gradients or transitions thus obtaining crucial information about connectivity among natural populations. By separating historic and recent gene flow, global and local changes may be identified which lead to changes, sometimes resulting in a loss of biodiversity. As an advantage, landscape genetics does not usually require to distinguish discrete populations in advance. Analyses are performed at population as well as individual levels.

We will collect and integrate field data with GIS and climatic data, e.g. to find current and potential faunal corridors in Trentino. For key species, hypothesis will be discussed about population dimensions, measured as effective population size by means of molecular markers, with regards to expected temperature increase, precipitation decrease and increasing human impact on the territory. As outcome, predictive maps for the next 50 years will be created which display the expected population changes from today to future. As a result of our landscape genetics studies, the role of landscape variables in shaping genetic diversity and population structure will be better understood. The outcome is relevant for managing properly the genetic diversity of threatened and endangered populations. The study of genetic differences would permit to locate biodiversity hotspots or, in a time series, to investigate whether levels of biodiversity has changed. Finally, using Partial Mantel tests the effect of spatial distance between populations will be partialled out and the correlation between the genetic and environmental distance-matrices will be estimated (see e.g. Balkenhol et al., Ecography, 2009). This will allow us to detect not only the location of genetic diversity spots but even the ecological processes accounting for the pair-wise differentiation of populations.

A4-wp3 - Ecological Modelling

Participants: Luca Bolzoni, Roberto Rosà

Mathematical and statistical models have been playing a very important role in the development and understanding of ecology and epidemiology.

To investigate the temporal responses to various stressors on the dynamics of endangered species in Trentino, a series of theoretical models will be developed within this activity. In particular, the dynamics of host-parasite interaction will be explored considering the effect of climate and land-use changes. A series of models will be developed to examine the potential competition between different host species mediated by parasites including the effect of habitat fragmentation on host-parasite interaction. Afterwards, seasonal dynamic models will be implemented for addressing the effect of climatic variables (e.g. temperature and precipitation) on parasite life cycles and the consequent impact on host population dynamics. A case study for these class of models will be the system including alpine galliformes (e.g. Tetrao tetrix and Alectoris graeca saxatilis) and their nematode parasites (e.g. Ascaridia spp.) known to affect host dynamics.

A different class of predictive models will be developed to assess the direct consequences of biodiversity changes to human health. In this case the model case study will include wild rodents, such as Apodemus flavicollis, being a reservoir of several zoonotic pathogens transmissible to humans, both directly and through tick bites. Epidemiological models for rodent-borne and tick-borne diseases will be developed to examine the dynamics of different zoonosis maintained by rodents in Trentino and transmitted to humans directly or through ticks. The establishment and potential spread of different diseases will be evaluated and particular attention will be posed to the effect of host biodiversity, both in terms of species richness and specific host density, on disease transmission.

The general approach will be based on a combination of model development and analysis, parameter estimates and sensitivity analysis. Interactive multi-platform software for numerical and statistical analysis will be used to produce simulations taking into consideration different scenarios.

---