It depends. If the rocks have to stay in the field (or stay intact), you should use a kind of silicone, (used by dental technicans, it's expensive (but cheaper than preparation equipment) and you can choose different properties). If the rocks are exposed at the sea, you should use a fast setting silicone.

If you can waste the rocks, you should use an epoxy for cast and cover and dissolve the limestone with hydrochloric acid, formic acid or acetic acid (but take care that you deal with limestone, the rocks on your picture look like siderite or similar carbonates are involved and this will only hardly be dissolved by the acids mentioned above)

Sorry I didn't notice this question when it first appeared, but here is a late answer:

For specimens that have fine detail, my experience is that the best fidelity comes from using liquid latex, which is readily available from art supply stores, or from many online sources.  Simply brush on a coat of the thick white  liquid, and allow it to dry to an amber color. Repeat this process to get at least three coats of latex,  five or six would be even better. The resulting latex mold will capture fine detail. To capture the overall 3-d shape of the specimen surface,  I usually make a backing layer composed of Plaster of Paris, commonly layering one several layers of cotton gauze ("cheesecloth") for small specimens, or burlap for big specimens (like dinosaur tracks).  The plaster "lacket" won't stick to the latex layer, so at the end, you can lift off the plaster, then peel off the latex layer.

For quick molds  in the field, and  especially for large specimens, I commonly use silicone caulking compound, available at local hardware stores. Here in the USA, I look for Dupont Silicone caulking compound, which must not be confused with the  Dupont  Silicone II formulation. The difference is that the original silicone caulk will not stick to a damp surface. You can moisten the specimen, then spread on a generous layer of silicone. The best way is to use a caulking gun to squirt out  enough silicone to cover the surface, and then use a wet sponge to pat the silicone to form an even layer. Use a thickness of about 1 cm for small specimens, maybe 2 cm for large specimens. The outer surface of the silicone will solidify  quickly, but it is important that the full thickness of the silicone has solidified before you peel the mold off the specimen. On a hot sunny day, I have sometimes been able to pull the mold after a few hours, but I prefer to wait 24 hours if that is feasible.  The clear and opaque versions of the silicone caulk work equally well.  But never use the Silicone II formulation, because that product does stick to damp surfaces so the mold ends up being  permanently cemented to the specimen.  Many silicone  and polyurethane preparations are sold for molding purposes, but these commonly require the addition of a catalyst. These have the drawback of requiring some kind of parting  liquid  (oil, wax, polyvinyl alcohol, etc.), and they are relatively expensive.  In contrast, neither the liquid latex or the Dupont Silicone caulk require a parting agent, the materials are inexpensive and readily available, and they yield very high quality molds. The molds are durable, allowing many replicate plaster  or resin casts to be made from each mold.