In all fairness, classical conditioning (associating an external stimulus with either a reward or pain) is pretty basic - the model system for studying it is a sea slug. Operant conditoning (associating one's own behavior with reward/pain) is slightly more complex, but is still present even in invertebrates.

That said, testing higher-order intelligence in different species is often difficult, especially for species very different from us, so just because we haven't observed it doesn't mean it isn't there. Conversely, however, complex behaviors that seem outwardly intelligent can be produced from little more than a combination of conditioning, memory, instinct, and our own anthropomorphic biases.