In 1987, Squier proposed a taxonomy of memory for the first time and upgrade it in 2004. furthermore in 2010, Michaelian modified it and proposed a new taxonomy for it. But Spatial memory is not considered in both of them.
I think that from the point of view expressed by Joachim "spatial" could be considered in terms of the type of content of the stored information (similar to verbal, visual and so on). However, there are some specific features of spatial memories that makes it quite special (e.g. phylogenetic origin and relevance, is it declarative?, or non declarative?, global map-based vs. landmarks-based strategies,...).
Anyway, from my point of view it should be considered as a separate construct when assessing subjects in clinical neuropsychology. A spatial learning proxy can be measured with tasks such as the Rulit-Light Trail Learning Test, or more properly with more complex experimental navigation virtual reality-based measures such as the task used by Fried et al. (2012) see the link.
thanks for your answer. that's right. I think, combination of 2 or 3 categories such as declarative and procedural groups may be needed for categorize this kind of memory.
of course if we see the view of Eichenbaum (2010), it would be related to the cortical-hippocampal circuit.
H. Eichenbaum, "Memory systems," Wiley Interdisciplinary Reviews: Cognitive Science, vol. 1, pp. 478-490, 2010
Conscious retrieval of any location could be related to the episodic memory associated to that experience. I would understand it like a type of memory directly dependent of the medial temporal lobe (cortical-hippocampal circuit, as you suggested above).
Of course, Eichenbaum papers can help to address this question.
Buszaki gave a while ago a talk arguing nicely that episodic memory evolved from spatial memory, i.e. what is where. Our animal studies also suggest that spatial memory is a precursor of episodic memory. Many species know where smth is or how to come from A to B and the region sufficient for it is the hippocampus. Humans, some bird species and also some other mammalian species have extended this spatial memory to also encode the temporal domain - fulfilling Tulvings definition of what-where-when or episodic memory.
Spatial memory cannot be strictly assigned to one of the classification subsystems; it is, indeed, part of several of these categories, since it involves aspects of non-declarative memory (procedural), declarative (semantic and episodic memories), as well as of both short and long-term memory. please see blow paper.
"Spatial memory: Theoretical basis and comparative review on experimental methods in rodents.-2009"