In the case of tracing metal in estuary using stable isotope composition I can’t help, because that was never my interest and I can’t recommend any good publication about this topic. In the case of phytoplankton, the method to separate it from water column is simply filtration, the procedure is i.e. https://www.researchgate.net/publication/225650442_Chlorophyll_a_extraction_from_freshwater_algae_-_A_reevaluation and after that, for example, they use acetone extract method. In the case of stable carbon and nitrogen isotope I know one paper where samples phytoplankton were investigated. The authors in the case of big shallow lakes took samples of setson and assume that in lakes with high phytoplankton biomass, the seston isotopic signal largely represents the isotopic signal of phytoplankton, whereas in lakes with low phytoplankton, particulate organic carbon originating from terrestrial or macrophyte dominates the signal it mainly correspond to phytoplankton but in the case of macrophyte dominated shallow and not big lakes (https://www.researchgate.net/publication/235347313_Bimodality_in_stable_isotope_composition_facilitates_the_tracing_of_carbon_transfer_from_macrophytes_to_higher_trophic_levels).