I tend to be very generous when it comes to labeling diets. Animals are not perfect boxes to never spill out of their strict defintions, nor are their diets, produced as they are from a variety of different sources. You’ll notice that I’ve been talking about what exactly defines some particular diets, and what they involve. These extend from my interests in determine if, in fact, it is possible to determine if an animal is an ovophage/ovivore/egg-eater. What mechanisms exactly of the jaw, animal, environment go into defining a given diet? Are they all the same for different animals? Can we use one set of parameters to then determine the result for any given animal, or are some diets just that much more special than others?
Of course, the first step is that first step: Determining what your definitions are. So this post will have four parts that have to be addressed when it comes to dietary definitions. Since I’ve been talking about piscivory, I’ll couch the discussion in terms of piscivory for ease of reference (it also makes the puns work). It turns out there are four basic things we must know, and they must mostly agree with one another: Habitat, Ecology, Association, and Morphology. These terms are using their basic definitions.
Where There’s Water, There’s a Way – Habitat
All piscivores require access to water. Their food lives in it, so this is a given. And how animals interact with this medium differs from group to group (or animal to animal). Areas of plentiful water imply areas with plentiful aquatic prey; and that means aquatic predators. Habitat and mode of capture of prey, involving mobility, have a strong relationship to morphology (those will be discussed below).
Large expanses of water require extensive mobility, either flight or speed, while deep waters require mechanisms to hide from predators or avoid detection from prey, or both. Competition in large bodies of water can be avoided simply by exclusive range of predators: isolated feeding grounds allows multiple different animals to exploit a similar resource. Smaller bodies of water require less mobility, but because of the small size of the feeding area, the diversity of foods may be reduced, preventing diversity of predators. Roving predators, or animals suited to moving over large areas, would tend to dominate their range over other competitors of the same food stuffs.
Competition can also be reduced by specialization. Larger bodies tend to have greater mobility in the water column, which allows a vertical separation of prey and thus predators. Smaller bodies, on the other hand, seem to have more stark vertical separation: upper levels are richly populated and diverse, whereas lower levels tend to be anoxic and depleted of plant life, thus hostile to typical prey species at the lower end of the web. Nectonic feeders (those specializing in a form of suspension feeding on detritus and bottom-lying foods that have fallen there from above) and benthic feeders (those specializing in the organisms that dwell at or within the bottom sediments) should be more common in lakes (but not swift rivers, which tend to have swiftly moving lower reaches) but also occur in near-shore lagoons, bayous and other broad reaches of slow-moving water, which is also water in which anoxic zones are larger. Catfishes are a good example of both.
There’s Always a Bigger Fish – Ecology
Aquatic trophic systems among fish tend to involve a series of ever larger consumers involving a variety of reptiles, turtles, fish proper. The bottom predator tends to be nectonic or benthic, feeding on the lowest and smallest foods: detritus, organisms in the muck of the bottom, food suspended in the water column, etc. But once a fish reaches a certain body size, only larger predators can reasonably eat it (clearly, though piranhas are not “reasonable,” though normally they consumed food smaller or of a size with them).
Archpredators[n1] in the aquatic environment will always be sparser and rarer, as in a terrestrial system. If recovery of large predators appears plentiful, either it is being oversampling, or there is a larger fish. (Which “fish” sometimes being non-piscine, such as leopard seals or orca in the arctic waters.) They are not always the largest animal in the ecosystem, and many “largest” aquatic predators feed at the bottom of the trophic web: whale and basking sharks, many rorquals, right and grey whales, are all suspension or benthic feeders, occasionally eating larger actual fish, such as humpback whales consuming schools of herring.
Of course, the alternate is also true: A clearly carnivorous animal without likely prey species implies the presence of prey species. You’re just not looking hard enough if you cannot find them. Even consumers in an austere, harsh nectonic environment have predators themselves. In the deeps of the ocean, there are only a few “top” predators, and they are usually the largest: giant squids and giant whales (there are also some abyssal-dwelling sharks, but they aren’t the largest of sharks, yet get fairly large).
“Is that a Fish in Your Pants, or …?” – Association
If you have a fish in your stomach, there’s a good bet you ate it. However you did is irrelevant. Some animals try to swallow things that aren’t good for them, and usually suffer as a result. But swallowing things foreign to you, as many generalist feeders might do, and making something useful of the venture, goes a fair way towards survival.
Predation is about access, and means, but it’s also about capability. Some pythons and anacondas can eat large, fully grown crocodilians, and we know this because we’ve seen them do it. Well, we’ve seen some try. Choking to death on a piece of food, like say a fish, or a salamander, or even a piece of plant, have been found in the fossil record. Additionally, remains in the body cavity, either in the general area of the stomach (forward in the cavity) or intestines (rearward) indicate successful consumption.
Association of food is of moderate value when it comes to inferring diet: We know that that animal ate that thing, and that’s mostly it. What we do not know, with a fossil, is how frequently that animal ate that type of food, or how frequently its species did so. The more fossils with similar associations, the greater the strength of inference, but this inference value decreases the more fossils are found that lack such association. This lack of association is mostly meaningless, but it should be borne in mind that some types of food cannot necessarily be recovered in fossils, especially when the food item has no soft parts. Animal prey without soft parts can including anything from sea cucumbers, jellyfish, and so on. Some soft-bodied animals, such as a variety of “worms,” have hard parts in the form of phosphatized or calcified jaw parts, as also found in squid and octopus, belemnites and ammonites, and nautilus. What of plentiful clams, whose innards may be removed without consumption of the shell? Or shells that don’t preserve well, that we might find them but never know they were there?
Association is on the “easier” end of analysis when it comes to fossil inference of diet. But it’s not easy. But another component of association is behavior. When two organisms are found to associate in a particular way (say, a tiger stalking some unfortunate future meal) then they leave behind evidence of this association. Even without direct association (nesting of one animal within another, as in the famous Xiphactinus/Gillicus “fish within a fish” fossil FHSM VP-333) the general association of one type of predator with one type of potential food increases the likelihood that the one consumed the other on a general basis. Fossils of bites either healed over or not can also indicate a preference for mode of capture, thus whether there is an incidental process of consumption, or a regular and thus learned behavior, which strongly implies preference.
Before I continue, another form of association includes the presence of various “-ites”: coprolites are removed from the body, but contain bits of partially digested foods, and their size may indicate to what they belonged; regurgialtites leave the body in the opposite direction as coprolites, and usually involve passing of undigestibles, which owls and snakes are well known to do; cololites are pre-coprolites that remain within the body on discovery, but are found as tubular constructs as inferred by their position inside the intestines; in contrast, gastrolites (not to be confused with gastroliths) are often misshapen masses, and may be associated with stones, and are the contents of the crop or stomach. The latter two are always associated with the body in question, but the former two are often not.
Bromalites (the remains of food formed into masses) is a sort of trace fossil, and as such represents a subset of ichnites which are evidence of association. Ichnology also involves fossil trackways, and while fish do not typically leave locomotory tracks (but can), ichnites in general may also permit some degree of dietary inference, such as the infamous Glen Rose River tracksites at Paluxy, Texas, which suggest a large sauropod being pursued by a large carnosaur. Ultimately, “Association” is “Behavior,” but in fossils, only the traces can be determined, and there’s some degree of uncertainty.
It’s Not Just Teeth – Morphology
Morphology may be separated into two groups: acquisition and processing. Food acquisition is the process in which food items are obtained. Processing is as the name on the tin suggests, the rendering or conversion of the item from “acquired” into “eaten.” For carnivorous animals, secondary acquisition methods generally involve the limbs, but in the absence of these there are also anglerfish “lures,” the use of the tongue as a lure in some turtles, etc. But most predators use the limbs, which include claws. In water, limbs are primarily adapted for swimming, so the method of acquisition is generally always the mouth. In this way features of the jaws useful in inferring piscivory have made the bulk of literature regarding the dietary specialization. Because of this, I will spend some time on the topic.
First, method of capture by the jaw. There are two methods: prehension and engulfing. The first of these is the act of inhibiting the prey without consuming it. This is useful when the prey is fairly obnoxious to deal with, or large. And you do this with either teeth or with a pointed element of the jaw, such as a beak. These can also be used for processing, but I will get to that in a bit. Piercing teeth are the most commonly cited tool used by piscivores.
Piercing teeth work best when they are long and when they are pointing in the direction of the movement of the jaw. Because the jaws are paired upper and lower and both are required, prey restraint through piercing requires the teeth to enter at an angle that comes close to becoming vertical when the jaw is fully shut. This reduces splay of the teeth. An angled tooth, on the other hand, makes a poor piercing tool, as the object being held in the jaws is held along the length of the tooth, rather than at its tip, which increases bending strain. One of the more cited characters in anatomical analyses of fossil vertebrates for piscivory is the rostral orientation of teeth. This is true to a point, but only when backed up with recurvature of the crown, as in gharials (Gavialis) or river dolphins (Inia, Platanista), whose teeth curve so that the tips point towards the opposite set of teeth. For example, sea otters (Enhydra) have rostrally-oriented lower canines and incisors whereas other mustelids typically have orally-oriented teeth (as in river otters like Lutra); sea otters, however, are not piscivores, but durophages, and river otters are generalist carnivores favoring fish. Among fish, vertical and recurved teeth is pretty common (pike, gar, barracuda, but I show the pike topminnow above, Belonesox); as it is among aquatic snakes, for example Boa, Anaconda, and Hydrophis, the latter which is adapted for aquatic life so well it has lost the ability to move on land though the typical. In Belonesox especially the morphology of the upper jaw (palatoquadrate) is arched, allowing the upper teeth to curve inward with a hooking of the jaw (which isn’t normally visible while the jaw is closed), whereas the mandible is straight. This is even true in sperm whales (Physeter), in which only the mandibular teeth (very infrequently, the maxillary) are present, and still the teeth are recurved; the dwarf sperm whale, Kogia, has both upper and lower teeth bearing this morphology. Recurvature of the teeth and their conical form generally suggest piscivory, but it is more likely their arrangement is a greater inference on diet than their form:
For fish various, dolphins and other cetaceans, crocodilians, and even some terrestrial mammals, the conical recurved teeth are generally numerous, roughly the same size, and interlock. The shape of the jaw itself may be narrow, or broad, but also shallow and if so will likely have a vaulted palate of some form.
Prehension by the jaws (oral prehension) also occurs with beaks, most notably in turtles in which many aquatic predators have a hooked upper bill (e.g., Dermochelys, Macroclemys), but also in many pelagic birds (e.g., Phalacrocorax, Diomedea). However, hooking of the bill also occurs in taxa not particularly piscivorous, or directly determined not to be, such as most parrots. Thus the combination of a hooked bill – or a particularly hooked one – and aquatic habitus increases such an inference in diet.
Perhaps an even better indicator than the use of teeth or a beak is the engulfing method of prey capture. Instead of oral prehension, suction feeding is employed. This process involves the rapid opening of the mouth, creating suction, which can be aided by the depression of the throat, opening it wider. These things happen in some fish and turtles, and in these taxa the teeth are not particularly large (if present). Examples of suction feed occur in some of the fastest fish in the sea, marlin, sailfish, and other billfish (Xiphiidae, Istiophoridae), but also in many “lie-and-wait” predators (such as sessile turtles such as the mata-mata, Chelus). The distinction between the two also involves some distinctions in morphology. Typical suction feeding involves relatively static position of the predator, and suction causes the prey to move towards the predator; however, alternately, “ram-suction” has been used to describe predator motion anf engulfing moving prey, which involves merely the predator being faster than the prey. Relative motion of prey to predator is the dividing line, but exists on a continuum. Sit and wait versus pursuit. Pursuit predators, such as penguins, here using Aptenodytes and similar, are toothless, while comparable to the narrow, elongated jaws of some sharks, such as the mako, Isurus, or the more “primitive” Mistukurina, the goblin shark. The ability to “suck food in” emphasized a mode of prey acquisition in which teeth are simply not necessary, and this is firm evidence of a pisicvorous diet.
It should be pointed out that suction feeding doesn’t always point to eating fish specifically, but can also be used for any small animal. However, it is generally used against relatively small-bodied but mobile prey species; generally, fish. I will get into specific morphologies regarding suction feeding later, but the best indicators are robust depressor muscles and not adductor muscles, controlling rapid depression of the mandible; muscles of the palate and jaw that hold the jaw open during forward motion (against a current or the animal’s own forward momentum), and relatively few if any teeth, as well as slender mandibles. Many suction feeders have broad jaws, rather than elongated ones, and in some cases may be wider than they are long, as in toadfish (Opsamus), many anglers, and a variety of selachians. But this is all handwaving generalization, and catalguing those species with diets is a tricky thing given how few species have been studied on all variables I’ve discussed so far, especially the mechanism of prey capture versus type of prey.
Some mammals have been deemed suction feeders, such as walrus (Odobenus) and manatees (Trichechus), as they both employ rapid depression and opening of a vault-like oral cavity to pull in food fast, and this extends the premise to fossil taxa such as the cetaceous walrus-mimic Odobenocetops. However, the mouths of these animals open ventrally in a broad head, as they do in most rays, which are indications of a benthic feeder. Still, sessile organisms tend to be soft-bodied or at least soft within a shell, and can include fish. However, none of these taxa may be considered piscivores; Trichechus is an herbivore.
But second, there is also method of capture by the limbs. This one is trickier. Of all piscivores that tend to capture with the limbs, or a combination of jaw/limbs, they are non-aquatic foragers. These include the flat-headed cats Prionailurus and the giant otter shrew Potamogale. Both swim, but are clawed, and both use jaw + claw to acquire food. These methods are very similar to use of the limbs to capture prey on land. A two-handed prehension technique inhibits prey escape, and the pairing of claws on either side mimic the jaws of oral prehension. Manual prehension in some predators can be adapted: the greater bulldog bat, Noctilio leporinus, uses its feet, which as in most “microbats” are strong hooks for holding on to sheer surfaces or suspension, hanging by the feet.
Additional features have been cited as leading to piscivory: a long and low skull, a flattened head, shallow placement of the eyes or other sense organs, hooking of the rostrum (and not just the teeth). Most extant piscivores, however, vary on these factors. For example, among birds one may contrast penguins (pursuit aqua-fliers) and mergansers (pursuit divers) – one has smooth, slightly curved jaws without serrations and a low profile, with shallow-set eyes; the other has a high profile, high-placed eyes, and a shallow, straight rostrum with serrated margins. Cormorants, like anhingas, are another, with a moderate profile, high eyes, and deeply-hooked but shallow rostrum. One explanation is that for some predators, the rostrum should be shallow as it reduces the snout profile and thus drag when it swishes through the water, swift movements to grasp prey. Penguins, however, are ram-suction feeders, as are many fish, including billfish, tuna, not a few sharks, etc.
When next I return to this subject I should be approaching more specifics with regards to diet, in the regions highlighted. Even broad comparison to extant taxa is useful, but I think ultimately problematic. Piscivory is best quantified by behavioral data and association, as all that is being argued is the consumption of a small range of aquatic prey. It’s also a little biased: If a shark eats fish, it’s a piscivore; but if it eats, say a seal, it’s a carnivore. It matters not that the mechanisms, environment, and habit are identical. Ultimately, the term “piscivore” is, in my opinion, fairly useless. But I’ll get to that next.
[n1] “Archpredator” doesn’t refer to a lord high ruler of all Predator aliens from the Alien/Predator movie franchises, but to animals that are literally at the top of the food chain: Nothing much eats them, while they’re alive; or rather, nothing much hunts them, as adults. Some animals (e.g., lions) may be targeted as prey when young (by hyenas), but as adults they are generally free from assault, and this is true of orca, wolves, etc. We’re excluding humans from this equation, as the relationship of humans to these species tends to be trickier: Humans would qualify as archpredators, but we may also be prey to sharks, lions, tigers, etc.