Natural scientists have been always attracted by the study of the phenomenon of Life, since it displays a plethora of curious and yet puzzling behaviors. In the last decades it has been registered an increasing interest in the investigations of ecological and biological systems by the Physics community. This stems from the fact that the physical discipline of Statistical Mechanics offers many tools, frameworks and ideas that have turned out to be naturally adapted, as well as very efficient, to deal with systems affected by an huge degree of complexity, like living systems are. In this Thesis we embrace such a perspective and so we tackle ecological and biological topics employing a Statistical Mechanics mindset.% and we focus on three different remarkable features, which are hallmarks of the complex nature of living systems. We firstly model ecological communities in which several different species compete for the consumption of a shared pool of resources with the aim of understanding how the huge biodiversity empirically encountered can originate. To do so, we extended the celebrated MacArthur's consumer-resource model to account for spatial contributions, originating from a variety of ecological mechanisms, in an effective way. Thanks to this, we show analytically the model predicts several species coexisting while competing for a limited number of resources, in complete agreement with evidences coming from empirical observations. This is solely due to the modification we introduce, based on both physical and ecological arguments, since such a result can not be obtained within the classical formulation of the model. Then, we move our attention to study the universal features of self-organized regular spatial structures, which can be found in both empirical and theoretical ecological investigations. Due to their wide diffusion also in other scientific fields, we search for any universal behavior in their spatio-temporal evolution, regardless the microscopic peculiarities characterizing a certain system. We provide a mathematical framework able to state whether such patterns emerge or not. More interestingly, in the pattern formation phase of the model, we are able to show that it exists a regime in which the evolution of the envelope of such spatial structures on long timescales and large spatial scales is model independent, i.e., it is governed by an equation, whose shape does not dependent on the dynamics details. Finally, motivated by real-world biological scenarios, we build a theoretical framework, which acquires the form of a generalized Langevin dynamics, accounting for demographic stochastic contributions and temporal delays effects. Hence we model systems whose evolution, subjected to noisy effects, is determined also by the past states visited by the system. We demonstrate how such a framework predicts quite naturally the emergence of almost regular oscillating behaviors, in the form of noise-induced cycles, in the temporal evolution of the system. We then apply these theoretical findings to understand the experimental results studying gene expression regulatory networks, in which noise and delay contributions indeed are at stake.
Gli scienziati sono sempre stati attratti dallo studio del fenomeno della vita, poiché essa mostra una pletora di comportamenti curiosi ed enigmatici. In particolare, negli ultimi decenni si è registrato un crescente interesse da parte della comunità composta dai fisici nei confronti dei sistemi ecologici e biologici. Questo deriva dal fatto che la disciplina della Meccanica Statistica offre molti strumenti, framework e idee che si sono rivelati naturalmente adatti, oltre che molto efficienti, per trattare sistemi affetti da un enorme grado di complessità, come sono i sistemi viventi. In questa tesi abbracciamo tale prospettiva e quindi affrontiamo tematiche bio-ecologiche utilizzando una mentalità meccanico-statistica. Innanzitutto proponiamo un modello che descrive comunità ecologiche in cui diverse specie competono per il consumo di un pool condiviso di risorse, con l'obiettivo di capire come possa avere origine l'enorme biodiversità che si incontra empiricamente in molti ecosistemi. Per fare ciò, abbiamo esteso il celebre modello di MacArthur per tendere conto in modo efficace di contributi spaziali, originati da una varietà di meccanismi ecologici. Grazie a ciò, dimostriamo analiticamente che la dinamica prevede la coesistenza di diverse specie in competizione per un numero limitato di risorse, in completo accordo con le evidenze provenienti dalle osservazioni empiriche. Questo è dovuto esclusivamente alla modifica che introduciamo, basata su argomenti sia fisici che ecologici, poiché un tale risultato non può essere ottenuto nella formulazione classica del modello. In seguito, spostiamo la nostra attenzione sullo studio delle caratteristiche universali delle strutture spaziali regolari auto-organizzate, che possono essere trovate in indagini ecologiche sia empiriche che teoriche. Data la loro ampia diffusione anche in altri campi scientifici, ricerchiamo l’esistenza di un comportamento universale nella loro evoluzione spazio-temporale, indipendentemente dalle peculiarità microscopiche che caratterizzano un certo sistema. Forniamo quindi un framework matematico in grado di stabilire se tali pattern emergano o meno. Inoltre, nella fase di pattern-formation del modello, siamo in grado di mostrare che esiste un regime in cui l'evoluzione dell'involucro di tali strutture spaziali su lunghe scale temporali e grandi scale spaziali è indipendente dal modello, cioè che tale evoluzione è governata da un'equazione, la cui forma non dipende dai dettagli della dinamica. Infine, motivati da scenari biologici reali, costruiamo un quadro teorico che acquisisce la forma di una dinamica di Langevin generalizzata, tenendo conto dei contributi stocastici demografici e degli effetti dei ritardi temporali. Quindi modelliamo sistemi la cui evoluzione, soggetta a effetti stocastici, è determinata anche dagli stati visitati in passato dal sistema. Dimostriamo come un tale framework predica abbastanza naturalmente l'emergere di comportamenti oscillatori quasi regolari, sotto forma di cicli indotti dal rumore, nell'evoluzione temporale del sistema. Applichiamo poi questi risultati teorici per comprendere le evidenze provenienti da esperimenti che studiano reti di regolazione dell'espressione genica all’interno delle cellule, in cui i contributi del rumore e del ritardo sono effettivamente in gioco.
Approccio fisico-statistico a tematiche ecologiche e biologiche / Garlaschi, Stefano. - (2021 Dec 22).
Approccio fisico-statistico a tematiche ecologiche e biologiche
GARLASCHI, STEFANO
2021
Abstract
Natural scientists have been always attracted by the study of the phenomenon of Life, since it displays a plethora of curious and yet puzzling behaviors. In the last decades it has been registered an increasing interest in the investigations of ecological and biological systems by the Physics community. This stems from the fact that the physical discipline of Statistical Mechanics offers many tools, frameworks and ideas that have turned out to be naturally adapted, as well as very efficient, to deal with systems affected by an huge degree of complexity, like living systems are. In this Thesis we embrace such a perspective and so we tackle ecological and biological topics employing a Statistical Mechanics mindset.% and we focus on three different remarkable features, which are hallmarks of the complex nature of living systems. We firstly model ecological communities in which several different species compete for the consumption of a shared pool of resources with the aim of understanding how the huge biodiversity empirically encountered can originate. To do so, we extended the celebrated MacArthur's consumer-resource model to account for spatial contributions, originating from a variety of ecological mechanisms, in an effective way. Thanks to this, we show analytically the model predicts several species coexisting while competing for a limited number of resources, in complete agreement with evidences coming from empirical observations. This is solely due to the modification we introduce, based on both physical and ecological arguments, since such a result can not be obtained within the classical formulation of the model. Then, we move our attention to study the universal features of self-organized regular spatial structures, which can be found in both empirical and theoretical ecological investigations. Due to their wide diffusion also in other scientific fields, we search for any universal behavior in their spatio-temporal evolution, regardless the microscopic peculiarities characterizing a certain system. We provide a mathematical framework able to state whether such patterns emerge or not. More interestingly, in the pattern formation phase of the model, we are able to show that it exists a regime in which the evolution of the envelope of such spatial structures on long timescales and large spatial scales is model independent, i.e., it is governed by an equation, whose shape does not dependent on the dynamics details. Finally, motivated by real-world biological scenarios, we build a theoretical framework, which acquires the form of a generalized Langevin dynamics, accounting for demographic stochastic contributions and temporal delays effects. Hence we model systems whose evolution, subjected to noisy effects, is determined also by the past states visited by the system. We demonstrate how such a framework predicts quite naturally the emergence of almost regular oscillating behaviors, in the form of noise-induced cycles, in the temporal evolution of the system. We then apply these theoretical findings to understand the experimental results studying gene expression regulatory networks, in which noise and delay contributions indeed are at stake.File | Dimensione | Formato | |
---|---|---|---|
tesi_definitiva_Stefano_Garlaschi.pdf
accesso aperto
Descrizione: tesi_definitiva_Stefano_Garlaschi
Tipologia:
Tesi di dottorato
Dimensione
12.1 MB
Formato
Adobe PDF
|
12.1 MB | Adobe PDF | Visualizza/Apri |
Pubblicazioni consigliate
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.