We study observational signatures of non-gravitational interactions between the dark components of the cosmic fluid, which can be either due to creation of dark particles from the expanding vacuum or an effect of the clustering of a dynamical dark energy. In particular, we analyse a class of interacting models ($\Lambda$(t)CDM), characterised by the parameter $\alpha$, that behaves at background level like cold matter at early times and tends to a cosmological constant in the asymptotic future. In our analysis we consider both background and primordial perturbations evolutions of the model. We use Cosmic Microwave Background (CMB) data together with late time observations, such as the Joint Light-curve Analysis (JLA) supernovae data, the Hubble Space Telescope (HST) measurement of the local value of the Hubble-Lema\^itre parameter, and primordial deuterium abundance from Ly$\alpha$ systems to test the observational viability of the model and some of its extensions. We found that there is no preference for values of $\alpha$ different from zero (characterising interaction), even if there are some indications for positive values when the minimal $\Lambda$(t)CDM model is analysed. When extra degrees of freedom in the relativistic component of the cosmic fluid are considered, the data favour negative values of $\alpha$, which means an energy flux from dark energy to dark matter.

We investigate the observational effects of a quintessence model in an anisotropic spacetime. The anisotropic metric is a non-rotating particular case of a generalized Godel's metric and is classified as Bianchi III. This metric is an exact solution of the Einstein-Klein-Gordon field equations with an anisotropic scalar field, which is responsible for the anisotropy of the spacetime geometry. We test the model against observations of type Ia supernovae, analyzing the SDSS dataset calibrated with the MLCS2k2 fitter, and the results are compared to standard quintessence models with Ratra-Peebles potentials. We obtain a good agreement with observations, with best values for the matter and curvature density parameters $\Omega_M = 0.29$ and $\Omega_k= 0.01$ respectively. We conclude that present SNe Ia observations cannot, alone, distinguish a possible anisotropic axis in the cosmos.