Free Octopus by Roland C. Anderson

charted the removal of light pieces of crab shell by waves and currents for the common octopuses. Ambrose (1986) made artificial middens near his two-spot octopuses’ dens, and watched hermit crabs take away the snail shells to use as homes. So visiting an octopus den once a week to assess intake by counting hard-part remains offers an incomplete picture of actual meals. The only way to know consumption from sampling leftovers is to know about the relationship between what the octopuses ate and what you found at their home within a couple of days.
    Another way to find out what an animal has eaten is to kill it, open the stomach, and sample the contents of the gut; you can also examine the stomachs of already dead museum specimens. But this approach doesn’t yield an accurate description of prey choices, because octopuses only eat the soft parts of prey: they scrape flesh away from the shell or skeleton, and they digest some of it before they eat it. What the researcher has to work with is often semidigested lumps of tissue. Biochemical serological analysis is needed to get a clear idea of what the food species is and where there was a different array of species than the midden samples of other octopuses. And you’ve killed the octopuses to find the answer to a behavioral question, which is a drastic solution, and you have only one sample.
    The underlying problem with the question of prey choice is that octopuses eat a lot of different prey species. Ambrose (1984) found that his two-spot octopuses from Bird Rock ate fifty-five prey species over several years. We found twenty-eight prey species for twelve common octopuses in one small bay in Bermuda over four weeks, and we have collected seventy-five prey species so far from the common octopuses of Bonaire. In addition, what an octopus of one species takes as prey varies across the range of that octopus. Giant Pacific octopuses, for example, range from California to Alas ka in North America and over to Japan. In Japan, they eat fish, shrimp, and crabs; on Vancouver Island, they specialize in crabs, cockles, and Pacific littleneck clams (Protothaca staminea); in Alaska, they eat crabs 75 percent of the time. In each of these cases, there were remains of over twenty-five prey species in middens of a small group of octopuses. Clearly, octopuses like variety, and also sample a different variety in different areas.
    Octopus Garbage Heaps
    I conducted a study (1991a) of what common octopuses ate in Bermuda and what I found outside the dens a couple of days later. A team of volunteers followed two octopuses constantly from dawn to dusk for about ten days. Each day, we recorded where the animals went and what they ate outside their dens when they stopped to snack. We picked up the shell remains, and recorded what they took back to their dens. We then recorded the daily disappearance of prey remains from the den midden caused by sediment sifting over them or waves washing them away. After ten days, I dug through the midden to see what had stayed and what was buried. I remember coming by one den and seeing the eight bright blue shells from a chiton meal (a chiton has eight shell valves), then watching on subsequent days as the shells slipped down the midden slope and gradually got covered by debris.
    From these observations, I made a flowchart of food fate, from capture to visible prey remains at the den. About one-third of prey was eaten in places other than the den, but there was no size selection. Both big and small prey were carried home, depending mostly on how far away the octopus was when it found the prey. Size mattered after pieces hit the midden, since small, light crab shell pieces were more likely to be washed away than heavy zebra mussel (
Arca zebra
) shells. Ultimately, I could backtrack from ten shells of six prey species in the midden visits by a researcher every five days to how many of each species were actually eaten. So I know how much food was probably