The Fisheries Blog explains (boldface mine):
Species re-assessment. You can’t know how to manage your species if you don’t know what species it is in the first place. The discipline of ichthyology is rapidly evolving, and we are continually discovering new fish species. Oftentimes, we can’t solve a problem without collecting morphology or genetic data from the original type specimens that are stored in natural history museums. Other times, we can go back and look at species that we thought were Species A, when new information suggests they are actually Species B. This can only be done with the help of scientific collections.
Assessing invasive species. We often see the damage caused by invasive fishes in because we have pre-invasion data from the recipient ecosystem. However, we don’t know the effects of many fish invasions because they occurred before the era of quantitative fisheries science. With museum collections, we can often reconstruct historical species distributions and understand occurrence patterns before and after invasions. We can also run stable isotope or fatty acid analyses on museum specimens to determine how their diets have changed in response to the invader.
Changes in contaminants through time. Museums also give us a nice time series of historical environmental conditions. For example, studies have examined tissue of museum specimens to understand how mercury contamination in both marine and freshwater fishes has changed over time. We can relate these changes back to significant environmental disturbances or regulations to evaluate their effects.
Evolutionary response to human threats. Fishes can adapt their morphology and life history quite quickly to deal with human pressures such as overfishing. Although data exists for some well-documented fishes, most species can only be assessed by re-measuring museum specimens.
It’s worth noting that these are very applied reasons. If you’re keen on translational science, then collections are critical.