In 1938 Thompson noted that there were few dipterous parasitoids of Diptera in contrast to the frequent attack by Hymenoptera on Diptera. Among the Tachinidae, Admontia, Siphona and Trichoparia have been reared from tipulid larvae, and several species of the first genus are common on that host in certain areas. Other records of attack on Diptera must be questioned; and considering their large size and exposed position, it is unusual that the larvae of Syrphidae are not subject to attack by tachinids (Clausen 1940/62).

Insect orders with aquatic larvae are also almost immune to attack by Tachinidae; but the few aquatic Lepidoptera are readily parasitized, and the species responsible show pronounced adaptations for life in water. In many of these orders, the larvae are always immersed in water, often at a considerable depth. However, some of them leaf it for pupation in the mud or sand near the water's edge, at which time they would be vulnerable to attack, particularly by species having planidium type larvae (Clausen 1940/62).

Many Tachinidae show great host specificity, but there are some species that have a much wider host range than is known in any other group of parasitic insects. An example of the latter is Compsilura concinnata Meig, for which ca. 100 different hosts are known in the United States alone (Webber & Schaffner 1926), these representing 3 orders and 18 families. Generally, it may be that the species of this family are less restricted in their host range than the Hymenoptera.

Many native Tachinidae show a degree of parasitization and effectiveness comparable to that attained by introduced species. In northern Japan, Centeter cinerea Ald. destroys ca. 90% o of adult Popillia japonica beetles in alternate years, within 6-10 days after their emergence. These are in years of beetle abundance, while when populations are low the parasitization is much lower. The species is undoubtedly responsible for suppressing the pest to a nonproblem level in Japan (Clausen 1940/62). In North America, the red-tailed tachinid fly (reported as Winthemia quadripustulata F.) frequently destroys 50% of the armyworm population, and sometimes parasitization nears 100%. Trichopoda pennipes F. similarly builds up to a high level at destroys up to 80% of the adult squash bugs of several Pentatomidae. Ernestia rudis Fall. in Europe parasitizes a high percentage of the pine moth, Panolis flammea Schiff. Paradexodes epilachnae Ald. shows the remarkable capacity for increase of tachinids in the field. Colonies of 100-200 were able to build up to a parasitization of 50% or more in relatively heavy infestations of Epilachna over a radius of several miles from the point of release within two generations.

Overall the Tachinidae may be considered beneficial, for most species are primary parasitoids of plant pests. Unlike the major families of parasitic Hymenoptera, which have a varying portion of their species hyperparasitic, there are none in this category among Tachinidae. However, certain species are harmful because of their direct attack on hosts that are themselves beneficial. Several tachinid species that attack silkworms in Asia are very harmful. Before adequate control was developed, the loss of silkworms from tachinid parasitism in Japan was sometimes 80%. The tachinids also serve as vectors of the pebrine disease. Other cases of parasitization on beneficial insects are found in the species that attack adult Carabidae and honeybees. However, these are not widely distributed and usually are not abundant enough to seriously affect the host population.

Many species of Tachinidae have been used successfully in biological control. An landmark example is the biological control of sugarcane beetle borer, Rhabdocnemis obscura Boisd. in Hawaii through the importation of Ceromasia sphenophori Vill. from New Guinea in 1910. Also, in 1925 the biological control of coconut moth, Levuana iridescens B. B. in Fiji by Ptychomyia remota Ald. was achieved even though the parasitoid attacked a related host, Artona in its native Malayan range. In many other cases less spectacular results utilizing tachinids have, nevertheless, led to some reduction in the targeted pests. In the United States a high parasitization of the gypsy moth and other Lepidoptera has been obtained by Compsilura concinnata Meig., imported from Europe during 1906-1911. This parasitoid is credited with having significantly reduced infestations. Chaetexorista javana B. & B. of Japan was established in the northeastern United States where it exerts partial control of the oriental moth, Monema flavescens Wlk. although complete control was prevented by the parasitoid's inability to tolerate occasional extreme cold in winter (Clausen 1940/1962).

Tachinidae have been deployed for control of the sugarcane moth borer, Diatraea saccharalis F. in the neotropics. Metagonistylum minense Tns, or "Amazon fly" was originally discovered in the Amazon basic of Brazil. It was colonized in British Guiana and islands of the West Indies since 1933, where it has evoked a significant reduction of the infestation in some areas, Guyana in particular.

Tachinidae is a very large family with over 12,000 species known by 1993. They are distributed worldwide. The greatest number of species have been described from the tropics. Important morphological characters include a bristly body, especially on thoracic dorsum and 4th to 6th abdominal segments; postscutellum of mesothorax well developed, protruding posteriorly; pteropleural and hypopleural bristles present. The body is small to medium in size, gray or dull colored. Cell R-5 is narrowed or closed at the wing tip.

Most tachinids are primary, solitary, endoparasitoids, but some are gregarious parasitoids. They have a wide host range, with all major groups of insects serving as hosts. They are important to biological control, as >16 species have been introduced into the United States alone. Many more have been employed worldwide, mainly against Lepidoptera.

A discussion on host preferences of Tachinidae is probably best done on the basis of principal subfamilies. The Exoristinae are the dominant group, both in number of genera and in species known to be of importance as parasitoids of crop pests. Host preferences cover an extensive range, with the majority of species parasitic in lepidopterous larvae and in adult beetles of the families Chrysomelidae, Scarabaeidae and Carabidae. Genera parasitizing caterpillars include Anetia, Compsilura, Winthemia, Sturmia, Zenillia and Exorista. On adult beetles, Eubiomyia, Centeter and Chaetophleps are representative. Relatively few genera and species parasitize larvae of sawflies. A number of species attack larvae of Curculionidae, Chrysomelidae and Tenebrionidae, and sometimes there are species parasitic in larvae of Tipulidae and Vespidae and in adult phasmids, earwigs and locusts. Species of genera Doryphorophaga, Meigenia and Paradexodes attack chrysomelid larvae, and several species of Siphona are included among the few parasitic natural enemies of tipulid larvae. Important subfamilies are the Gymnosomatinae, Tachinae, Rutiliinae, Dexiinae and Oestrinae.

Biology & Behavior

Tachinidae have sustained extensive investigation since the beginning of the 20th century. Early detailed studies were by Nielsen (1909, 1912, 1918), Pantel (1910, 1912) and Baer (1920, 1921). Townsend (1934-1939) contributed much information on the reproductive behavior and egg of 1st instar larvae. Tachinidae show exceptional uniformity in host relationships as compared with parasitic Hymenoptera and Diptera. The larval stages are entirely internal during the feeding period, with one exception being Myiobia bezziana Baran., which Beeson & Chatterjee (1935) found to be an external parasitoid of caterpillars of a wood boring cossid in India, which was nevertheless questioned by Clausen (1940).

Most Tachinidae are solitary. Even in gregarious species it is only in rare instances that more than 3-4 develop in a single host, although sufficient food seems available for a larger number. A maximum of 16 individual Eubiomyia calosomae Coq. was recorded from a single Calosoma beetle, while up to 28 Palpostoma subsessilis Malloch reach maturity in scarab beetles in Australia. The largest number recorded was 110 Achaetoneura samiae Webber from a single Samia cocoon, and 550 A. frenchii Will, were reared from 44 S. cecropia L. cocoons. Surplus 2nd or 3rd instar larvae are usually killed in cases of superparasitization, resulting from an apparent overcrowding or starvation, but also at times by direct combat. If the number is very high, all the larvae may die and the host continue to maturity or also die. Clausen (1940) remarked that in one case a total of 147 dead larvae of Achaetoneura was found in a single S. cecropia larva.

Female Reproduction.-- Females have several modifications of the internal reproductive system which relates to the type of eggs or larvae deposited (please refer to Clausen, 1940, for diagrams). The basic type is one that produces the heavy shelled macrotype egg and in which not much uterine incubation occurs. C. cinerea is typical of this group. Each ovary has 9-10 ovarioles and a short oviduct leads to the uterus. The latter is also short and membranous for its entire length, with the stalks of the three spermathecae attached near the middle. In gravid females, each ovariole may have one mature egg and a series of immature ones. One mature egg is usually present in each oviduct and a single fertilized one in the posterior part of the uterus. Daily egg production is low, but extends over a protracted period.

Species that oviposit on leaves produce microtype eggs, which have several adaptations. Zenillia libatrix Panz. is representative. Dowden (1934) found each ovary to comprise 80-100 ovarioles. In young females each of these contains ca. 14 eggs in various developmental stages. The paired oviducts are long and slender, and the long posterior uterus is thick-walled for most of its length. The spermathecae are attached near the bases of the paired oviducts. After fertilization when eggs descend to the uterus, the uterus becomes greatly enlarged and may have several thousand eggs packed irregularly within. The coiled and expanded uterus may fill a large part of the abdominal cavity. All species of this group have many ovarioles, ranging to 460 in Leschenaultia exul Tns. (Bess 1936). Many genera of the Exoristinae have this kind of reproductive system and lay microtype eggs. Epidexia (placed in Dexiinae by Townsend) is the only member of the group known to produce microtype eggs.

Anetia nigripes Fall represents a third type that injects its larvae into the host's body. Each ovary consists of 12-14 ovarioles, and in unmated females the anterior and posterior uteri are about equal in length, with the spermathecae attached near the juncture. After mating, the eggs descend past the spermathecal openings into the posterior uterus, which becomes very much elongated, and lying in four coils. Eggs which most recently passed into the uterus lie transversely in an even row, but as they progress downward, they lie longitudinally and usually paired, with the head end directed cephalad with respect to the body of the parent female. Thus, the larva is ejected with its caudal end first. In gravid females of some larger species this "strap-like" uterus is very long, reaching its greatest development in Latreillimyia bifasciata F. where it measures 110 mm (Townsend 1936). Eggs range from fully incubated at the posterior end of the uterus to undeveloped at the anterior end (Clausen 1940/1962). This kind of reproductive system allows for the deposition of a relatively small number of larvae daily, although larviposition may extend over a long period. However, in Ernestia ampelus Wlk. and Compsilura concinnata all the eggs in the uterus are in the same stage of development. This allows the deposition of the full quota of larvae in 2-3 days (Tothill 1922). Maggots of this species lie in the reverse position in the uterus and are thus ejected with their heads first (Culver 1919).

Many of the tachinids which deposit their maggots on foliage or on the surface of the soil represent a fourth type. The uterus is somewhat coiled and strap-like before fertilization. After fertilization it becomes much distended and filled with enormous number of eggs in all stages of development, these often lying in precise transverse rows in the anterior portion and longitudinally in the posterior section (Clausen 1940/1962). The eggs of Echinomyodes are arranged with great precision in as many as 24 rows. This kind of reproductive provides for the deposition of a large number of larvae in a short period of time (Townsend 1936). Among species which deposit larvae or fully incubated eggs, the posterior uterus serves as an incubation chamber, with its walls abundantly supplied with trachea to satisfy the oxygen requirements of incubating eggs.

The rhythmic development of the eggs of P. epilachnae was observed by Landis (1940). The total number of ovarioles ranges from 16-54, each containing 6 eggs in various developmental stages, which represents the full reproductive capacity. Each egg has a brood relative in the entire series of ovarioles, and together they descend into the oviduct at about the same time. The successive broods descent in turn, and they may be distinguished in the uterus by the stage of incubation attained, the first batch being fully incubated and each following one being less developed than that preceding it.

Females of specie that inject eggs or larvae into the host body require an extensive modification of the external reproductive structures. The form taken depends on the kind of host and the amount of force necessary to penetrate the integument. A simple adaptation is that of species attacking caterpillars, which have relatively thin and flexible integuments. Adaptation for penetration of such hosts is found in such common genera as Compsilura, Anetia and Lydella. The 6th abdominal segment is developed into a curved, very pointed, thornlike process. This structure, or piercing organ, is deflected downward and in some species lies along the mid ventral line when not being used. It is grooved along the outer convex side, and the ovipositor glides along this groove in the deposition of eggs or larvae. Therefore, there are two distinct acts involved in larviposition, (1) the puncturing of the host integument and (2) the insertion of the ovipositor in the wound for laying an egg or larva. In species that larviposit they are usually placed directly between the peritrophic membrane and the cellular wall of the mid intestine rather than free in the body cavity of the host, and are then left to seek their ultimate destination (Clausen 1940/1962).

Chaetophleps setosa Coq. shows a striking adaptation for deposition of the egg or larva internally. This is a parasitoid of chrysomelid beetles of genus Diabrotica, which was described by Walton (1914) as Neocelatoria ferox Walt. and studied by Bussart (1937). The 6th abdominal segment is modified into a very long piercing organ, more than 1/2 the length of the abdomen. The 2nd abdominal segment is greatly extended ventrally into a laterally compressed structure which bears at its tip a large number of heavy, flattened, spine-like processes directed a bit caudad. During larviposition, the fly pounces on the beetle, and the latter turns on its back, after which its abdominal region is grasped in pincer-like fashion between the piercing organ and the extension of the 2nd segment. The ovipositor is then driven through the elytra and into the abdominal dorsum. However, many beetles are attacked while in flight or after alighting but while the wings are still spread and the ovipositor is inserted through the thin dorsal integument of the abdomen (House & Balduf 1925). Bussart (1937) found that oviposition occurs only while the beetles are in flight. This modification of the abdominal structure is clearly an adaptation for holding a hard bodied host during insertion (Clausen 1940/1962). Celatoria diabroticae Shim. also attacks Diabrotica beetles and is similarly modified. Clausen (1940) noted that a similar adaptation is found in the Conopidae, which parasitize adult bumblebees and some larger wasps.

Adult Behavior.-- In most Tachinidae adult emergence occurs during the early morning, although in some species of crepuscular or nocturnal habit emergence is in late afternoon and early evening. Adults feed on honeydew secreted by scale insects, leafhoppers and aphids and also on various plant secretions, particularly nectar glands. Some Dexiinae, such as Prosena sibirita F., possess a very long probosis which is used to feed at blossoms, mainly Umbelliferae. In the laboratory Landis (1940) found that females of Paradexodes confined in cages and fed sugar and raisins lived longer if the cage was dirty. From this it was inferred that the yeasts and other materials obtained from waste material have a nutrient value for the flies.

Tachinidae only infrequently feed on host body fluids, which is possible only in species having the piercing type of ovipositor. Host feeding was observed in Doryphorophaga doryphorae Riley, a parasitoid of Colorado potato beetle larvae (Bruneteau 1937); and it is found also in Anetia nigripes (Clausen 1940/1962).

Although crepuscular and nocturnal species frequently correlate their attack with similar habits of the host, there are exceptions as shown by Hamaxia incongrua Wlk. This parasitoid is active in late afternoon and early evening, even though one of its principal hosts, Popillia japonica Newm., is wholly diurnal and feeds during the morning and early afternoon. Other scarab hosts, particularly Sericinae, are wholly night feeders, but there is not much difference in the extent of parasitization because of this seemingly favorable behavior. The Australian Palpostoma subsessilis is also crepuscular, and Cryptomeigenia theutis Walk. is nocturnal as are adult Phyllophaga, its host.

Tachinids usually mate soon after adult emergence and in most species this is during the morning hours during bright sunshine. However, Carcelia gnava Meig. and other species mate at dusk. In the laboratory mating is often best secured by caging 1-2-day old males with newly emerged females. Temperature, light and humidity are important for influencing mating, the optimum range often being quite narrow. Various artificial stimuli have been tried to secure mating, such as exposure in vials to bright sunlight, vigorous shaking of cages, lining cage walls with green cloth, etc. Cleare (1939) in studies on Metagonistylum, determined that light intensity is the most important factor and that mating occurs only within a very narrow range.

Males are able to inseminate a large number of females, as shown by Dowden (1933) for A. nigripes, where one male successfully inseminated at least 13 females over ca. 4 weeks. Late mating revealed a progressive reduction in the proportion of the eggs that were fertilized, however. Females have been observed to mate repeatedly in several species (Clausen 1940/1962).

Gestation varies considerably in Tachinidae. A minimum of 2 days was recorded for Ptychomyia remota, and Winthemia required a little over 2 days. These tachinids deposit unincubated eggs, and it should be expected that the gestation period is short. The shortest period recorded for those depositing fully incubated eggs or maggots is that of Palpostoma subsessilis. This species was found to contain fully developed larvae within 4-6 days after adult emergence (Burrell 1935). Usually larviparous species have a gestation of 8-14 days, although this extends to as much as 4 weeks in Ernestia ampelus. In the latter all maggots in the uterus are of the same developmental stage, and deposition may be completed in 2 days or extend over one month (Tothill 1922).

Townsend (1908) distinguished five modifications of tachinid reproductive behavior, based on the position of placement of the eggs or larva with respect to the host, as (1) host oviposition, (2) leaf oviposition, (3) suprocutaneous host larviposition, (4) subcutaneous host larviposition and (5) leaf larviposition. This work was based largely on dissection of gravid females, and it was found that no only can the type of egg be determined by such dissections, but valuable clues as to the oviposition habit may be secured. Pantel (1910) presented a different classification, distinguishing 10 groups and using as a basis the female reproductive system, the type of egg, the stage of incubation of the egg and placement of the egg or larva at deposition (please see Clausen, 1940, p. 345 for details). Townsend (1934) listed 39 groups, the majority of which represent Tachinidae, on the same basis employed by Pantel, but utilizing also the general characters of 1st instar larvae (Clausen 1940/1962). This researcher distinguished oviposition from larviposition. Oviposition was considered to refer to the deposition of all eggs adapted for attachment to surfaces, whether flattened or provided with a pedicel, and of which the embryo may be in any developmental stage. Therefore, the deposition of all microtype macrotype and membranous eggs are defined included in oviposition. On the other hand, larviposition refers to the deposition of fully formed maggots, followed by their immediate activity, irrespective of whether they are naked or enveloped in the membranous chorion (Clausen 1940/1962).

In Tachinidae, reproduction may be oviparous, ovoviparous or larviparous, and various adaptive modifications occur as a consequence of these differences. There are four general types of eggs, the microtype, macrotype, pedicellate and membranous. These are associated with certain morphological modifications of parent females and serve to aid the parasitoid to reach the body cavity or some internal host organ (Clausen 1940/1962). Females of species depositing incubated eggs or larvae directly on their hosts and require the stimulus of the host presence for normal oviposition or larviposition can, under stress, deposit them at random in order to reduce pressure in the uterus. Those which inject them into the host body will retain them indefinitely if hosts are not available, and they may finally die as a result of penetration of the body cavity by imprisoned larvae, as was observed in Ernestia ampelus (Tothill 1922).

Reproductive Capacity.-- This varies greatly among different groups and species, being directly related to the position in which the eggs or larvae are deposited with respect to the host, and to hazards encountered before the larvae reach the host body cavity. The minimum deposition of eggs or larvae occurs in species where females inject their eggs or larvae into the host body or deposit them directly on the body. In this group are many of the most common and important species, such as Compsilura concinnata, Trichopoda pennipes, and Winthemia quadripustulata. In these species reproductive capacity is c. 100-200, with some depositing <100. Phorocera agilis R.D. has a capacity of >200 eggs, laid at the rate of 4-5 daily (Prell 1915, Burgess & Crossman 1929).

A second group includes species that deposit larvae or fully incubated eggs in the vicinity of the host. Bigonicheta setipennis Fall. places the eggs very near the host, or at times on it, and its total is relatively low (ca. 25), while others in which the association is not so close produce a higher number. Those placing eggs or larvae on foliage and of which the hosts are fee living caterpillars deposit 400-1,000 eggs or larvae. When hosts are enclosed in a tunnel in a plant stem, as in cane borers, and the larva is deposited near the entrance hole, the number is usually around 1,000. Gravid females of Theresia claripalpis v.d.W. contain >500 eggs and larvae (Jaynes 1933), and Metagonistylum minense parasitic on Diatraea and having the same general habits as Theresia, produces 500-700 maggots (Myers 1934b).

A third group are parasitoids of white grubs, which deposit their eggs or larvae on the soil surface. Prosena sibirita often contains over 800 larvae and eggs in various stages of development. Townsend (1934) found that ca. 2,000 eggs were in the uterus of a female Microphthalma disjuncta.

A fourth group includes species which deposit microtype eggs and those in which the larva is attached to the substratum by a membranous cup, which consists of the old eggshell, enveloping the caudal segments. The microtype eggs must be ingested by the hosts, while the fixed larvae are dependent on passage of the host larva within reach. The chances of reaching the host are more or less equal, as shown by the reproductive capacity of the two forms. Reamur (1738) estimated that a female Echinomyia contained 20,000 maggots, but this figures was considered too high by later researchers of this species. Clausen (1940) remarked that Von Siebold estimated 7,000 for Echinomyia fera L. However, Townsend mentioned 13,000 for Echinomyodes, which is the highest number determined by actual count. He also found ca. 3,2000 eggs and maggots in the uterus of Eupeleteria maginicornis Zett., which was thought not to represent the full reproductive capacity. Records for species with microtype eggs range from 2-6,000.

This method of reproduction in which the minute eggs are deposited apart from the host larvae and must be ingested by the latter before hatching, is known to occur in a large number of species, principally in the Exoristinae. Townsend (1908) recorded 14 species and Pantel (1910) listed European species of 8 genera with this habit. Since then a large number in many genera have been found to reproduce in this manner. Well known genera are Exorista, Gonia, Frontina, Parachaeta, Masicera, Sturmia, Gaedia, Chaetogaedia, Leschenaultia, Prosocilipes and Pales.

This kind of oviposition was first observed in S. cilipes Macq. (= sericariae Rond.) by U. Sasaki in 1873, and the first record is that by H. Pryer in his catalogue of the Lepidoptera of Japan, published in 1884, in which he stated, "I have noticed that the Uji, a diptera, which is parasitical upon it and causes an immense amount of damage, deposits its eggs about the larva on the leaves and not on the insect." (Clausen 1940/1962). Dr. C. Sasaki, son of the discoverer, published an extended account of the habits of S. cilipes in 1886. Such behavior was so bizarre at the time that little credence was given to it for many years. It was not until 1908 when Swezey described the behavior of Chaetogaedia monticola Big. and Townsend that of several other species in the same year, that full credit was accorded to Sasaki's contribution. Clausen (1940) remarked that the parallelism between the course of events here outlined for Tachinidae of this type, and that which takes place in the Trigonalidae in Hymenoptera, is especially interesting, because in both cases the eggs are minute, hard of shell and deposited on the food plant of the host. They are capable of remaining viable for long periods, and are consumed by the host and finally hatch in the digestive tract from which position the larvae migrate into the general body cavity.

Host Stage Attacked.-- Most Tachinidae attack the larval stages of the host. This behavior is consistent among those which parasitize Lepidoptera, and none is known to attack the pupa directly, although some complete feeding in that stage. Several species, such as Zenillia libatrix, delay development beyond the first instar until host pupation. Those which deposit macrotype eggs on the host body usually limit themselves to the late larval instars, while others may gain entry at almost any time during the larval period. Dexia and Prosena are able to parasitize scarab grubs in any stage, but they do not complete feeding until the latter are mature. Tachinids parasitizing chrysomelid larvae, such as Paradexodes, sometimes successfully attack the prepupa and pupa (Clausen 1940/1962). Edelsten (1933) reported an exceptional case of emergence of a tachinid maggot from an adult moth. A female Zygaena lonicerae Esp. emerged normally but lived only two days, during which she laid 30 eggs. Ten days later a mature maggot of Phryxe vulgaris Fall. emerged from the body, after having completely consumed the contents. Previously the emergence of a larva of this species form a living female of Nyssia lapponaria Boisd. was reported. Emergence of Manduca atropos L. moths from pupae that had yielded tachinid maggots, the latter having emerged from the wing pads, has been also observed (Clausen 1940/1962).

A large number of tachinids attack only the adult stage of their hosts, which is true of almost all those which parasitize Orthoptera and Hemiptera. It is interesting that many species attacking adults tend to limit their oviposition to the female sex., which may be important from the point of view of natural control. Centeter cinerea lays 80-95% of its eggs on female beetles of Popillia japonica Newm., and other examples of this behavior are Hyalomya aldrichi Tns., attacking mostly adult females of the false chinch bug, Nysius ericae Schill., and Thrixion halidayanum Rond., which is restricted to female Phasmidae. By contrast, most chrysomelid hosts of Chaetophleps setosa are male (Clausen 1940/1962).

Silvestri (1910b) studying Erynnia nitida R.D. found a very unusual seasonal differentiation in host selection, where the larvae of the elm leaf beetle were attacked by the summer broods of the parasitoid and only the adult beetles by the last brood. Clausen (1940) noted a comparable peculiarity in behavior with a few species of Hymenoptera.

Among tachinids that deposit eggs directly on the host body, most are often found to be largely confined to a particular part of the body. Species of a single genus may differ in this respect, even though they attack hosts of the same group which are similar in size, form and behavior. Therefore, Centeter cinerea places most of its eggs on the dorsum of the thorax of female P. japonica (Clausen et al. 1927). While ovipositing, the female parasitoid usually attacks pairs of beetles who are mating and dashes diagonally across the female's thorax, lowering the tip of the abdomen momentarily to deposit the egg. By contrast, C. unicolor Ald. attacking Anomala and Phyllopertha beetles, places its eggs ventrally on the posterior portion of the abdomen (Parker 1934).

Among Tachinidae attacking caterpillars, the variation in position is also great. On free living caterpillars, either the last two thoracic segments or the last abdominal segments are usually chosen. In W. quadripustulata the thoracic position on Cirphis unipuncta Haw. is a provision for the eggs' protection, as a high percentage of those placed farther back are crushed by the host's mandibles (Allen 1925). The presence of these eggs apparently causes some irritation, which causes the host to try to brush them off or destroy the. An unusual position for egg placement was recorded by Ainslie (1910) for Exorista larvarum L. on larvae of Hemileuca oliviae Ckll. The latter is attacked only while in motion and at the moment when the posterior portion of the abdomen is raised. The egg is usually placed on the sole of the psuedopod, within the crescent of hooklets. On Pentatomidae and other Hemiptera, the tachinid egg may be placed on the side of the abdomen or thorax, which is common, on the venter of the prothoracic margin, as in Gymnosoma fuliginosa R.D., or on the dorsum or sides of the abdomen while the wings of the host are spread, as in Phasia crassipennis L., Siphona geniculata DeG. and S. cristala F. (Roubaud 1906), which parasitize tipulid larvae, oviposit on the stigmatic crown, presumably because this is the only portion of the host body exposed (Clausen 1940/1962).

In Carcelia gnava the pedicellate egg is attached by the tip of the pedicel to a hair of the caterpillar host (Clausen 1940), while in C. evolans Wied. the egg is placed on one of the thoracic segments of the bagworm host, a position which is obligatory because that is the only part of the body that is ever extruded from the bag (Skaife 1921b).

In hosts that inhabit soil, the eggs or larvae of the tachinids are laid on the soil surface, although Davis (1919) found that the female of Microphthalma disjuncta Wied. places them in crevices. They are nevertheless probably placed in proximity to host grubs in considerable numbers rather than singly. Species attacking hosts in plant stems, such as Theresia and Metagonistylum, which attack sugarcane moth borer, place larvae near the entrance of the host tunnel. They must burrow through the frass that fills the entrance before reaching host larvae. This behavior is found in a great many species attacking hosts that are concealed but that have an open entry or later make holes for other purposes in fruits, stems or seeds (Clausen 1940/1962).

Female Rondanioestrus apivorus Vill. pounce on worker bees while they are in flight, but touch the body only lightly to lay the maggot on it (Skaife 1921a). Tachinids that lay macrotype eggs seem to show no discrimination in their choice of hosts. This often results in individual hosts receiving an excessive number of eggs. A certain portion of these eggs is lost through molting of the host, this varying with the length of the incubation period and the interval between host molts. If the egg incubation period is the same as the length of the host larval stage, there would be virtually a complete loss of the eggs before hatching, while if the egg stage is, for example, 3 days and the host larval stage 6 days, the loss from this cause would be ca. 50% (Clausen 1940/1962). It seems that early portions of the host stage, immediately after molting, are preferred for oviposition, which would reduce the loss considerably.

At time there is a failure in successful parasitization even when oviposition is extensive. Toward the end of outbreak periods of the nun moth, Lymantria monacha L. in Europe, almost every caterpillar bears tachinid eggs and yet the attack is rarely successful, with the hosts developing and emerging normally. In a collection of 235 gypsy moth caterpillars, each bearing 1-33 eggs of Exorista larvarum, only four parasitoids were produced. Another collection of 252 did not yield a single parasitoid. Although only part of this loss can be explained by the molting factor, it probably accounts for the loss of a considerable portion of the reproductive capacity of species laying unincubated eggs on caterpillars (Clausen 1940/1962).

One field collected larva of Datana minestra Drury yielded a maximum of 228 macrotype tachinid eggs (Clausen 1940). A larva of Samia cecropia with 40 tachinid eggs was still able to reach adulthood. Tothill et al. (1930) noted where 72 eggs of Ptychomyia remota were deposited on a single Levuana larva, in which only one parasitoid can develop to maturity. Observations on Winthemia indicate that the number of eggs laid on different hosts and on various instars of the same host vary directly with the host's size (Allen 1925). Clausen et al. 91933) showed in Centeter cinerea that there was no selective oviposition, but that the egg distribution appeared random.

Microtype eggs are laid on plant foliage that serves as host food and adhere to the leaf surface by a mucilaginous material that is partially water soluble. Sometimes a particular plant or plant group serves as the oviposition stimulus, which is thus independent of the host itself. Other times the attraction seems to be to foliage bearing, or visited by, host larvae. Dowden (1933) observed that cut leaves, as well as the presence of host larvae, stimulate oviposition by Zenillia libatrix, which simulates the condition existing while the host larvae feed. Eggs of Racodineura are laid on any plant material upon which earwigs have fed the previous night. In most species, the eggs are laid on the undersides of leaves, scattered about, but in other species the are placed at the leaf margins. Females of Gonia capitata Deg., parasitic on Porosagrotis in North America and Europe, lay most of their eggs on the upper sides of leaves of Graminae, in particular the bluejoint grass, Agropyron smithii (Strickland 1923). However, the host feeds mainly on the cultivated grains and attacks the bluejoint grass only when other preferred vegetation is not available. It also feeds extensively on alfalfa, but the parasitoid does not oviposit on that plant. The parasitoid's value is thus restricted, for host larvae on their preferred food plants are able to avoid attack. The extent of oviposition of Gaedia puellae Nishik. on mulberry foliage is correlated with the infestation of aleyrodid, Bemisia myricae Kuw., on the secretions of which the female flies feed (Nishikawa 1930). The silkworm, principal host of this tachinid, does not occur on mulberry in the field, although species of Acronycta, Bombyx and Porthesia, some of which are usually present, are more normal hosts (Clausen 1940/1962).

Immature Development (Egg Incubation, Hatching and Host Entry).-- There is a wide range of behavior regarding incubation, hatching and host entry. Activities of 1st instar larvae in penetration of the host vary as much as any other group of parasitic insects (Clausen 1940/1962). The frequent occurrence of partial or complete uterine incubation of the several types of eg produced, which is relatively rare in other dominant parasitic groups, particularly Hymenoptera, serves as one means to overcome or avoid certain hazards that might otherwise be disadvantageous during external incubation.

Macrotype eggs generally undergo the entire embryonic development outside the parent female's body. Occasional rare exceptions this occur, such as in Ptychomyia remota (Tothill et al. 1930), where there is a partial and variable degree of uterine incubation. The normal period of external incubation in this species is 36-50 hrs; but some eggs have been found to hatch in 30 min., and other required 4 days. Among other species in which the incubation period has been determined, it most frequently requires 2-3 days, with a minimum of one day. The eggs of Eubiomyia calosomae are almost fully incubated at the time of deposition, usually hatching in <24 hrs., though some hatch in <3 hrs (Collins & Hood 1920).

Two ways exist in which hatching and entry into the host body are accomplished by larvae from macrotype eggs. Centeter, Trichopoda and Meigenia and others have the indehiscent form of macrotype egg where the larva bores directly downward through the thin chorion on the ventral side of the egg and through the heavily chitinized integument. Of course there is no external evidence to indicate that hatching has occurred. This method is particularly common among those species attacking adult Hemiptera and Coleoptera. Larvae from eggs of this type are usually provided with teeth on the distal margin of the mouth hook, which serve to rasp (see Clausen, 1940 for diagram). An exception is shown by E. calosomae, in which young larvae escape from the egg through a hole in the thin ventral chorion. However, instead of continuing into the host body immediately, they emerge from beneath the egg and enter at some other point on the host. In the second form, hatching occurs by the lifting of a definite lid, or operculum, at the anterior end of the egg. The fracture takes place along a horizontal line around the front of the egg, a bit below the median transverse line and often extending slightly dorsad at each end. The larva emerges partly from this opening and with the caudal segments still enclosed within it, braces itself and penetrates the integument just in front of the egg. Ernestia, Phorocera and Winthemia have this behavior,a nd it is probably common among the species attacking caterpillars and other relatively thin-skinned hosts (Clausen 1940/1962). In some species the larva abandons the eggshell altogether and enters through the intersegmental membrane or at some other vulnerable spot.

In the majority of tachinids that lay microtype eggs, uterine incubation is also partial or complete, and the entire quota of the female may be present in the uterus and partly or completely incubated before any of them are laid (Clausen 1940). Such species, as Zenillia libatrix, are thus able to lay a large number in a short time. Such eggs, whether or not incubated at the time of laying, are protected from desiccation and injury by the heavy and variously sculptured chorion. Hatching does not occur until they are ingested by the host larva. They remain viable for 3-5 weeks in most species, with a record from Gonia ornata Meig. being alive 2/5 months after laying, at which time the caterpillars of the 2nd generation of Euxoa were present in the field (Sakharov cited by Clausen, 1940). This ability to persist in an inactive condition for a long time is advantageous to the species and counteracts, partly at least, the disadvantage of being laid apart from the host. Several genera and species with apparently mature eggs in the uterus show very little evidence of embryonic development, being laid before this is complete (Townsend 1908, Clausen 1940). In these species, ingestion of the eggs by host larvae cannot lead to successful parasitization unless the eggs are at least 1-2 days old.

Clausen (1940) pointed out that opinions vary concerning the manner in which hatching of microtype eggs takes place. The chorion of the ventral side of the egg is thin and membranous, contrasted with a thick walled dorsum. Townsend implied that the action of the digestive juices provides the stimulus for hatching and considered that the heavy chorion served to protect the embryo from injury by the host mandibles while the egg was being swallowed. Swezey (Clausen 1940) questioned this, and believed that the chorion was cracked by the mandibles of the host, thus allowing the larva to escape. He supported his conclusion by the snapping open of the shell, and the escape of the larva, when pressure is applied to it. Also, hatched larvae were found in large numbers in the crop of caterpillars almost immediately after the eggs were eaten. Severin et al. (1915), Nishikawa (1930) and Dowden (1933) immersed eggs of Chaetogaedia in the fluids ejected from the mouths of Heliophila larvae; hatching of some of the individuals resulting in less than one minute and 97% hatching within 3 hrs. The same results was obtained with the juices of other caterpillars. The presence of such fluids stimulates the larva, and the thin ventral chorion of the egg is broken as a result of its movements. Many eggs hatched when immersed in distilled water for 36 hrs, and this was brought about by an increased turgidity of the larva (Clausen 1940/1962). The egg absorbs a lot of fluid, often resulting in a doubling in size. In some species the thin ventral chorion bulges out, blister-like, so that the greater part of the larval body is enveloped by only a thin membrane. As a result, the rupture of the heavy dorsal chorion is not required.

Nishikawa confirmed the conclusions of his colleagues with respect to Gaedia, stating that hatching occurs only after immersion in the digestive juices; but he found that a higher percentage hatch resulted with a higher concentration. The mouth hook apparently breaks the ventral chorion of the egg. Dowden (1933) reported that the mere immersion of the eggs of Z. libatrix in the digestive fluids of the host fails to induce hatching but does result in a pronounced swelling which renders them susceptible to rupture due to variation in pressure within the host's digestive tract. Hatching takes place in this species at any point in the digestive, while in others it is mostly in the fore intestine.

Most Tachinidae in which the eggs undergo partial or complete incubation while still within the uterus of the parent produce the membranous type of egg, and in some cases actual hatching of all eggs in the uterus occurs. Female Prosena sibirita never deposit eggs and the eggshells are retained in a "brood pouch," whereas in Dexia ventralis they are voided at the time of larviposition. In the latter, uterine incubation is usually complete, with hatching taking place before larviposition. During periods of extensive reproductive activity, the brood pouch may be emptied before all the eggs are fully incubated, and some of them may require as much as two days of further development before they can hatch (Clausen 1940/1962).

Larvae that are laid externally as such or that arise from the membranous type of egg usually enter the host through the intersegmental membranes or at some other point where the integument is thin. This is especially true of the species attacking heavily armored hosts, such as beetles and locusts. In several species, such as Lixophaga diatraeae Tns., Siphona cristata, Eubiomyia calosomae and S. geniculata, it was found that entry is through a spiracle. Clausen (1940) expressed some doubt about this, however. Strickland (1923) noted that many of the planidia of Bonnetia comta Fall. are bitten off and killed by the cutworm larva while they are trying to penetrate the integument, and Muesebeck (1918) noted that brown-tail caterpillars make frantic efforts to dislodge or destroy the larvae of Sturmia nidicola Tns. as they bore into the body. This mortality factor does not operate in the case of species with microtype eggs, and is comparable to the loss of macrotype eggs through the same agency. Most hosts, however, show no discomfort during the time maggots are penetrating.

Stimuli attracting planidium type of larvae to their hosts are not completely understood. Many attach themselves to almost any moving object within reach. Those of Archytas analis F. are attracted to many species of caterpillar in which they cannot develop, but they show no interest in certain other species. Generally, the larvae that find the host and enter its body through their own efforts do so immediately after coming into contact with it. Complete penetration is often achieved in 15 min., though some species require more time. The variation is probably due mainly to a difference in the thickness and toughness of the host's integument. However, in the case of the planidia of A. analis, which, when they reach the host exude a liquid that fastens them horizontally to the skin of the host. They may remain in this position for 24 hrs. or longer before attempting to penetrate (Clausen 1940/1962).

First instar larvae of Dexia ventralis Ald. and other species of Dexinae, must search through soil for their hosts and have the planidium type larva. The penetrate the integument at almost any point as soon as the host is located. Experiments have shown that the larvae do not discriminate among grubs, but that differences in parasitization of different host species and instars is related to the thickness and hardness of the integument. In this way, under comparable conditions there was a parasitization of only 18% of mature Anomala grubs and 85% in 2nd and early 3rd instar grubs of Phyllophaga, the integument of the latter being very thin and bare. Although initial parasitization was high, the parasitoid could not mature in the latter, however.

Many species, such as Bonnetia comta, Ernestia ampelus, Archytas analis and Eupeleteria magincornis, use the egg shell as a cup-like device that serves to anchor the larvae to the substrate while awaiting a host. The shell is fastened to the leaf or other surface by a mucilaginous substance from the colleterial glands of the parent female. it closely envelopes the caudal end of the larva (see Clausen, 1940 for diagram). B. comta larvae stand erect in the shell, even when resting, with the anterior segments retracted, whereas larvae of A. analis lie horizontally on the substratum when resting (Allen 1926). The presence of the collapsed eggshell is not essential to the well-being of the planidia. Both of these species frequently leave the shell entirely when excited, such as occurs when a host approaches. They are still able to assume the erect position at will. It does not seem that this adaptation serves any essential purpose, however (Clausen 1940).

The majority of species that inject their membranous eggs directly into the host body partially incubate their eggs in the uterus. However, a few regularly lay them before much embryonic development has occurred. The pedicellate egg, which is attached to a hair or to the caterpillar's integument, hatches quickly after deposition. The young larva seeks out a vulnerable spot on the host where it enters. Its activities thus do not differ much from those of other forms of larvae from membranous eggs, except that it is spared effecting the initial contact with the host.

Larval Activity in the Host.-- Clausen (1940) mentioned that there is much less diversity in behavior among Tachinidae once the host is penetrated. This was because for all larvae the same general medium is inhabited, even though the final destination after entry differs a lot among different types of larvae and species. Most species do not associate with any particular organ, yet others have a regular habit in this regard. The organs with which they may be associated are nerve ganglia, gonads, salivary glands, intestines, muscles and fat body. The greatest adaptations are found in microtype larvae, which find their way into the body cavity from the intestine. Young Sturmia cilipes larvae enter one of the nerve ganglia of the silkworm, usually the 2nd to 5th, and during its stay of about 1 week in this position causes a proliferation or enlargement of the ganglion, which also changes in color to white.

Clausen (1940) remarked that the occurrence of these larvae in the ganglia resembled the same localization of attack by some species of the hymenopterous family Platygastridae. In Gonia capitata, the maggot first remains for a period of 4-28 days in the mesenteron of the intestine without feeding. Then it progresses to the salivary gland, and finally reaches the supraesophageal ganglion. A few species are known to inhabit the salivary gland throughout this period. Gaedia puellae, parasitic on silkworms, has the habit and may occasionally be found in the reproductive organs. The period passed within the salivary gland by some species varies directly with the age of the host larva, which may be from 4-22 days. The maggot of Leschenaultia exul may reach the gland within 2 hrs. after ingestion of the eggs by the caterpillar, and it remains there for 8-10 days. It lies in the gland itself, rather than in the duct, and a pronounced malformation is produced. The first-instar larva of Zenillia libatrix may be found in a muscle, the salivary gland or occasionally in a histoblast. S. scutellata R.D. consistently passes the first stage in a muscle, and its presence results in hypertropy of the tissue. In Ghaetogaedia monticola this period is passed in a sac, apparently formed by an enlarged tracheal tube, near one of the host spiracles. However, this may be only a respiratory funnel plus a membranous sheath. First instar larvae of Racodineura antiqua Meig. lie free in the body cavity of the host, thereby departing from the usual habit of this group.

The planidium and tachiniform larvae which gain entry to the host through the body wall, either by direct penetration or by injection by the parent female, are not known to enter nerve ganglia or the salivary gland at any time. Maggots of Plagia trepida Meig. and P. ruricola Meig. enter a muscle and pass the entire first stage therein, a habit they have in common with S scutellata. However, in such instances the tissue is killed and thus no enlarged pouch or sac is formed (Thompson 1915b). The larvae of Rondanioestrus apivorus, parasitic in adult honeybees, feed in the abdomen throughout the developmental period, while those of the sarcophagid, Myiapis angellozi Seguy, are often found lodged in the thoracic muscles (Seguy 1930). The majority of these species lie free in the body cavity of the host, for the greater part, if not all, of the first stage, but a number are intimately associated with the intestine (Clausen 1940/62). Young maggots of Compsilura concinnata are always found between the peritrophic membrane and the cellular wall of the mid-intestine, and the same is found in Anetia hyphantria Tot. A. piniariae Hart and A. nigripes are found in the mid-intestine, frequently attached to the walls by spiracular hooks. The young maggot of Zygobothria nidicola (Muesebeck 1922) lies free in the body cavity for 10-14 days after penetration and then enters the esophagus, where it lies dormant for ca. 9 months, while that of Archytas analis remains for a period up to 15 days between the skin and the hypodermal layer before entering the body cavity. The larvae of Chaetophleps setosa are often located in the fat body (Clausen 1940/62).

There may be an obligatory association of certain species with a salivary gland, muscle, nerve ganglion or intestine, but not when in a gonad or the fat body (Clausen 1940). Species associated with a definite host organ as 1st instar larvae usually leave it right before or after the first molt, and most then assume a fixed position in the hot body for the rest of the feeding period, which is related to respiration.

A large number of species have 1st instar larvae that pass a long period of time without feeding or apparent growth, especially those which are association with host organs. However, one this starts it is extensive before the molt occurs. Zenillia libatrix, e.g., increases from 0.23 mm. to 2.0 mm. in length before the end of the 1st instar.

Some free living species show migratory behavior during development, as shown by Centeter cinerea. The egg is laid on the thoracic dorsum of the female beetle, and young larvae bore directly downward into the host. Feeding and the first molt occurs in the thorax, and the 2nd instar then enters the abdomen right after the molt. It slowly works its way to the tip of the abdomen, turns and reenters the thorax. The host dies and the second molt occurs. The thorax contents are consumed after which the larva reverses its rout to complete feeding in the abdomen. However, in male hosts the second molt occurs in the abdomen rather than the thorax. First instar larvae of A. analis persist in the host caterpillar until the latter pupates, after which it molts and then positions itself in a wing pad, in which it causes an easily recognized characteristic bulge. The respiratory funnel is formed at this position (Clausen 1940/1962).

In some tachinids the larvae are partially or completely enveloped by a membranous sheath, which like the respiratory funnel, is of host origin. It is soft and flexible, of varying thickness, and almost opaque in species that induce its greatest development. Clausen (1940) believed it could be the result of a defensive reaction on the part of the host similar to that where phagocytes attack a foreign body, and thus it differed from funnel formation, which results from healing. The sheath of Sturmia is made of hypodermal cells, leucocytes, and compressed fat cells and envelops only the funnel and the posterior portion of the body (Muesebeck cited by Clausen, 1940). In Siphona cristata and other species, young overwintering larvae are completely enclosed in the sheath, but older individuals have the sheath open at the anterior end. A closed sheath also envelops young larvae of Acita diffidens Curr. and Winthemia quadripustulata, which occur in caterpillar bodies only in midsummer, and thus the sheath must be permeable to host blood from which larvae derive nourishment.

Host death does not necessarily occur in the same stage as that which is originally attacked. Among Lepidoptera, initial attack is frequently on larvae, usually when they are half grown or larger. In most host species, death does not occur in that same stage, but a large number consistently pupate. A few tachinid species are indiscriminate in this regard, and the stage at which the host is killed depends on the age of the larvae when parasitized. Many tachinids kill the host early with respect to their own stage of development, often when they are in the 2nd stage, and death is followed by a rather complete liquefaction of the host body, this conditions being distinct from putrefaction.

Cirphis unpuncta caterpillars parasitized by Winthemia die two days after the parasitoid larvae penetrate, but the latter are still able to complete development. Death of worker bees parasitized by Rondanioestrus apivorus occurs suddenly, often they are stricken while in full flight and die within minutes after falling to the ground. The mature larva emerges from the body within 10 min. thereafter. Because of the continued activity of the affected bees until the parasitoid larvae are mature, it is possible that the latter feed only on the body fluids and that, when ready to emerge, they cut the nerve cord and thus cause almost instantaneous death (Clausen 1940/1962).

There are several variations in the manner of emergence of the mature larva from the host, which depends on the host stage and whether or not it is alive. In larval hosts, the mature tachinid larvae usually make an incision in the ventral area of the abdomen, at which point the integument is very thin. Some researchers believed that this was done by use of mouth hooks, but other thought that it occurs by pressure of the caudal end, aided by the solvent action of body secretions. The emergence of Bessa selecta Meig. is with its rear end foremost (Nielsen 1909). Microphthalma michiganensis Tns, attacking scarab grubs, dissolves a large opening in the body wall of the grub permitting exit (Petch & Hammond 1926). The aperture may be made some time before actual emergence and may be used in the meantime for respiration. Emergence from pupae of Lepidoptera usually occurs at some point on the body venter and sometimes from the wing pads.

In hemipterous hosts, many species are still alive at the time of emergence of mature tachinid larvae. In Nezara and Anasa, parasitized by Trichopoda pennipes, Eurygaster by Clytiomyia and Dysdercus by Alophora and Catharosia, the larvae leave the body through the anal opening or through the intersegmental membrane close by. The host does not die for several days. Coleopterous hosts also show a similar condition at times. The larvae of Minella chalybeata Meig. emerge from the chrysomelid beetle, Cassida deflorata Suffr. through a dorsal aperture between the 1st and 2nd abdominal segments. Because the vital organs are not affected, the host does not die until later. The death of parasitized earwigs also follows, rather than precedes, parasitoid emergence from the body, and maggots exit through the intersegmental membranes near the posterior abdomen. Larvae of Thrixion emerge from the body of the phasmid host through the wound at the side of the thorax which had been previously used for respiration, and thus the mechanical injury that occurs at this time is slight.

Several species, such as Zenillia pexops B.B., that attack larval hosts in exposed situations, pass winter as mature larvae within the dry host skins. Thus they are exposed to sudden changes in both temperature and humidity and must adapt to such conditions. These larvae are golden yellow in color owing to large quantities of fatty substances stored within the body, and the integument is much heavier than in species not so exposed (Clausen 1940/1962).

Tachinidae generally have three larval instars, but four and possible five may occur in some as in Microphthalma michiganensis. In Actia diffidens, Paradexodes epilachnae, and others, it was noted that the inner wall of the puparium is lined with a distinct transparent membrane, the true cast skin discarded after a short but definite prepupal stage (Clausen 1940).

Pupation.-- Tachinidae pupate at variable sites, especially those attacking Lepidoptera. Most form the puparium outside the host, those from free-living caterpillars usually enter soil, while others developing in stalk borers, leaf rollers, etc. usually pupate within the burrow or leaf roll. Some such as Voria ruralis Fall, Echinomyia fera and Sturmia nidicola pupate in the host larval skin. In solitary parasitoids the puparium is oriented the same way as the host, but if gregarious the puparia lie transversely in an even row. Species reaching maturity in the host pupa often form the puparium within the pupal case. Pupation on sawfly larvae usually occurs either within the larval skin or in the soil. When attacking Coleoptera, tachinids usually pupate within the larval skin. When adult beetles are hosts, most species pupate within the abdomen of dead beetles, with the head at the posterior end of the abdomen, thereby facilitating emergence. This is especially obvious in Centerer, Hamaxia, Trophops and Erynnia. Some gregarious species also pupate in the host body, especially Palpostoma subsessilis, where as many as 28 develop in one beetle, and Cryptomeigenia theutis. In Eubiomyia calosomae, that may produce up to 16 individuals in a single Calosoma beetle, ca. 40% pupate within the host abdomen, and the rest do so externally in the space between the abdomen and the elytra. A few solitary species emerge from the host beetles for pupation, among which are Degeeria funebris Meig. and Minella chalybeata. Mature larvae of Chaetophleps emerge from the dead or dying chrysomelid host through an incision at the juncture of the head and thorax and pupate in the nearby soil (Clausen 1940/1962).

Tachinids that attack Hemiptera usually emerge from the host body to pupate in the soil. However, Trichopoda pennipes differs in that the summer generation pupates in the soil while puparia of overintering broods are found in the dead hosts. Tachinids attacking earwigs pupate outside the host body as do the few species with dipterous hosts. Mature larvae of Prosena, Dexia and other white grub parasitoids leave the body and pupate in the soil, ca. 2-4 cm. below the host remains (Clausen 1940/1962).

The normal habit of dipterous larvae, which enter the soil for pupation is to reverse their position so that when the puparium is, formed the head end points upward (Thompson 1910). Flies emerge by using the ptilinum and the backwardly directed spines of the body in order to force their way through the soil. Landis (1940) found that in Paradexodes, 46% of the pupae had their head end directed upward, but 29% had it directed downward, and 25% were laying horizontally.

When pupating within the skin of a larval host, gregarious tachinids often adopt a regular transverse position. Those pupating outside the host are usually scattered, or they may be closely packed but without any order. An exception is found in Sturmia cubaecola Jaenn. and S. protoparcis Tns. where several dozen puparia from one host are cemented together in an upright position in a disk-like mass (Greene 1921).

Temperature and humidity changes affect tachinid pupae variably. Those pupating in soil are protected from sudden changes and are not able to tolerate prolonged exposure to temperatures or humidities that differ a lot from those experienced under natural conditions. Resistant forms are found in those that pupate above ground and pass winter as pupae. Such pupae withstand large fluctuations in both temperature and humidity. A maximum emergence is secured in culture from species pupating above the ground when they are held under comparatively low humidities, while with most soil-inhabiting forms an almost saturated atmosphere is most favorable. Maximum emergence from puparia of Parasetigena segragata Rond, was secured by Grösswald (1934) from material which had been stored at 7°C. and 100% RH. Experiments with W. quadripustulata resulted in 33.4% emergence at 7.1% RH and 100% at RH of 73-100% (Hefley 1928). The viability of puparia of both summer and winter broods is proportional to the atmospheric humidity up to 73%, above which it remains at 100% (Clausen 1940/1962). By contrast the emergence of host adults from healthy pupae that are found in the soil is inversely proportional to RH. Therefore, optimum conditions for the host are disadvantageous to the parasitoid and vice versa. Parasitoid pupal stage duration is also much influenced by changes in RH; at 17°C. the range was from 26.3 days at 7.1% to 15.7 days at 73%, and then lengthened to 20 days at 100%.

Larval & Pupal Respiration.-- Cutaneous respiration satisfies the oxygen requirements of tachinid larvae within the host body. They temporarily tap an air sac or tracheal branch of the host, or they establish a fixed source of exchange with the outside air through the host tracheal system or directly through the integument. The latter involves the formation of a respiratory funnel within which the caudal end of the parasitoid larva, having posterior spiracles, is enclosed. Cutaneous respiration is the only way in which young larvae embedded in muscular tissue, a ganglion or a salivary gland are able to secure their oxygen requirements. As they mature this source probably proves inadequate and there is a greater need and another source is secured.

Many tachinids that are free living in the host body have posterior spiracles equipped with sharp, heavily chitinized hooks, which are used to puncture an air sac or tracheal branch. This is especially common among the parasitoids of adult beetles, and permits free movement, for the connection is only temporary. Many of these larvae persist in a free state throughout the larval period, although the spiracular hooks may be lacking in the later instars. Spiracular hooks in Anetia spp. are present on the 1st instar larvae, while in Centeter spp. and Hamaxia incongrua they are found only in the 2nd instar. They are employed to puncture an air sac to provide for respiration in Centeter, whereas in Anetia piniariae and Compsilura concinnata they also serve to hold the larva in a definite feeding position.

The larval respiratory funnel, in which the posterior end of the body with the functional spiracles is fixed, is an adaptation of general occurrence in Tachinidae. However, it is found elsewhere in only a very few highly specialized parasitic species of the closely related Sarcophagidae, but in no other families of parasitic Diptera. The only case of the development of an apparently similar relationship elsewhere is found in the chalcidoid family Eucharidae, of which two species of Orasema are internal parasitoids of larvae (see Clausen, 1940 for diagrams). Funnels in Tachinidae may be integumentary in origin, giving direct access to the outside air, and formed at the point of entry of the young larvae or at some other point by a larva that has already passed a period of free live in the host body. Or it may be of tracheal origin, usually arising on one of the main lateral trunks or on a spiracular stalk, but sometimes on a tracheal branch or air sac.

There are two classes of respiratory funnels, based on the manner of origin . Those which develop at the point of entry of the parasitoid larva into the host body are called primary and are always integumentary. Others arising as a result of the activities of the larva from within the body are secondary. The latter may be either intergumentary or tracheal in origin. However, there is not much difference in form or function of the two (Pantel 1910-1912).

When the respiratory connection is made with the formation of a funnel is variable. In a large number of species it takes place at the time of initial entry into the host, and funnel formation can frequently be distinguished within a few hours after penetration. Species that inhabit Lepidoptera and have this habit are Ernestia ampelus, Bonnetia comta, Ptychomyia remota, Phorocera agilis, Winthemia quadripustulata and Sturmia inconspicua Meig. Bigonicheta setipennis parasitic in earwigs and Dexia ventralis in scarab grubs similarly fix themselves at the point of entry. However, in hosts that are in the adult stage when attacked, the immediate formation of the funnel at the point of entry is rare.

Funnels formed at the point of entry are often relatively consistent in position for a given species. The planidium of Bonnetia comta usually penetrates the integument of the noctuid caterpillar host on the dorsum of the 1st thoracic segment. Bigonicheta does this in the intersegmental areas of the thorax. In these cases the location of the funnel is a matter of choice by the planidium, but in species that lay macrotype eggs it is determined by the location of the egg on the host body. In Winthemia and others that have a dehiscent egg, the larva remains with its posterior end in the eggshell while penetration is being attempted, and the funnel is thus found just in front of the egg.

Those species making the respiratory attachment after a period of free life in the host or of confinement in a definite organ, it may be either with the integument or with the tracheal system. There is an approximately equal division between the two points of attachment. Among species having caterpillar hosts, Actia diffidens Curr. is invariable found in an integumentary funnel on the mesothorax, while Sturmia nidicola and Leschenaultia exul are located in the posterior region of the abdomen (Prebble 1935). Gonia capitata and Archytas analis, which form the funnel only after the host has pupated, are located in a wing pad, while Zenillia libatrix chooses a point between any of the ventral plats of the head sclerites.

Of species making their attachment to the tracheal system, S. cilipes and Compsilura concinnata do so with the short stalk leading from a spiracle to the main longitudinal trunk, often close enough to the spiracle as to be considered connected with it. Most species make their attachment to the longitudinal trunk itself, but near the base of a spiracular stalk, and in caterpillars and coleopterous larvae a spiracular stalk in the 1st or 2nd abdominal segment is the preferred location (see Clausen, 1940 for diagram). The connection is with one of the smaller tracheal branches permeating the fat body of the Pyrausta larva in Zenillia roseanae B.B., whereas Eubiomyia calosomae forms the funnel on a tracheal branch in the beetle's metathorax. An unusual adaption was recorded by Matthey (1924) in Exotista larvarum L., where occasional larvae are found to have broken the longitudinal tracheal trunk of the host and to have used the broken end of the trunk itself as a funnel enveloping the posterior body. Gymnosoma rotundatum L., parasitic in Pentatomidae, makes its connection during the intermediate larval period, with one of the air sacs in the thorax (Clausen 1940/1962).

Species making respiratory connections with the tracheal system rarely reveal any external evidence of their presence until near the end of feeding. If the funnel is formed close to a spiracle, it may sometimes be visible. However, integumentary funnels are usually visible almost immediately after the connection is made, not only because of the actual perforation but by the dark funnel showing through the host integument if the latter is thin and not heavily sclerotized (Clausen 1940/1962).

The tracheal funnel represents a defensive reaction of the host to the irritation caused by making the perforation and to the persistence of the posterior end of the body of the parasitoid in the wound. Prell (1915) called it a wound-scab formation, which if true should cause it to have a constant makeup whether it arises from the integument or a trachea. Tothill et al. (1930) believed that the tracheal funnel of Compsilura consisted of an inner chitinous layer, a median hypodermal layer and an outer basement membrane.

There is much variation in size and form of the funnel, both among species and among different parasitoid instars. It is relatively short, flat and almost button-like in Thrixion (Pantel 1898), parasitic in Phasmidae. However, in most cases it is cup-like and closely envelops several posterior segments of the parasitoid body. It increases gradually in size with larval growth, and the basal portion may appear as a slender stalk. The funnel is usually much darkened in color, this being more pronounced near the point of attachment where the wall is thickest. In a few species the funnel is almost colorless. At times it has a distinctly segmented appearance due to a different in size and form to accommodate later instars (Clausen 1940/1962).

One or both molts may occur during the period of connection with the funnel. When this takes place, the exuviae remain in the form of a wrinkled lining on the inner wall of the funnel instead of being matted into its base, where they would interfere with respiration. The first exuviae are more often ejected from the mouth of the funnel (Baer 1921). This is possible in species forming the funnel at the point of entry, for the opening is often large. Anetia nigripes, shows an unusual kind of molting. The 1st instar larva, lying in the mid gut of the caterpillar host, forms a transverse split just above the caudal spiracles, after which the skin is cast over the head (Clausen 1940/62).

When the host molts, the union between the funnel and integument is severed, and thus no interruption of function or injury to the parasitoid larva occurs. However, sometimes it was found that the funnels of 1st instar larvae have been pulled out of the wound as the skin was cast; but the larvae remained in situ, with posterior portions of the body extruded from the wound. They had no problem with forming new funnels at the same spot.

In Compsilura, the initial perforation in the tracheal tube is made by the larva's mouth hooks (Tothill et al. 1930). In Thrixion and other species this is accomplished with the posterior end of the body (Pantel 1898). The 3rd instar larva of Paradexodes epilachnae in the larva of Epilachna varivestris Muls. frequently abandons the funnel that it has used and makes a final respiratory opening on the host's dorsum. To do this the kidney shaped spiracular plates are flexed against the inner side of the integument until it is cleared of muscular and other tissue. They are then pressed closely to the surface, and a partial vacuum is created, which causes the integument to come into close contact with the serrate edges of the plates, and the rasping action of the plates results in perforation (Landis 1940).

Usually a single respiratory attachment is made during the life of the larva, and its position remains fixed in the host body from the time of formation of the funnel until it is abandoned for feeding in the 3rd stage. However, Archytas analis, occupying a funnel in the wing pad of the host pupa, leaves this one in the early 3rd stage and makes a second breathing pore in the head or anterior thorax. In Paradexodes, the first funnel, arising from a spiracular stalk, is formed at the end of the first stage, and this is abandoned at the second molt and attachment made to another spiracular stalk. This second funnel is for a portion of the stage only, following which a third opening is made, this time in the integument, or the larva enters the intestine and lies at either end, using the natural openings of the host to respirate (Clausen 1940/62).

Initial funnel formation may occur right after entry into the host or during a later larval stage. When this is done is relatively constant for any given species. In most species it is during the first stage, but it occurs only after the first molt in Leschenaultia exul, Zenillia roseanae and Gonia capitata. A small group of species do not make a definite attachment or induce the formation of a respiratory funnel at any time. Among these are Anetia piniariae and A. nigripes, which are parasitic in caterpillars Hamaxia incongrua, Centerer cinerea and Palpostoma subsessilis in adult scarab grubs (Boas 1893), Trichopoda pennipes in adult Pentatomidae and Fortisia foeda Meig. in Lithobius. Dissection is usually not adequate to detect the larvae, but it is thought that most of them if not visible from the exterior, are somehow attached to the internal tracheal system.

Cutaneous respiration is normal in the first larval stage of many species, especially those inhabiting one of the host organs, and spiracles do not occur in Racodineura and a few other species. Tachinids that are free living as larvae secure their air during later stages largely by means of frequent temporary connections with an air sac or tracheal branch. The time when the connection with the respiratory funnel is broken varies with the species and the physical condition of the host. Clausen (1940) thought that the determining factor was probably whether or not the larva could reach its food source during this period. In Degeeria luctuosa Meig. on Haltica beetles, the connection is maintained until the end of the 3rd stage. This is possible because the attachment is to a tracheal trunk near an abdominal spiracle and all organs of the abdomen on which the larva feeds are within reach. On the other hand, the larva of Eubiomyia calosomae is attached to a tracheal branch in the metathoracic region. Thus, to complete feeding, the connection is broken very early in the 3rd stage, and thereafter it wanders free in the body of the live host. The persistence of the respiratory connection reaches an extreme in Siphona geniculata, where larvae sometimes maintain their connection with the funnel even after emergence from the host (Rennic & Sutherland 1920).

Other adaptations that relate to respiration are known. The young larva of A piniariae which inhabits the mid gut of the host, resumes feeding in springtime which is marked by the cutting of two openings in the intestinal wall. One of these is near the head for feeding purposes, and the caudal spiracles are thrust through the second opening. In this position many tracheal branches are held by the spiracular hooks, and the air supply is derived from them. The mature larvae of Ginglymyia acrirostris Tns. that are found in aquatic larvae of Elophila fulicalis Clem. (Lloyd 1919), extrude the large stalked spiracles completely through the dorsal integument of the host (see Clausen, 1940 for diagrams). Just before pupation, the host replaces its thin, web-like covering with a heavy, oval, roof-like structure, making up the cocoon, which has several openings at each end. Water passes freely beneath this covering, and a large bubble of air formes at its center. The forked spiracular structure of the parasitoid extends from the middle region of the host body into the bubble, thus securing for the mature larva and pupa an adequate air supply. Clausen (1940) speculated on the manner in which this adaptation may have arisen. Transitional forms are not known, and these presumably would not fulfill the requirements. It is probable that the parasitoid adapted concurrently with the host, for the latter is one of the few species that has adapted an aquatic mode of life.

Another respiratory adaptation is in the 1st instar larva of Plagia trepida (Thompson 1915b), which bears a well-developed anal vesicle in the form of a large plate occupying the greater portion of the ventral surface of the last segment and in the center of which the anal opening is located. This vesicle is formed of large cells appear like the striate wall of the middle intestine. This larva normally lies in a dead host muscle, and the great development of the vesicle is in response to the need for oxygen. Other larvae confined to host organs are bathed in abundant secretions, resulting partly from the hypertrophy of the parts involved, and thus the need for respiratory adaptations is not great. Pantel (1898) recognized a respiratory purpose for this structure in Tachinidae. It has not been found in any genus of the family except Plagia but is present in bilobed form in Conopidae. The structure appears to be homologous with the anal vesicle of some parasitic Hymenoptera, where it is highly developed (Clausen 1940/62).

Information on tachinid pupal respiration reveals that the spiracles of the puparium, representing those of the 3rd larval exuvia, are not utilized directly. There is no connection with them. Species having extruded prothoracic cornicles may obtain their air supply from outside, but the extensive development of the internal spiracles in all species indicates that these are the principal organs serving the purpose. Where the cornicles do not penetrate the puparium, the larval spiracles may remain sufficiently open to permit the passage of air into the general cavity of the puparium. Snodgrass (1924) concluded that the anterior larval spiracles of Rhagoletis pomonella Walsh provide the channels through which air enters the puparial chamber. At molting, the tracheal branch or stalk that leads to the spiracle is withdrawn from the body, and it remains distended, even though broken, on the inside of the puparial shell. Clausen (1940) thought that the same adaptation may occur in Tachinidae.

Adult Emergence.-- As true for many Diptera, emergence from the puparium is by an expansion of the ptilinum which forces off the two parts of the operculum. The fly then works its way out. Soil penetration or other material is accomplished by alternate expansion and retraction of the ptilinum, aided by the backwardly directed spiens of the head and thorax, and by use of the legs. The way adult flies emerge from puparia that are still enclosed within host cocoons or burrows has caused speculation. Where the puparium is contained within the dead host skin, this is usually thoroughly dried and adheres closely to the outer surface of the puparium. It is broken by the outward pressure of the expanding ptilinum. In dead beetles on or in the soil, the intersegmental membranes of the abdomen are so thin and weakened by decay that the fly has no difficulty in forcing the segments apart (Clausen 1940/62).

Cocoons of Tenthredinidae and Lepidoptera ore different, however, for the walls are composed of varying quantities of silk and other material. This makes them too tough and heavy to be broken by pressure alone. The manner in which emergence is accomplished by Diplostichus janitrix Htg. from the cocoons of Diprion pini L. differs. After feeding in the body of the host, the parasitoid larva emerges and prepares for a later exit by cutting a circular groove with its mouth hooks around the inner wall of one end of the cocoon (Robbins 1927). When the fly emerges from the puparium, it forces this cap away, which is only lightly attached by the outer silken layer. Cocoons containing unparasitized sawfly larvae do not show this inner groove, and adult sawfly emergence is effected by cutting away a cap, which is a bit larger than that removed by Diplostichus, with the mandibles (Prell 1924) . De Fluiter (1932, 1933) found that the parasitoid may emerge from either end of the host cocoon, thereby establishing definitely that the cap structure is related only to the parasitoid. However, most do emerge from the anterior end.

Even Lepidoptera with very heavy walled cocoons sustain tachinid parasitization. Chaetexorista javana seems to have no difficulty in emerging from the egg-like cocoons of Monema flavescens, that are very hard and tough and can be cut only with a sharp knife. The cap, which is usually forced off by the host itself, is similarly removed by the parasitoid. The circular line of union of the cap with the rest of the cocoon is relatively weak, and an oral secretion may serve to soften the lining so that the pressure the fly is able to exert causes the braking away. Some Lepidoptera that spin silk cocoons leave a definite opening or loosely woven area at the anterior end of the cocoon, through which the parasitoid larva or adult is able to emerge without trouble. Winthemia datanae Tns. parasitic on Samia cecropia shows that sometimes there is difficulty in emergence with some hosts. The larvae seem unable to penetrate the cocoon wall, and they die from desiccation even before pupation.

Life Cycle

Tachinids have life cycles ranging from 10 days in Metagonistylum minense in the tropics to a full year for others. Most multibrooded species complete the cycle in 3-4 weeks in summer. The incubation period is vary variable, for many species are larviparous or ovoviviparous, and the microtype eggs, though fully incubated at oviposition, must be eaten by the host before they can hatch. In some species they remain viable for as long as 2 1/2 months. The unincubated macrotype eggs hatch in 2-4 days. Free living larvae, which await a host or must search for it, are able to survive for up to 10 days in the case of Ernestia ampelus.

Summer broods have short larval feeding periods, in some cases being only 4-6 days as in W. quadripustulata, 7 days in Lixophaga, and 6-8 days in Centeter spp. However, 12-16 days is more frequent for feeding of summer broods. In Bigonicheta the period is nevertheless variable, ranging from 21-90 days, the duration depending on the amount of food available in the individual hosts (Mote et al. 1931). In contrast, the developmental period of Paradexodes epilachnae larvae is not affected in this way, and the duration of feeding is the same, whether in young or mature larvae, prepupae or pupae (Clausen 1940/62).

Summer developing broods have a pupal period ranging from 5-7 days in Hyalomya aldrichi to 25-30 days in Sturmia nidicola, with 8-12 days as a general average for the family. Dowden (1933) observed that the pupal stage of female Zenillia libatrix is 1-2 days longer than for males, which Clausen (1940) thought was probably a consistent difference for the sexes generally.

Tachinids hibernate mainly in the pupal stage, although there are some departures from this. The next most common behavior is in an early larval stage within the living host larva or pupa, which occurs commonly in lepidopterous hosts. Of the Dexiinae, the single brooded Prosena sibirita is in the first stage within the host grub, while Dexia ventralis and Microphthalma are in the second stage during that period. Species which attack adult beetles that persist through the winter are usually in the early larval stage within the body. Species of Chaetophleps, Degeeria, Erynnia, Eubiomyia, Erynnia, and Stomatomyia show this behavior. Several species of Winthemia hibernate in the mature larval stage in the soil. Zenillia pexops (Wardle 1914) that develops in sawfly larvae, overwinters in its mature larval stage within the dried host integument, and a few other species have this behavior. Hibernation of a large number of North American species were recorded by Schaffner & Griswold (1934). North American tachinids do not overwinter as adults in temperate regions, although in Japan and Korea Hamaxia incongrua apparently does so (Clausen 1940/62).

Hibernation in the pupal stage is most often the result of low temperatures, but this is not the only factor responsible. A large number of species pass into a state of diapause irrespective of prevailing temperatures. Sometimes this can be broken prematurely by subjecting pupae to a period of pronounced cold. Thompson (1928) found that Bigonicheta setipennis appears to have two types of puparia. In one the fly develops and emerges rapidly, while the second persists until the following spring, when the adult flies appear. Emergence of the first form cannot be indefinitely retarded by low temperatures, nor can that of the second type be forced by high temperatures. According to Pantel, the two types of pupae ar found most often in species restricted to a single host or to a very small number of hosts and is an adaptation that increases the opportunities of the brood for finding hosts, inasmuch as a portion emerge during autumn and the rest in springtime.

The generations per year ranges from one in a large number of species of temperate regions to 8-10 in Lixophaga (Scaramuzza 1930) and Prosopaea indica Curr. in the tropics. The number in Metagonistylum is probably much higher, for host stages suitable for parasitization are available throughout the year, and there is no hibernation or estivation (Clausen 1940/62). Species parasitic in adult chrysomelid beetles and Pentatomidae often have a much greater number of generations per year than do their hosts. This is especially true in those which overwinter in an early larval stage in their hibernating hosts. Degeeria luctuosa has this habit and completes its development early enough in springtime to produce an additional generation on the hibernating brood of beetles. Chaetophleps setosa usually passes through 5 generations annually under temperatures prevailing in Illinois. Clytiomyia helluo F. completes 4 generations, and sometimes 5-6 in Eurygaster, which has an annual cycle (Jourdan 1935).

For most species the seasonal cycle of the parasitoid is correlated with that of the preferred host, but in quite a few cases thee is an obligatory alternation of hosts. This is true especially among multibrooded species that pass winter in the early larval stages in the bodies of the live hosts. Therefore, a species having this habit would be unable to exist solely on a host species that hibernates in the egg stage. Compsilura concinnata is one of these parasitoids. The gypsy moth is a favored host, but the parasitoid would be unable to persist on it alone. An exceptional adaptation to bridge the winter has been developed in Erynnia nitida, a parasitoid of the elm leaf beetle. The two summer generations develop in the larvae whereas the overwintering brood develops in the hibernating beetles (Silvestri 1910, Clausen 1940).

In Dexia ventralis the seasonal is of interest because of an unusual alternation of hosts (Clausen et al. 1927). This species is a solitary internal parasitoid of some scarabs in Asia. In Korea there are normally three generations annually. The larvae produced by the spring brood of females attack the grubs of Popillia spp., the 2nd generation occurs in Serica spp. and the overwintering generation in miridiba koreana N. & K. Thus, the successive generations during the season develop each upon a different scarab subfamily. However, part of the population has only 2 generations annually, and in this case the overwintering host is P. castanoptera Hope rather than Miridiba. Popillia grubs pupate ca. one month later than those of Miridiba, and Dexia development is delayed also, so that the following generation is directly on Serica, the intervening one on Popillia being omitted. The parasitoid thus shows considerable adaptability, but the complexity of the annual cycle and the host preferences suggest that the species requires at least one alternate host to become numerically abundant. Adults appear in the field around the same time as adult beetles of the host brood on which they had developed. Therefore, with a single host having a strictly annual cycle, only 1st instar grubs would be available for attack for a prolonged period. This makes it difficult for the planidia to locate them in soil, and they are not in a suitable physical condition for extensive parasitization. Clausen (1940) therefore questioned whether the species would be able to maintain itself on a single host species unless the latter had at least a partial 2-year cycle, as a result of which grubs in a stage of development suitable for parasitization would be available at all times.

Sex Ratio & Parthenogenesis.-- Sexes are not superficially easily distinguished in Tachinidae. Reports from New Zealand on Hystricina lupina Swed. by E. S. Gourlay (Clausen 1940) shows a sex ratio of 4:1 in favor of females. In Paradexodes the ratio is ca. 50:50, with a tendency toward proportionally more females at higher temperatures.

Parthenogenetic reproduction was reported in G. puellae by Nishikawa (1930). He found that uninseminated females lay few eggs, usually <10, in contrast to several thousand laid by mated females, and that these unfertilized produced normal larva when ingested by silkworms. Unmated Ptychomyia remota females sometimes lay eggs, though these do not hatch. Compsilura concinnata may puncture the host caterpillar, as in normal larviposition, but no eggs or larvae are laid. it seems to be the general habit to retain the eggs in the ovarioles until after mating. If unmated, the eggs are broken down and resorbed. However, Webber (1932) found that the eggs of unmated females of Carcelia laxifrons Vill, Phorocera agilis, etc. descent into the uterus and may be laid, whereas this does not occur in Sturmia inconspicua.

There are wide fluctuations in relative abundance of the two sexes during different times. In some multibrooded species, females predominate during autumn. Allen (1926) concluded that a definite sexual segregation took place in Archytas analis in the field at times. It is believed that most females migrate to new areas soon after mating, in case suitable hosts are not abundant locally, leaving males to mate with whatever females may emerge later. The dispersal tendencies of females seems lacking in the males. Allen (1925) noted the occurrence of large gyrating swarms of Winthemia quadripustulata, consisting entirely of males, in localities lacking in host infestations and food sources.

Parasitism Effects on Hosts.-- Tachinid parasitism effects adult hosts variably, depending on their age at the time of attack and the rate of the parasitoid's larval development. During early larval stages, feeding is principally on body fluids and fat bodies, which inhibits development or causes atrophy of the reproductive organs of the host, a condition of parasitic castration similar to that in Hymenoptera. In C. cinerea attacking Popillia beetles, parasitization usually occurs very soon after the beetles emerge, and this combined with the lapse of only 6 days from egg laying to host death, ensures that little or no oviposition by the latter will occur. Overwintering beetles of Galerucella that harbor the young larvae of Erynnia nitida are killed in springtime soon after feeding starts, and no eggs are deposited. In Eubiomyia calosomae, with several generations per year and which attacks Calosoma beetles having an adult life of 2 or more years, the summer broods kill the host in 9-12 days. Overwintering parasitized beetles die within a few days after the beginning of activity in springtime. The effect on the host population is much less than in the cases previously noted, because a large portion of the reproductive potential may already have been realized (Clausen 1940/62).

When parasitizing Orthoptera, parasitoid larval feeding seems to be restricted mostly to the blood, which results primarily in a reduction of the fat bodies. Mature larvae of Ceracia ajarifrons Ald. have been taken from several locusts that had oviposited during the preceding 12 hrs (Clausen 1940). Pantel (1898) had shown that the degeneration of the reproductive system of female Phasmidae as a consequence of parasitism by Thrixion was only temporary and that in some cases these females were again able to produce and deposit eggs after the parasitoid larvae had left their bodies. In the European earwig, which is frequently attacked in the late nymphal instars, there is a partial atrophy of reproductive organs, and death usually occurs before eggs can be laid.

Clausen (1940) noted that in the case of parasitism of Hemiptera, the false chinch bug, Nysius ericeae Schill. is rarely able to deposit eggs when parasitized by Hyalomya aldrichi, but adult females of Anasa tristis DeG., that contain larvae of Trichopoda pennipes oviposit, seemingly without serious interruption, until the final larval stage of the parasitoid is attained. The parasitized individuals of the autumn brood do not reach sexual maturity, however (Worthley 1924).

For detailed descriptions of immature stages of Tachinidae, please see Clausen (1940/62).

References: Please refer to <biology.ref.htm>, [Additional references may be found at: MELVYL Library]

Aguilar, J. D. 1957. Revision des Voriini de l'ancien Monde (Dipt. Tachinidae). Ann. Epiphytes 3: 235-70.

Allen, H. W. 1925. Biology of the red-tailed tachina-fly, Winthemia quadripustulata Fabr. Miss. Agr. Expt. Sta. Tech. Bull. 12. 32 p.

Blanchard, E. E. 1956. Parapolios grioti, nuevo Actiino útil argentino (Dipt.). Rev. Soc. Ent. Argent. 19: 45-6.

Burrell, R. W. 1935. Notes on the habits of certain Australian Thynnidae. J. NY. Ent. Soc. 43: 19-28.

Clausen, C. P. 1940/1962. Entomophagous Insects. McGraw-Hill Book Co., Inc., NY. & London. 688 p. [reprinted 1962 by Hafner Publ. Co., NY.].

Cole, F. R. 1969. The Flies of Western North America. Univ. Calif. Press, Berkeley & Los Angeles. 693 p.

Dupuis, C. 1956 (1953). Variations convergentes ou comparables de certains caractères des tachinaires, notamment des Phasiinae (Dipt. Larvaevoridae): leur signification taxonomique différente selon les lignées. Proc. Intern. Cong. Zool. 14: 474-76.

Dupuis, C. 1957a. Contributions à l'étude des Phasiinae cimicophages (Diptera Larvaevoridae). XIX. Etude de Cylindromyia pilipes (Lw.) s. str. Cahiers Nat. (n.s.) 13: 9-22.

Dupuis, C. 1957b. Idem. XXI. Notes taxonomiques et biologiques diverses. Cahiers Nat. (n.s.) 13: 71-9.

Hertig, B. 1940. Monog. Z. angew. Ent. 16: 1-188.

Herting, B. 1957. Die Raupenfliegen (Tachiniden) Westfalens und des Emslandes. Abh. Landesmus. Naturkde. Münster Westf., Jhg. 19. 40 p.

Landis, B. J. 1940. Paradexodes epilachnae, a tachinid parasite of the Mexican bean beetle. U. S. Dept. Agr. Tech. Bull. 721. 31 p.

McLeod, J. H., B. M. McGugan & H. C. Coppel. 1962. Commonwealth Inst. Biol. Contr., Tech. Comm. 2: 1-216.

Mesnil, L. P. 1955. Contributions à l'étude de la faune entomologique du Ruanda-Urundi (Mission P. Basilewsky, 1953). Diptera Tachinidae. Ann. Mus. R. Congo Belge 40: 359-67.

Mesnil, L. P. 1956. Trois nouveaux Tachinaires d'Afrique (Dipt. Tachinidae). Entomophaga 1: 76-80.

Mesnil, L. P. 1957. Nouveaux Tachinaires d'Orient. Mém. Soc. R. Ent. Belgique 28: 1-80.

Mesnil, L. P. 1962. Die Fliegen der Palaearktischen Region 8: 753-848.

Paramonov, S. J. 1957. Notes on Australian Diptera. XXIII. Notes on some Australian Ameniini (Dipt., Tachinidae). Ann. & Mag. Nat. Hist. 10: 52-62.

Peris, S. V. 1956. Notas sobre Acemyiini (Dipt. Tachinidae). Graellsia 14: 1-7.

Reinhard, H. J. 1956. A synopsis of the tachinid genus Leucostoma (Diptera). J. Kan. Ent. Soc. 29: 155-68.

Reinhard, H. J. 1957. New American muscoid Diptera (Sarcophagidae, TAchinidae). Ent. News 68: 99-111.

Tothill, J. D., T. H. C. Taylor & R. W. Paine. 1930. The coconut moth in Fiji. London Publ. 269 p.

Townsend, C. H. T. 1934-1939. Manual of myiology in twelve parts. Pt. 3, 4. (8 vols.), Sao Paulo, Brazil.

Zimin, L. S. 1957. Révision de la sous-tribus Ernestiina (Diptera, Larvaevoridae) de la fauna paléarctique. I. Rev. Ent. URSS 36: 501-37.