History: The first member of the Troodontidae to be described was Troodon formosus in 1856 by Leidy, but all that was known by then was a single sharp tooth. Until the discovery of more fossil material, that tooth was placed within quite a few different families and species. Originally believed to be part of a great shark's dentition it was subsequently assigned to Megalosaurus and then to the pachycephalosaurid Stegoceras. In 1945 C. M. Sternberg finally and correctly argued that the tooth could only belong to a small carnivorous (meat eating) dinosaur. Only in the 1980ies the discovery of more skeletal material revealed that Troodon formosus (at that time named "Stenonychosaurus inequalis") was indeed a small theropod. D. A. Russel and R. Seguin (1982) were the first to complete a full restoration of the animal. By that time Saurornithoides mongoliensis (Osborn 1924) and Saurornithoides junior (Barsbold 1974) had already been described. Since it was clear that the three animals were very similar to each other they were placed in a family "Saurornithoidae" which was to become the now accepted Troodontidae lateron.

Feathers and endothermy ("warm-bloodedness"): In 2005 the discovery of Jinfengopteryx elegans prooved once and for all that Troodontidae were definatly adorned with feathers since that remarkable fossil had feather impressions preserved. Although the animal was placed inside the Avialae (primitve birds) by the authors of the first description (Qiang et al. 2005) Michael Mortimer and others argued very convincingly that it fits much better within the Troodontidae (see "The Theropod Database" under "Links"!). But what good would feathers be for an ectothermic ("coldblooded") animal? No good at all! Feathers are a very effective way to retain body heat through insulation, but this insulation at the same time prevents the body from obtaining heat from it's surroundings. Since ectothermic animals depend very much on environmental heat it would be rather hazardous for such animals to develop any sort of insulating cover. Modern reptiles for example greatly depend on quickly heating up their bodies to their preferred temperature in the morning, specially in moderate climates or environments where temperatures drop sharply over night ( for example deserts). If they cannot do that for whatever reason, many will not even become fully active, and some tropical reptile's digesting systems will not even work properly if their body temperature drops below a critical minimum. The commonly known Iguana (Iguana iguana iguana) cannot digest food properly once it's body temperature sinks below 20 °C. If the animal has to remain at such low temperatures over a prolonged period of time, it will get sick and eventually die! Thus feathers are a very strong indicator for endothermy.
Another important argument for endothermy in non-avian dinosaurs (and a direkt connection between feathery integument and endothermy) can be made of inspections of embryos in fossil eggs of Troodon formosus. All embryos within a clutch show a similar stage of development (Varicchio et al. 2002). There are two possible explanations for this observation, but one of them can be excluded with almost 100% security:
A - All eggs of a clutch were laid at the same time (this possibility can be excluded, see below at "behaviour").
B - The eggs were laid during an intervall of several days. Because of the difference between the temperature of the body of the mother and that of the environment, development of the embryos within the eggs was temporarily halted (the embryos didnt develop much due to the lower temperature). Only after clutch completion the mother - and/ or father - animal began brooding (wether or not both parents participated in brooding can not be stated with sufficient security at this point). Only now the temporarily halted development of the embryos continued. This observation can be made for (to my knowledge) all modern avian dinosaurs (birds). This is of essential importance for all those species which's chicks ar precocial since alternating hatching-times within a clutch would pose an almost insurmountable problem for the parents because of the need to look after the hatchlings on one side, and the necessity of further brooding for those eggs which havent hatched yet on the other! (Varicchio & Jackson in "Feathered Dragons", 2004)
Besides this one there are several other indicators pointing towards endothermic non-avian dinosaurs:
The bones of ectothermic animals show distinctive "growth rings" (similar to those shown in trees) which many non-avian dinosaur bones lack (Bakker and many other authors). However, growth rings are also known from a few endothermic animals and thus arent considered as good indicators by some authors (for example Ruben et al. 2005).
Another feature found in endothermic animals may be an even better indicator for dinosaur endothermy. Bones of endotherms show a dense system of specialized blood vessels while those of ectotherms show only few or even no such blood vessels at all in their bone structure (Bakker 1972). Since many dinosaur bones show the endothermic distribution of these specialized blood vessels, and some even show more of them then what is seen in modern endotherms, the picture points once again towards endothermic non-avian dinosaurs.
The upright posture that was kept by all dinosaurs also points towards endothermy. While modern (and extinct) ectotherm's legs are splayed to the sides, all living endotherm's and extinct non-avian dinosaur's legs are positioned directly under their bodies (Bakker 1972, Schweitzer 2005).
Moreover only endothermic animals evolve into obligate bipeds (animals that must walk on two legs) (Schweitzer 2005).
Another indicator for dinosaur endothermy is the predator to prey ratio. Since ectotherms need less energy (because they dont have to "waste" a lot of energy for keeping their body temperature stable) the predator to prey ratio is significantly higher in ecosystems where most top-predators are ectotherms then in endotherm-dominated ecosystems (Bakker 1972, 1986; Farlow date unknown). Since most non-avian dinosaur faunas studied apparently show predator to prey ratios similar to those of modern mammal dominated faunas, it is likely that these ratios indicate dinosaur endothermy. However it can not be excluded that these ratios may be the result of bias in the fossil record (Farlow, date unknown).
Lately the bones of Tyrannosaurus rex have been analysed regarding oxygen isotopes (radioactive oxygen atoms/molecules). The analysis results suggest that the animal's extremities (arms, legs, tail) were kept at an elevated temperature, close to that of the body's core. This in turn suggests that the animal may have been endothermic (Phil Bigelow on the DML 2005).
A remarkable fossil named "Willo" studied by a team of scientists (amongst them Dale A. Russell) revealed structures when scanned using the latest computer tomography techniques that the team interpreted as the remains of a four-chambered heart and a single systemic aorta. Such a configuration strongly points towards elevated metabolism rates for dinosaurs and in return towards endothermy (Fisher et al. 2000).
It may even be possible that endothermy was a common trait for the ancestors of avian-, non-avian dinosaurs and modern crocodylians, the archosaurs. All modern reptiles have a three-chambered heart - with one exception: The crocodylians. Seymor et al. (2004) argued very convincingly that the four chambered heart found in modern crocodylians may be a remainder of their archosaur ancestors and that crocodylians reverted to ectothermy when they occupied the semi-aquatic "wait and ambush" predator niche. If archosaurs were endotherms as the avian dinosaurs living today (birds) are, then it would be almost certain that the extinct non-avian dinosaurs were also endothermic animals (see Seymor et al. 2004a; Hillenius et al. 2004 and Seymour et al. 2004b for a good sequel of argument and counter-arguement!).
However there is one argument that appears to speak against endothermic dinosaurs: Respiratory turbinates. These very delicate (and thus not preserved in fossils!) structures are common amongst most modern endotherms. Ruben et al. argued in their 1998 publication that the lack of respiratory turbinates allows the inference that dinosaurs were ectothermic animals. These structures found in the nasal cavity (the cavity right behind the opening of the nose) function to warm up air before it reaches the lungs and "recycle" some of the humidity as the air is exhaled. Ruben et al. argued that, besides the fact that those turbinates cannot be expected to preserve in fossils, the nasal cavities of non-avian dinosaurs were generally to small in order to make it possible for them to possess respiratory turbinates. Apparently some non-avian dinosaurs had nasal cavities that could have been very well big enough to support respiratory turbinates (for example: Gregory Paul at the DML 1998). Paul (same source as before) also states that Ruben et al. used reconstructions of highly fragmental specimens in order to demonstrate the "narrowness" of non-avian dinosaur nasal passages and that these passages are actually quite a bit larger then thought when examined in better preserved fossils (Velociraptor mongoliensis). My personal view on this is that a) the absence of obvious evidence for turbinates is no evidence for their absence, b) quite a few modern endotherms show no such structures at all.
After all we know at least one group of dinosaurs that is still alive today and are also endothermic: The avian-dinosaurs, better known as birds!

Habits: The possibly best-documented parts of troodontid behaviour are their brooding and nesting habits. It seems that Troodontidae, like modern birds (1 egg) but differing from modern reptiles (several eggs), produced two eggs at a time in their paired oviducts (Horner 2000). The eggs were propably laid one at a time inside a shallow trough that was encircled by an about 10 cm high rim approximatly 15 to 40 cm outside the clutch perimeter (Varrichio et al. 1997, 1999). In MOR 963 (a fossil nest belonging to Troodon formosus) the diameter of the nest is approximatly 1,6 x 1,7 meters (Varicchio & Jackson in "Feathered Dragons", 2004). Thus the nests of Troodontidae were very similar to those of modern ratite birds (for example Ostriches and Emus). Some weeks ago I had the chance to take a look at an Ostrich nest at a nearby park (Safaripark Hodenhagen, Niedersachsen, northern Germany) and thus was able to look for similarities with my own eyes. A rim around the nest perimeter was not obviously visible, which may be due to the coniferous forest soil the nest was built on. However, the nest mainly consisted of a shallow trough which propably didnt reach more then 5 cm below ground level. A clear, almost circular perimeter around the clutch (which consisted of four eggs) was visible, very similar to the one described above. If the eggs would have been elongated, half buried in the ground (as described by various authors for Troodon formosus) and no Ostriches were visible in the area, the nest could have been easily mistaken as belonging to Troodon. The rim observed around nests identified as belonging to Troodon formosus is thought to be part of a nest structure independently constructed of the egg-laying process and representing an additional investment in marking a particular area around the clutch. It may have serverd to reduce the risks of predation and flooding (Varicchio et al. 1999, Horner 2000). The eggs were half buried in the ground in an upright position and remained "fixed" like that during incubation. The reason for this fixation is thought to be that dinosaur eggs lacked chalazae (chords of albumen that hold the embryo in place) and thus were vulnerable to movement which may have damaged the embryo inside the egg (Varicchio & Jackson in "Feathered Dragons", 2004). The finding of an obviously brooding Oviraptor (a theropod quite closely related to Troodontidae) stretched out right on a clutch of Oviraptor eggs provides strong evidence that active incubation (as seen in modern birds) was a common behaviour amongst non-avian maniraptoran theropods and thus for Troodontidae in general (Norell et al. 1995). The finding of an adult Troodon formosus alongside a nest belonging to the same species provides further evidence for the hypothesis that Troodontidae may have actively incubated their eggs. However, the fossil in question was not preserved in a brooding position. Thus other possible interpretations (stealing of the eggs by another Troodon, or simple protection behaviour for the nest) can not be excluded with sufficient security. (Varicchio & Jackson in "Feathered Dragons", 2004)
Troodontidae hatchlings are regarded as capable of locomotor activities (able to walk or run around right after hatching) and precocial (left their nests right or shortly after hatching) (Horner 2000). Neonate (very young animal) theropod toothmarks found on adult carcasses are interpreted as indicating parental feeding (Bakker 1997, Horner 2000). The (rare) findings of nests and young in association with adult remains are widely regarded as evidence for parental care in several dinosaur lineages. It may very well be that parental attendance of eggs and brooding behaviours evolved long before the aves (birds) amongst archosauria (Tullberg et al. 2002). Since biparental care systems can be found in both extant archosaurs and avian dinosaurs (birds) it is possible that male Troodontidae assisted the females in guarding their nests and taking care of the young (Tullberg et al. 2002).
Equipped with conspicous large eyes Troodontidae may have been most active at night times (Russell et al. 1982). Russell (same source as before) also suggested that Troodon formosus may have mainly preyed upon small, nocturnal mammals and emphasized on the relatively high encephalisation quotient (brain- to body-mass ratio) found in most Troodontidae as a possible indicator for such behaviour. However, shed teeth of Troodontidae are sometimes found in the Two Medicine Formation of Montana closely associated with isolated skeletal and egg remains of various small ornithopods (herbivorous dinosaurs from the Ornithischia order) (Horner 1994). This indicates that troodontids preyed on these propably diurnal (active during daylight) animals and possibly on their eggs (Holtz 1998). From what I was able to find in the literature it appears to me that the question wether Troodontidae preferred a nocturnal or diurnal lifestyle remains unresolved. Maybe they actually didnt prefer a special time of the day after all and were active as occasion served?
Whatever their preferred time of activity was, there can be little doubt that Troodontidae, just as their sister family Dromaeosauridae, led an active, alert life. They were propably not as agile as the Dromaeosauridae but instead were propably more adapted for speed, the arctometatarsalian condition, longer legs and slightly different femur to tighbone ratio being the main indicators for this assumption.
The visiual capabilities of Troodontidae were in all likelyhood very well developed (see above). Although similar to that of modern reptiles in some aspects, the middle ear construction in Troodontidae shows similarities to that of modern birds in others (Barsbold 1983). Thus it is likely that the sense of hearing was quite well developed in Troodontidae and acoustic means of communication possibly played a larger role in their daily life then as in other theropods (such as Allosauroidea for example), in which's life sound propably played a rather minor role except perhaps for the lowfrequency but high volume methods of long-range communication (roaring) important to ritualized behaviors (Rogers 2005). As for Dromaeosauridae, I could find no information regarding the sense of smell of the Troodontidae, but as stated for Dromaeosauridae it was propably not very much developed as is the case with other animals that rely mainly on their optical sense. Gregarious habits and pack hunting behaviour are very likely for Troodontidae, but I could not find any mention of fossil records for these in the literature available to me so far.
Finally I would like to get to the remarkably "abnormal" features found in some Troodontidae teeth. At this point there is no way around for me to mention and honor the remarkable work on this special subject done by Thomas R. Holtz, Jr. (accompanied by Daniel L. Brinkmann and Christine L. Chandler) and to thank him for the kind way in which he responded to my related question on the DML! You can find a downloadable .pdf version of the 1998 publication of these three scientists "Denticle morphometrics and a possibly omnivorous feeding habit for the theropod dinosaur Troodon" on Thomas R. Holtz, Jr.'s website (see "Links"). All said below (if not referenced otherwise) is solely based on this publication!
The teeth of theropod dinosaurs show (in the overwhelming majority of species) characteristic serrations (more or less sawtooth-like, in some cases hooked, exaltations on the forward- or backward-pointing sides and often even on both sides of the tooth). In many cases the wear patterns of teeth tell stories about the possible behaviours of the animals they belonged to (for example Schubert et al. 2005). Additionally the serration density, coarseness and appearance (sometimes) allows to identify the owner of the tooth/teeth down to family- and (in some rarer cases) even down to species-level (Farlow et al. 1991, Sankey et al. 2002, and many others)! I would also like to specially mention the work of Joshua B. Smith et al. regarding dental morphometrics in theropod teeth and their implications for species identification published in 2005. Moreover tooth serrations also can tell something about the feeding habits of the owners of the respective teeth. The serrations found in teeth belonging to herbivors (plant-eaters) are much coarser, larger and less densily distributed on the tooth-surface then are those of (primarily) carnivorous (meat eating) animals. And here it gets really exciting: The teeth of some Troodontidae (Troodon formosus propably being the most prominent example) show serrations that rather plot with (known) herbivorous dinosaurs concerning their coarseness, size and distribution on the tooth's (forward and backward pointing) sides, then they do with the finer, smaller and tighter packed serrations found in teeth belonging to theropods (known) to be primarily carnivorous! So some Troodontidae perhaps were herbivors after all? Very unlikely since there are at least five lines of evidence speaking strongly against a strictly herbivorous feeding habit for Troodontidae:
1. The enlarged predatory "sickle claws" on the second toes speak a very clear language saying: "I belong to someone who at least hunts from time to time!".
2. All fingers and the rest of the toes are also equipped with long, sharp claws which once more suggest "Hunter" and not "Plant Eater".
3. The lower jaws are similar to those of theropods in general and the overwhelming majority of those were carnivors (however, at least the Therezinosauria were herbivors and they possess a similar lower jaw construction too).
4. There is fossil evidence that Troodontidae fed on small ornithopods and possibly their eggs (Horner 1994).
5. The abdominal cavity of Troodontidae is not enlarged which is generally the case in herbivorous dinosaurs.
A possible (and I think most likely) solution to the riddle: Omnivory (feeding on both meat and plants)! The part mostly based on Holtz et al.'s 1998 publication ends here. Another piece of evidence that may point towards a possible omnivorous feeding habbit for the Troodontidae may be seen in the round objects which are visible in the holotype fossil of Jinfengopteryx elegans. Given the relatively good knowledge of the reproduktive tract of troodontids it appears to me impropable that these objects may represent eggs. It seems rather likely that these objects may represent seeds of a mesozoic plants within the digestive tract of the animal.
In summary I would like to conclude that Troodontidae apparently showed a lot of extraordinary adaptions such as the conspicous large eyes and high encephalisation quotient, additional investments in nesting and taking care of their youngs, a propably well developed sense of hearing and possible omnivorous feeding habits. Even amongst the already highly specialised Deinonychosauria, the Troodontidae seem to occupy a somewhat special position. I am very curious about what further study of this peculiar family of animals may reveal!