Fishful thinking: one paleontologist explains how an important ROM collection began with dinner.

Many people are familiar with the usual museum methods for collecting artifacts or specimens: excavate for dinosaur bones or ancient ruins, run expeditions to exotic or inhospitable places to find elusive or undescribed species. This is the stuff of National Geographic TV programs, even the pages of our own ROM magazine. But for more than 20 years now, I have been collecting museum specimens in a more commonplace locale--Toronto supermarkets.

In North America, our diet is quite restricted. We tend to eat only a few kinds of birds--chicken, turkey, quail--and mammals--cow, pig, lamb--and no amphibians or reptiles (how many of you have had frog legs or alligator more than once?). Even the fish in our supermarkets are not diverse: think salmon, cod, sole, maybe snapper and a few others, and most often they come as boneless fillets. But this is not the case in other cultures.

The Toronto area is home to some sizable ethnic fish markets serving their local communities: Kensington Market (primarily Caribbean), Markham (Chinese, Japanese, and other Asian cultures), and the Danforth and College Street West (the Mediterranean region). People of each culture are familiar with the fish from their home country, and wish to continue including them in their diet. The result: each market carries a diverse offering of whole fish species.

My own interest in fish began more than 25 years ago when I first started cataloguing a set of Pleistocene, or Ice-Age, fossils from Florida in the ROM's Paleobiology collections. Although most were mammal bones, there were some fish bones, too, and identifying them was particularly problematic. The department didn't have any comparative fish skeletons. So I consulted with the Museum's fish department, Ichthyology. They did have a few skeletons, primarily Ontario and Atlantic Coast commercial species, but very few from the Caribbean, the main source of fish found in Florida. But the ichthyologists gave me a good suggestion: if I needed Caribbean fish skeletons, why not just buy some fish from Kensington Market? With no budget at first, I began to purchase my dinner at the market, then save the skeletons for the ROM. I was amazed at the diversity of whole fish species for sale, and didn't know where to start. We needed them all. I was "hooked."

In 1988 I received a government grant to hire some summer students to prepare the skeletons for our comparative collection. The grant came with some additional money for fish purchases. Now I could go every week on a fishing expedition to Kensington Market, and on every visit I found a different species for sale. While Ichthyology staff assisted with identifications, all summer the team of students cooked, cleaned, and numbered fish bones.

Almost every summer since, we have hired one or more students or accepted volunteers to continue this work. This way, we have added hundreds of skeletons to the collection, all at minimal cost to the ROM.

Over time, word began to get out that the ROM had a sizable comparative fish bone collection. Archaeologists started visiting to study our skeleton collection (rather than our fossils!) to identify bones from archaeological middens, the garbage dumps of yore. Numerous times they came with puzzling bones only to find one of our fish skeletons was a perfect match. Several species of the jack family, in particular, often have strange swollen bones in their spine, ribs, and parts of their skull that are so odd they don't even look like fish bones. When they show up in middens or in fossil sites, they are usually set aside as unidentified whatzits. But now we had the solution to this puzzle. More than one archaeologist has left our collection sporting a wide grin. Sometimes (with my encouragement) these archaeologists have brought donations in thanks, even fish caught nearby their archaeological sites. Our collection continued to grow.

Then, a friend told me about the Asian superstores in north Toronto. I discovered a whole new set of fish species available there. At first I wondered whether we really needed these particular species, as we have essentially no Asian fossils in our collection. But when a colleague from the Tyrrell Museum of Palaeontology in Drumheller, Alberta, came to visit, he was overjoyed to see some of these Asian fish skeletons. The milkfish, for example, a popular food fish in the Philippines, represents a kind of primitive fish whose relatives are commonly found in the famous dinosaur fossil deposits of Alberta. But this kind of fish no longer occurs in North America, so he had never seen a skeleton before.

Several times now I have even found unusual fish species that--it turned out upon consulting ROM ichthyologists--were not even represented in the ROM's main fish collections (but are now). This just encouraged me in my odd hobby.

Although my collecting may sound a bit off-the-cuff, I carefully photograph, weigh, and take a standard set of measurements for each fish before cooking it. My kitchen ends up looking a bit like a science lab, much to the amusement of any visitors. I carefully consume only the meat, and save every bone. This is a bit of a labour of love, especially when eating more than one at a time, which requires organization so as not to mix them up.

I take the bones to the ROM, where I finish the processing, sometimes putting them in the ROM's infamous "bug room," where a colony of dermestid beetles cleans the bones by consuming every last morsel of flesh. Each skeleton, once clean and dry, receives a catalogue number (which is written on each bone that is large enough); this allows a researcher to take certain bones from two or more species and lay them side by side for identification purposes, but then to put back in the correct box for future reference. Since most bone or fossil identifications involve single bones, we store the dry comparative skeletons disarticulated, as separate bones, rather than together as a skeleton.

Why this fish diversity continues to be shipped to Toronto from afar is somewhat of a mystery to me, as sometimes the fish vendors don't even recognize all the species they are selling. So they are not the usual offerings, even for that region's cuisine. But the strength of a comparative collection is diversity, so these oddballs are exactly the kinds of species I am looking for, much to the amazement of the vendor--or maybe to their pleasure since they might not otherwise be able to sell them off!

To this day, I continue to cruise fish markets and still find fish species not represented in our comparative collection. With more than 500 fish families and more than 30,000 fish species known worldwide, we will never have them all. But our bone collection has grown to include more than 525 species from 126 families, not including those still in my freezer. Who knows what kind of fish I'll have to eat next?

Kevin Seymour is assistant curator in the Paleobiology section of the ROM's Department of Natural History.