Community diversity pattern depends highly on spatial scale. One of the reasons is that the balance between deterministic and stochastic processes arguably changes at different spatial scales. We do not know yet to which extent deterministic and stochastic processes affect species richness and abundance, and understanding processes behind biodiversity pattern formation is critical for predicting the system’s behaviour including its stability. To study community dynamics at local and regional spatial scales together, I am using a metacommunity model, in which local communities are connected by dispersal, and both local and regional processes take place. The model follows lottery competition between larvae for space to settle into, and environmental gradient in the system causes a difference in species fecundity. I analyse alpha, beta, and gamma diversity obtained from the simulations under neutral (without environmental gradient) and non-neural (with environmental gradient) conditions. I will show how spatial heterogeneity of species changes depending on the network structure of the larval dispersal connectivity and the intensity of deterministic and stochastic processes. This study will relate spatial structure and spatial biodiversity pattern, through investigating the effect of stochastic and deterministic processes on biodiversity.