Yes, in your datasheet you have mentioned polyclonal antibody.
Polyclonal antibodies (pAbs) are antibodies that are secreted by different B cell lineages within the body, oftenly from serum (whereas monoclonal antibodies come from a single cell lineage and having same epitope). They are a collection of immunoglobulin molecules that react against a specific antigen, each identifying a different epitope. Due to presence of different epitopes they can react with another antigen too like mouse and rat VEGF in your case.
The term 'cross reactivity' in your antibody datasheets essentially means that your anti-VEGF antibody, though it was raised against Human recombinant VEGF165, still could show binding interactions with similar antigenic sites on different proteins (of course has to be 'VEGF like', having the more or less similar epitope configurations), and these could happen among different species (in this case it is human, mouse and rat).
That means you can also use this particular antibody to probe against VEGF proteins in all the three species and can expect similar outcome.
Cross reactivity goes further than just inter-species binding of the same target protein.
All proteins, including Abs, have some degree of affinity for eachother, some weak and some very strong - the affinity of biotin with avidin is the highest known.
If you do a Western blot, for example, and you take a series of exposures at varying lengths - maybe 5 sec, 5 min and overnight, you will see more bands show up due to your antibody binding weakly to other proteins on the membrane. Of course some antibodies are very good to use because they have minimal cross reactivity but even many monoclonal Abs can cross react with undesired proteins. John R Underwood has published several papers on this topic.
Another concept you could read about is "affinity maturation".
In a simple term it means that the antibody will recognise your specific protein in those particular species.
For example if antibody X is raised in rabbit and it has cross reactivity for mouse, rat and human but not monkey, then that means that if you used this antibody for staining in mouse, rat and human tissue you'll be able to get labelling for your protein but not tissue from monkey.