Bird digestive systems deal with swallowed indigestibles by first extracting all the nutrition, calcium and other minerals the body needs, and then packaging and regurgitating the nonnutritious indigestibles. Their gut is too slender and delicate for sharp bits and bulky bundles to be passed out as droppings. A healthy fresh casting is usually neat, roundish, oval or sausage-shaped, glistens damply and smells of nothing particular, while a foul casting is often a sign of sickness.

Castings, also called pellets, are probably made by most birds sometimes. An Eastern Kingbird made castings after eating grasshoppers (Fig. 2). Crows and herons cast after eating small rodents (Fig 3). Kingfishers fed smelt and minnows produced delicate, lacy spindles (Fig 4). But most fish-bones and frog-bones are rapidly dissolved, so our ospreys made no castings, or at least not from the kind of fish we fed them (whatever we could get—herring, catfish, bass, perch, whitefish, live minnows and some unspecified supermarket species.)Mind you, ospreys eat by tugging and biting off small pieces. Loons, who eat their fish whole, produce gritty "whitewash" shoots in the water instead.

Most gulls will snap up whatever living or dead creatures they can find, and when fish were hard for us to get, we offered them mice. When we admitted a Herring-Gull that had a hook visible on X-ray in her gizzard, we deliberately gave her large amounts of fat mice. When she cast, the hook was probably ejected safely in the bulky fur; there was no trace of the hook in her final X-ray when she was given euthanasia for aspergillosis on the 14th day. (We have found gulls to be very prone to aspergillosis.) Was the hook rusting unseen in the tangle of vegetation of the aviary floor, or had she simply dissolved it?

This varies according to the size of the meal and is limited by the size of the mouth. A Snowy Owl in the wild produced the biggest pellet I have seen — 130mm x 36mm. A small male Great Horned Owl coughed up one 105mm long. But following a single small meal, a large raptor will only make a small casting. A Screech Owl made a casting 60mm x 15mm, a little Saw-whet Owl a casting of 43mm x 15mm, but a kestrel, about the same weight as a Saw-whet, makes a much smaller casting due to a narrower jaw. (Fig 5)

Castings can weigh quite a bit, too. A healthy 85-gram Saw-whet produced a casting of a compacted whole vole; it weighed 5 grams, about 6% of the owl's body weight. If that was produced by a 150-lb man, it would weigh about 4 kg!

The compressed clues include fur, feathers, bones, teeth, snakeskin, turtle scutes, crayfish legs and insect keratin. As there is often a wispy "tail" trailing at the bottom-end, the prey bits on the top end entered the stomach last and will therefore be least digested. Once when I popped pills into an owl's mouth, he cast soon after; the now damp, crumbling pills were tidily sitting on the blunter "head" end.

The first postadmission casting is a prized dietary record. Each newly admitted bird's unit was closely watched to find it before it got trampled on, and then it was teased apart to learn its secrets. Those bits we could not identify were bagged and kept for expert assistance.

One first casting identified an osseous jumble I had puzzled over on X-Ray. The abdomen of this large female Great Horned had been startlingly tented outward with long bones. Next morning she obligingly flung out a 75gram casting so big that it broke into several parts, revealing recycled rabbit, including whole legbones 75mm in length! (Fig 1) Quite a sword swallowing trick.

Other prey parts we have found in raptor castings: rat parts (think warfarin here!) rabbit and vole skulls, shrew and Black Squirrel jaws, legs, recognizable starling feathers, a Redwing Blackbird epaulet, corn from pigeon crops, snake plates, a small gartersnake, crayfish parts cooked red by the stomach acid cast by a Screech Owl, grasshopper and cricket legs, part of a Luna moth, beetle elytra and even scavenger beetles, suggesting the bird had been investigating carrion at the time. Once there were no less than ten mouse skulls in one casting.

A study in the U.S. found lead pellets in 70% of Bald Eagle castings, suggesting that the eagles were feeding on animals that had been shot.

Following the digestion of a meal or two (usually once a day or more, depending on the bird's metabolism) the packet begins to move up the esophagus. To complete its egress, a raptor briefly `retches' or yawns widely, leans over and drops or shakes it out (Fig 7). This explains why a thin, hungry raptor who has been eating voraciously may one morning appear dismayingly indifferent to a prey offering. But the bird knows: before the food can go down, the casting must come up. The stimulus of a proffered meal quickens the reverse motility of the stomach and esophagus, hurrying the casting up and out of the way.

Perhaps the casting has to be actually on its way up before food-taking is begun. In an unusual accident, a Red-tailed Hawk was netted for release shortly after eating a full meal, and aspirated and died. His crop was very full still, but the post-mortem revealed a large casting in the stomach. Either those two items were going to pass each other like a pair of elevators, or they were going to be amalgamated in the stomach for a later giant casting of two separate meals.

A similar puzzle was offered by a starved female Great Horned in an indoor unit, given mouse as usual right from the start. On her third day, she polished off a whole adult white lab rat. On her fourth day she was given 250g of dark lab mice, and on the morning of her fifth day she produced a handsome and healthy casting of nothing but white rat, with the skull and long bones unbroken. How had her system bypassed the intervening meal of dark mice? Had she absorbed their smaller bones entirely? What happened to the dark fur?

Castings have appeared from 1-17 days. The longer times are apparently related to thorough absorption of the bones and teeth for extra calcium. We fed all our raptors dead mice and rats; a nestling Great Horned Owl did not cast till the fifth day, and then it was a mere dusty pressing of fur. Though he had eaten hugely of mice every day, he had been severely deprived of calcium and phosphorus in previous captivity: his body absorbed every fragment of tooth and bone.

Another reason for delay is injury, particularly to the abdomen, which slows or stops digestion. On post-mortem I have occasionally found evidence of ruptured stomachs, some of which had then healed again. (See article on Hooked Loons.) The following owl was probably a survivor of such a blow.

An adult male Great Horned was brought to our clinic, thin, gaping slightly, weak, and with a protrusion below his chin as if he was about to hoot. But the lump was very hard, and the owl was having difficulty breathing. Quickly put on his back on the table, a dark obstruction could be seen jammed deep at the back of his throat. It was so hardpacked that I only drag out small tufts with forceps, so I stuck my gloved fingers right in and levered out the hardest, most massive casting ever, so foul that my retching stomach defeated my scientific attempts to weigh and measure the thing. Jammed in, and hard-packed, the broad, round shape was too wide to be passed without help, and was killing him.

Six weeks later, weathering had hardly reduced the odour of its decomposition, but by standing upwind I managed to examine the extraction. It was about 80mm x 80mm x 40mm and consisted of at least four fat microtus mice (voles) compacted, entire, and undigested!

It was two weeks before the owl could eat and cast properly, so I postulated a severe abdominal blow, rupturing the stomach, preventing digestion — perhaps even preventing descent into the stomach — and causing a long delay during which the rotting mass compacted. He recovered very slowly, and was banded and released.

Sometimes owls are found with porcupine quills. This story shows one reason.

A very healthy northern Great Gray Owl was struck by a vehicle and flung dead on a snowbank at the side of a road. Blood had run down from the haemorrhaging brain through nostrils and had flowed over beak, a common occurrence with such violent collisions. But I was surprised to find some porcupine quills lightly stuck in the skin just behind each hallux (the back talon) as well as a few on the back of the tarsus (lower leg) of one foot; they were undamaged and white, suggesting a light pass, of the underbelly. None were found in feet or anywhere else in body. On post-mortem I discovered he was indeed well-fed that day: the stomach was a hugely stretched bag with a very large mole at the bottom, a mouse in the middle, and an undigested shrew on the top!

This bit of reportage was siezed upon and reprinted by the late Peter Whelan, birding columnist of the Toronto Globe & Mail who called me, as he occasionally was wont to do late at night when searching for unusual bird tidbits. Naturalist Jim Duncan responded: the Great Gray had not been attacking the porcupine, he said, but targeting a shrew scavenging the road-kill; the owl was swooping with legs outstretched to snatch the shrew without stopping in flight. He was exactly right, as proven by the stomach contents.

Note, 2006. I recently had the pleasure of looking after a grandson's tree-frogs, and was surprised to find that they too make castings of hard bits—in this case, leftover exoskeletons of the crickets and mealworns on which they were being fed.