Episodic-like memory and divergent brain systems in mammals and birds

Abstract

In their review in PNAS, Allen and Fortin use a comparative neuroanatomical approach to gain insight into the evolution of episodic memory (1). The authors suggest that, as in humans, episodic-like memory in nonhuman mammals depends on a system consisting of the hippocampus, parahippocampal region, and prefrontal cortex (PFC), and, in birds, a similar system consisting of the hippocampus, area parahippocampalis, and nidopallium caudolaterale (NCL), the avian analog of the PFC. Although we agree that the behavioral evidence for episodic-like memory in birds is compelling, in contrast to Allen and Fortin, our interpretation of the available neuroanatomical evidence indicates that key components of this system are missing in birds. In mammals, the PFC is thought to assess the relevance of episodic memories recalled from the hippocampus via its connections with the hippocampus, parahippocampal region, and other higher association areas reactivated during recall and to, then, orchestrate behavior accordingly through projections to motor areas (1). Like the PFC, the NCL is reciprocally connected to higher association and motor output regions and appears to perform executive functions (2). However, of the laboratories (at least four) that have addressed this question, none has detected connections between the NCL and the area parahippocampalis or hippocampus (3) (reviewed in ref. 4). This may be related to the fact that the PFC and NCL are analogous structures that evolved independently and develop from different parts of the pallium; in mammals, the PFC develops from the dorsal pallium, whereas in birds the NCL and the rest of the dorsal ventricular ridge (DVR) develop from the lateral and ventral pallium (5). Interestingly, in birds, rather than developing into a higher-order multimodal association area, such as the PFC, the dorsal pallium develops into the hyperpallium, a structure composed of primary visual and somatosensory/motor subregions (5). Nonetheless, consistent with its developmental homology with mammalian dorsal pallial derivatives, the hyperpallium shares PFC-like connections to the hippocampus, area parahippocampalis, and motor output regions (5) and, as suggested by Bingman and coworkers, may bridge the apparent gap between the hippocampus and NCL by conveying spatial (visual and olfactory) information via known projections to the NCL (3). The NCL could then integrate this information with that received from other association regions in the DVR and, thereby, orchestrate adaptive behaviors. Although this pathway may provide a route for integrating information processed by the hippocampus, hyperpallium, and DVR, additional differences exist between birds and mammals that likely influence the nature of episodic-like memory in the two groups. Unlike mammals, wherein highly processed information from virtually the entire neocortex converges in the hippocampus, enabling it to serve as a node for rapidly encoding experiences and initiating their recall, the avian hippocampus only receives olfactory and visual input and is not privy to much of the information processed by association regions in the DVR (nidopallium and mesopallium), in addition to the NCL (4). Given these fundamental differences in neuroanatomy, to the extent that episodic-like memory depends on information reaching the hippocampus, such memories may be qualitatively different between birds and mammals.