Is there a difference between an abstraction and a property?
Yes.
An abstraction (SEP) is primarily an ontology, which means it is a web of belief constructed around ontological commitments with mereological phraseology. On the other hand, a property is primarily a characterization of an entity which might be an abstraction or might be concrete. For instance, the category 'fish' abstracts across multiples fish (organisms) and fishes (species). To a physicalist, there is no existence of a 'fish' in the abstract, because the term properly understood (under specific metaontological theories) is grounded in nominalist fictions. In fact, each 'species of fish' is also an abstraction so that 'trout' abstracts across multiple organisms that qualify as trout given a class of particulars. Thus, any abstraction, like the OSI model, as another example, is a description of entities and relationships among them. There is no physical OSI model, but rather a description of ontological criteria that allows a person to accept or reject membership of some given physical system of computation (SEP).
A property (SEP), on the other hand, is a characterization of an entity which may be either concrete or abstract. Consider a specific trout caught on a specific date and time in a specific place. According to prototype theory it will have a collection of properties that lend to it conforming to cognitive prototypes of a conceptual category 'trout'; it will likely have the property of 'being wet', 'having a tail of x cms', 'possessing a specific genome that conforms to "trout"', etc. The concrete properties are likely to be described in terms of language enriched with qualia (SEP), such as 'shiny', 'heavy', 'slippery', etc., but the abstract properties are themselves conceptual categories built from other concrete and abstract properties which might be subject to graded membership and concomitant sorites paradox.
So, abstractions and properties are not strictly hierarchical, but are two different classificatory axes. Some abstractions are ontologies with entities and properties, and some entities and properties are abstractions. Thus, there is the dichotomy of abstract and concrete, and there is the organization of entities with properties some of which are relationships, and how they relate is ontologically domain-specific. The first axis is language for evaluating the physical existence, and the second is for organizing language around identity of phenomena for use with reference.