Tu, Chengyi (2018) Resilience, Complexity and Cooperation in Socio-ecological System. [Ph.D. thesis]
Full text disponibile come:
Statistiche DownloadSimple MetadataFull MetadataEndNote Format
Advances in experimental technologies, both in the laboratory and in the field, are generating an increasing volume of ecologically and sociologically relevant data, spanning a wide range of scales, revealing recurrent emergence of patterns in these systems. This "data explosion"' is both a challenge (inventing new tools for their analysis) and an opportunity (identifying rules driving the functioning of complex systems). However, data alone do not necessarily lead to an understanding of the systems of interest. At present, we are in a situation where in front of a rich (but common to many systems) phenomenology we have innumerable models for very specific cases that call for a general vision. This challenge is very fascinating for physicists, that have in their veins the search for general principles of apparently different phenomena.
In particular, a very important property that seems to be shared by most of the socio-ecological systems is their ability to respond to perturbations, i.e. the system resilience. Cell biology, ecology, environmental science, and food security are just some of the many areas of investigation on the mechanisms increasing the system resilience. Nevertheless, not all socio-ecological systems display high resilience. In food security, the intensification of international food trade and local shocks in food production led to global food crises, and for example Suweis develops a framework to investigate the coupled global food-population dynamics and finds that the global food system is losing resilience (increasingly unstable and susceptible to conditions of crisis); In ecology, the concept of resilience has evolved considerably since Holling's (1973) seminal paper to describe the persistence of natural systems in the face of changes in ecosystem variables due to natural or anthropogenic causes. It has been suggested that in many ecosystems we are facing a lost of resilience and consequent loss of biodiversity. Therefore an important challenge is to understand what are the main drivers ruling the resilience of ecological communities, so that proper ecosystem management strategy can be developed.
From data is emerging that one of the key feature of socio-ecological system resilience may lie in the architecture of the interaction networks. The topology of the interaction network may actually represent the "parameter" that system somehow self-tunes so that the system's responses to stimuli is optimized with respect to some feature (e.g. stability). In inanimate matter, spins or particles always have their mutual interactions turned on (with an intensity decaying with their relative distance) and the network describing their interaction is dense, with most of the connections present. In contrast, if we consider for instance an ecosystem, species interact selectively even if they coexist at short distances, and the species interaction network is sparse, that is, most of the interactions are turned off. At the same time, the interactions that are turned on form non-random evolving structures that are the result of some optimization process under adaptive/evolution pressure. Thanks to massive databases now easy available, characteristics similar to those just mentioned for ecological networks, have been observed also in gene-interaction network, in neuronal networks and even in social networks. These networks are very different and yet share a crucial aspect: they all have undergone biological/social evolution that has driven their incremental complexity.
One particular long-standing question regards the relationship between stability (resilience) and complexity in ecological system. Many of the population dynamics modeling frameworks proposed in the literature cannot elude the celebrated May's theorem. This theorem, recently refined by Allesina and Tang states that the stability of the system depends on the product [SC], where [S] is the number of species and [C] is the fraction of non-zero pairwise interactions between species. This result leads to the so-called stability and complexity paradox debate: a system in order to be stable cannot be too large ([S] large) or too connected (large [C]). The paradox lies in the fact that empirically, ecosystems with a large number of species seem to be very stable. Moreover, recently it has been suggested that because of this stability paradox, in microbial ecosystems competition may play a much important roles than what expected until now. In fact in these models, competition has a stabilizing role in ecosystem dynamics, contrarily to cooperation that decreases the ecosystem resilience.
During my Ph.D. I have used a physicists approach based on complex networks and statistical physics, to study the resilience in Socio-Ecological systems, how it is related to the system complexity and what is the role of cooperation in the ecosystem dynamics. I have used a comprehensive approach that includes data mining, theoretical modeling (both computational and analytical) and statistical analyses. In particular, I have investigated the efficiency of a recently proposed framework to study the resilience of complex interacting systems, what the role of cooperation and competition in the universal patterns theoretically predicted by the model, and its validation with data. I have then focused on the long-standing open question of the relation between complexity and resilience in ecosystems, by specifically focusing on how the architecture of interaction networks may confer to living systems their ability to promptly react to to perturbations (e.g. increase resilience). To do that we have developed a stochastic population dynamics model, generalizing an interacting non-equilibrium model known as the voter model, and I have also studied the effect of cooperation on the ecosystem resilience. The results of my work suggest a novel picture on the relation between complexity, cooperation and resilience, challenging previous results in the literature.
Solo per lo Staff dell Archivio: Modifica questo record