Bonaventura, Luca

Protecting key hotspots of marine biodiversity is essential to maintain ecosystem services at large spatial scales. Protected areas serve not only as sources of propagules colonizing other habitats, but also as receptors, thus acting as protected nurseries. To quantify the geographical extent and the temporal persistence of ecological benefits resulting from protection, we investigate larval connectivity within a remote archipelago, characterized by a strong spatial gradient of human impact from pristine to heavily exploited: the Northern Line Islands (NLIs), including part of the Pacific Remote Islands Marine National Monument (PRI-MNM). Larvae are described as passive Lagrangian particles transported by oceanic currents obtained from a oceanographic reanalysis. We compare different simulation schemes and compute connectivity measures (larval exchange probabilities and minimum/average larval dispersal distances from target islands). To explore the role of PRI-MNM in protecting marine organisms with pelagic larval stages, we drive millions of individual-based simulations for various Pelagic Larval Durations (PLDs), in all release seasons, and over a two-decades time horizon (1991-2010). We find that connectivity in the NLIs is spatially asymmetric and displays significant intra- and inter-annual variations. The islands belonging to PRI-MNM act more as sinks than sources of larvae, and connectivity is higher during the winter-spring period. In multi-annual analyses, yearly averaged southward connectivity significantly and negatively correlates with climatological anomalies (El Niño). This points out a possible system fragility and susceptibility to global warming. Quantitative assessments of large-scale, long-term marine connectivity patterns help understand region-specific, ecologically relevant interactions between islands. This is fundamental for devising scientifically-based protection strategies, which must be space- and time-varying to cope with the challenges posed by the concurrent pressures of human exploitation and global climate change.

An improved version of the PDAC (Pyroclastic Dispersal Analysis Code) numerical model for the simulation of multiphase volcanic flows is presented and validated for the simulation of multiphase volcanic jets in supersonic regimes. The present version of PDAC includes second-order time and space discretizations and fully multidimensional advection discretizations, in order to reduce numerical diffusion and enhance the accuracy of the original model. The resulting numerical model is tested against the problem of jet decompression in both two and three dimensions. For homogeneous jets, numerical results show a good quantitative agreement with experimental results on the laboratory scale in terms of Mach disk location (Lewis and Carlson, 1964). For multiphase jets, we consider monodisperse and polydisperse mixtures of particles with different diameter. For fine particles, for which the pseudogas limit is valid, the multiphase model correctly reproduces predictions of the pseudogas model. We obtain that particles are in mechanical and thermal equilibrium with the gas phase and the jet decompression structure is in quantitative agreement with pseudogas results (Ogden et al., 2008b). For both fine and coarse particles, we measure the importance of multiphase effects with relation to the characteristic time scales of multiphase jets and we quantify how particles affect the average jet dynamics in terms of pressure, mixture density, vertical velocity and temperature. Furthermore, time dependent vent conditions are introduced, in order to achieve numerical simulation of eruption regimes characterized by transient jet behaviour. We show how in case of rapid change in vent conditions, volcanic jet structures do not evolve through a succession of steady state configurations and the transition between different flow conditions can result in the collapse of the volcanic column.