Living systems are not just static structures. We ascribe 'life' to them because they're dynamic. We can find another sort of circularity in their internal operations. These internal operations are 'circularly' interconnected in the same sense that the components are. There is something about the identity and unity of a living system which is maintained by these internal operations -- something which can be influenced by events in the living system's environment, but which is specific to the living system itself. You can move the living system to another environment, but (so long as it can successfully survive) this circularly-interconnected network of internal operations will persist. These operations evidences no intrinsic 'purpose' beyond maintenance of the living system's constitutional and configurational integrity.
The course of actions ('responses') observed for a given living system exhibits a sort of circularity in the sense it is (at least partially) repetitive. The exact trajectory of these courses of action is mediated 'internally' by the organism's capacities for action. In other words, what the organism will do (and remain living) will be circumscribed by the range of things the organism can do. Because these capacities are in turn qualified by the living system's circularities of form, configuration and internal operations, similar circumstances will result in similar actions.
Correspondingly, the course of situational transitions affecting the organism ('stimuli') is mediated 'externally' by those potentials the world affords. Even though it is the organism's own configuration which determines its capacities for action (and hence its specific actions), the 'environment' influences the overall course or trajectory of the situations encountered, and hence the series of resulting actions. As such, there is a 'circularity' in the reciprocal interplay between the living system and its 'environment'.
The 'circle' of this interplay cannot be reasonably said to have a starting point (except the point at which the living system originates). It cannot be said to have an ending point (except the point at which the living system ceases to be living). As such, we cannot predict the living system's course of activities based on 'first' or 'last' causes.
Because of this, the course or trajectory of reciprocal engagement between a living system and its observed 'environment' is not reducible to exclusive determination by one or the other.