Living entities have their own means of locomotion and are capable of reproduction. Furthermore, the habitat in which organisms are embedded is typically heterogeneous, such that environment conditions vary in time and space. In this thesis, theoretical models to understand the collective dynamics of living beings have been proposed and investigated aiming to address relevant questions such as population organization and persistence in the environment, using analytical and numerical techniques. Initially, considering an homogeneous habitat, in which the statistical properties of the environmental conditions are time and space independent, we study how spatiotemporal order can emerge in the population distribution due to nonlocal interactions and investigate the role of environment fluctuations in the self-organization process. Further, we continue our investigation assuming an heterogeneous environment, starting with the simplest case of a single habitat domain, and we obtain the critical conditions for population survival for different population dynamics. Considering a class of nonlinear equations, introducing temporal oscillations and interactions among the organisms, we are able to provide a general picture of population stability in a single habitat domain, challenging previous ecological concepts. At last, assuming a fragmented complex landscape, resembling realistic properties observed in nature, we additionally assume that individuals have access to information about the spatial structure. We show that individuals survive when patches of viable regions are clustered enough and, counter-intuitively, observe that population size is maximized when individuals have partial information about the habitat. Finally, since, analytical exact results are not feasible in many important situations, we propose an effective approach to interpret experimental data. This way we are able to connect environment heterogeneity and population persistence.