Predation models based on differential equations,such as the classical Lotka-Volterra predator-prey model, allow understanding the interactions between populations of prey and predators at a population scale. But what about predation models at an individual scale? How to link both macro and micro scales in order to comprehend these interactions? Only a few existing papers tackle these questions. Some authors proposed elements of response on how it is possible to apply classic population predation models to individuals, but strong assumptions limit the scope of the results. We propose here an approach that facilitates linking both micro and macro scales in predator-prey models. While the macro model is aspatial and continuous in time, the micro (individual-based) model is discrete in both space and time. The choice of spatial and temporal granularities requires an important consideration. The model reproduces phenomena evolving at two temporal scales: a rapid one focused on the interaction between individuals of each population (predation process), and a slower one for the growth and natural death processes, whose parameters remain calibrated at a population scale. Predation rate of prey and birth rate of predators are observed during the individual-based simulation according to local interaction rules: individual movement rules and the prey perception range of the predators (predation distance). This perception can vary with the introduction of shelters for the prey which allows exploring the effect of spatial heterogeneity on the dynamics of both populations. Both micro and macro models have been implemented on the NetLogo platform. Experimental simulations explore the influence of the perception range (from 15m to 45m) and of the density of shelters (from 0 to 70%) on the evolution of population sizes through time. The cyclic behaviours of the system are still observed. This work illustrates how we can link both spatial scales through methodological choices and gives an interesting path which can be explored in order to develop predation models focusing on testing the effects of spatial heterogeneity.