1. Long-term datasets are needed to evaluate temporal patterns in wildlife disease burdens, but historical data on parasite abundance are extremely rare. For more than a century, natural history collections have been accumulating fluid-preserved specimens, which should contain the parasites infecting the host at the time of its preservation. However, before this unique data source can be exploited, we must identify the artefacts that are introduced by the preservation process. Here, we experimentally address whether the preservation process alters the degree to which metazoan parasites are detectable in fluid-preserved fish specimens when using visual parasite detection techniques.
2. We randomly assigned fish of three species (Gadus chalcogrammus, Thaleichthys pacificus, Parophrys vetulus) to two treatments. In the first treatment, fish were preserved according to the standard procedures used in ichthyological collections. Immediately after the fluid-preservation process was complete, we performed parasitological dissection on those specimens. The second treatment was a control, in which fish were dissected without being subjected to the fluid-preservation process. We compared parasite abundance between the two treatments.
3. Across 298 fish individuals and 59 host–parasite pairs, we found few differences between treatments, with 24 of 27 host–parasite pairs equally abundant between the two treatments. Of these, one pair was significantly more abundant in the preservation treatment than in the control group, and two pairs were significantly less abundant in the preservation treatment than in the control group.
4. Our data suggest that the fluid-preservation process does not have a substantial effect on the detectability of metazoan parasites. This study addresses only the effects of the fixation and preservation process; long-term experiments are needed to address whether parasite detectability remains unchanged in the months, years, and decades of storage following preservation. If so, ecologists will be able to reconstruct novel, long-term datasets on parasite diversity and abundance over the past century or more using fluid-preserved specimens from natural history collections.

Study species

We used three species of marine fish for our experiment. Two species, Walleye Pollock Gadus chalcogrammus and Eulachon Thaleichthys pacificus, were provided to us by the University of Washington Burke Museum Ichthyology Collection (UWFC). Both fish species were collected by National Oceanic and Atmospheric Administration research cruises in Alaska, frozen, and shipped to UWFC, where they were stored frozen as they awaited cataloging. The Walleye Pollock were collected from the Bering Sea and Gulf of Alaska between 2000 and 2002 and the Eulachon were collected from Shelikof Strait in the Gulf of Alaska in March of 2002. On 11 and 12 May 2018, we collected English sole Parophrys vetulus by otter trawl in Port Madison, WA, in conjunction with the Fisheries Ecology course offered by the School of Aquatic and Fisheries Sciences at the University of Washington. English sole were euthanized, placed on ice, and frozen within 6 hours of collection. These three species of fish represent a variety of trophic strategies (suspension feeder, benthic predator, and pelagic predator, respectively), life history (i.e., short lived [Eulachon] and long lived [English Sole and Walleye Pollock]), habitats (benthic [English Sole] and pelagic [Walleye Pollock and Eulachon]), as well as body plans (i.e., flat fish [English Sole] and fusiform [Walleye Pollock and Eulachon]). Fish traits might influence the parasites that the fish can be infected with or the effects of preservative on fish tissues, and thus having a diversity of fish traits allows more robust conclusions regarding the effects of preservation on parasite detectability. In total, we collected 99 English Sole, 109 Walleye Pollock, and 70 Eulachon.

Experimental test of the effect of preservation on parasite detectability

We took an experimental approach to assess whether and how the preservation process affects the detectability of parasites. For each of the three fish species, individuals were randomly assigned to one of two treatments using a stratified design. Prior to randomly assigning fish to a treatment, fish were visually paired according to length within each species. Then, an individual from each pair was randomly assigned to the preservation treatment by a coin toss. This stratified random design equalized host size, a potential driver of parasite abundance, between treatments, and ensured that the mean, median, and range of host body sizes were similar between the two treatments.

Two treatments were included in this experiment. In the control group, frozen fish were thawed and dissected, and their parasites identified. In the treatment group, frozen fish were thawed and preserved according to methods used at the University of Washington Fish Collection (UWFC). We followed the preservation protocol used by the UWFC, by first placing completely thawed fish in 10% buffered formalin solution until the fish is fixed (absolute amount of time is variable due to variation in fish body size, range = 9–16 days). A fish was considered to be fixed when its tissues had become firm (particularly around the abdominal cavity) but retained some pliability. Over-fixation and decalcification are possible if fish are left in formalin beyond the time required to fix all the tissues. To prevent this from happening, fish were monitored during the preservation process at least every three days. Once the fish were fully fixed, they underwent two consecutive freshwater rinses, each lasting 24 hours. After this, the fish were placed directly in 70% ethanol, where they remained for at least three days prior to dissection. Across the entire preservation process, care was taken to ignore parasites that could be dislodged during the preservation process, to mimic actual preservation conditions in collections. After preservation was complete, fish were dissected, and their parasites identified.

Parasitological dissection

To avoid confounding the effects of preservation and dissection method, we used a consistent dissection protocol across both treatments. This could result in lower overall detectability (across both treatments) for parasites found in muscle, since candling was used rather than a muscle squash to avoid excessive destruction of the musculature of specimens – a variation in dissection protocol that is often used for museum specimens to preserve external morphology. Methods for all other tissues follow standard parasitological techniques (see below). We compared parasite abundance between the two treatments to assess the effects of preservation on detectability for each parasite taxon detected.

For each specimen, we made a ventral incision from the anus to the gill isthmus. We then removed all viscera for examination. The viscera were separated by organ and individually squashed between two glass plates and examined for parasites. Gills from the right or blind side of the fish were removed, placed in a vial with artificial seawater (for control group) or 70% ethanol (for preservation treatment), shaken to free parasites from the gills, and examined under a dissecting microscope. External surfaces and the buccal cavity of the fish were also examined for ectoparasites. The specimen was then spread open along its ventral incision and placed over a strong light to examine the flesh and skin for parasites. All examinations were conducted under a stereomicroscope to best capture the entire parasite burden and not simply the large-bodied parasite taxa and all parasites were identified to the lowest possible taxonomic level (hereafter, morphotypes) using standard parasite identification keys.

This deposition includes several datasets: one raw dataset on the abundance of various parasite taxa in Walleye Pollock (Gadus chalcogrammus; PollockParasites.csv), one raw dataset on the abundance of various parasite taxa in Eulachon (Thaleichthys pacificus; EulachonParasites.csv), one raw dataset on the abundance of various parasite taxa in English sole (Parophrys vetulus; EngSoleParasites.csv), and a dataset containing meta-data collected on each host fish individual sampled (HostData.csv).

Data are available in this repository, and are also permanently archived in GitHub: https://github.com/wood-lab/Fiorenza_et_al_2020_Ecol_Evol