General literature on larval morphology of two winged flies is for example, Hennig, 1973, Ferrar, 1987, Smith, 1989, Teskey, 1981, Courtney et al., 2000.

Although the imagines are usually in the focus of taxonomists, the larval Agromyzidae cause the major damage. In several cases they might be found before the adults can be reared or collected.
Therefore, it is useful to gather as much information as possible about the larvae. Although they normally do not yield much taxonomic information, in certain cases even the identification of the species infesting a given host plant may be possible on basis of the larvae.
Moreover, even if the taxonomic opportunities were left aside, the study of larval behaviour and morphology provides valuable insights into the lifestyle of Agromyzidae and their relation to plants.

As in other cyclorrhaphan flies, the larvae of Agromyzidae are soft-bodied legless maggots with worm-like movements. Head and mouthparts are compared to the 'standard insect head' extremely modified ( Ag oryzae larva.pct). Cyclorrhapha are easily identified by the absence of a head capsule and of biting mandibles. Instead, they possess a mandibular complex consisting of two closely connected parallel mandibles moving in a vertical plane. Thus, digging, scratching or sweeping are the only motions this kind of mandibles allow. Since they are operating in dorsoventral direction the leaf-mining larva normally lies on its side. The mandibles are connected with some internal sclerites, constituting the cephalopharyngeal skeleton (Teskey, 1981, Courtney et al., 2000) ( Chr milii ceph nat.pct). In spite of their apparent simplicity the species of cyclorrhaphan Diptera successfully established themselves in various habitats. They can be saprophagous in the soil, highly specialized filter feeders in aquatic environments, parasitoids, phytophages and even predators. In contrast, Agromyzidae are always plant feeders.

Apart from agromyzids, there are several members of other dipteran families with leaf-mining and other phytophagous life histories (e.g. Scathophagidae, Anthomyiidae, Tephritidae, Drosophilidae). This CD-ROM deals with Agromyzidae only.
An excellent diagnostic feature to distinguish agromyzids from other plant inhabiting maggots is the reduction or absence of the dorsal bridge of the cephalopharyngeal skeleton in Agromyzidae agrom ceph aabiens.pct. In non-agromyzid dipteran leaf miners and in other Diptera the dorsal bridge is usually large and prominent (e.g. on this picture of a particle feeder: drosophila ceph.pct). Furthermore, the shape of the mandibular complex of Agromyzidae and other Diptera is distinctive. The latter normally have got well developed abductor apodemes, whereas agromyzids do not possess these structures. In leaf-mining Agromyzidae the mandibular complex is normally higher than long.
Another visible character in most agromyzid larvae lies in the dorsal arm of the basal part of the cephalopharyngeal skeleton. It seems to be divided into two different parts due to the development of a window within the basal part. However, since the lower part frequently is reduced or hardly visible, the character is of limited use for recognition of the family Agromyzidae (but see descriptions of Subfamily Agromyzinae and Subfamily Phytomyzinae, agr-phy-larva.pct). For more information on the terminology of the cephalopharyngeal skeleton of Agromyzidae see agrom ceph aabiens full.pct.

Both, Agromyzidae and other mining Diptera often have several teeth (mouth hooks) at the tip of each mandible. Although the number can vary, most larvae have four mouth hooks in total, two on each mandible. Higher numbers are present mainly within Japanagromyza ( Japanag. Larvae.pct) and, to a lesser extent also in Agromyza. In several stem-mining species a reduction of mainly the ventral mouth hooks has taken place. In those cases the arrangement of more or less equal sized mouth hooks (e.g. Chr milii ceph nat.pct) was replaced by one or two strong and enlarged hooks, which are capable to cope with the tougher plant tissue within stems ( Mel lappae ceph nat.pct, Phb cambii Larva SEM1.pct). Corresponding to the larval environment of stem miners (borers), where tubes rather than flat mines are produced, the mandibles are hardly laterally flattened. Accordingly, the body shape of non-leaf-mining species is more rounded, in the case of stem miners long and cylindrical and in seed dwellers and gall producers sac-like.

A further characteristic of the larval mandibles is their widespread asymmetry. From the considerable deviation in the total size of both mandibles ensue different positions of the mouth hooks. The pattern of asymmetry is extraordinarily constant throughout the whole family. In all species of the subfamily Phytomyzinae the right mandible is longer than the left; the left one is larger in Ophiomyiini and Japanagromyza. Symmetrical or nearly symmetrical mandibles occur in Agromyza, Nemorimyza and occasionally in small numbers in some phytomyzine genera. In the following series of scanning electron micrographs examples are shown:

A albipennis Larva SEM.pct
Cer denticornis Larva SEM2.pct
Chr nigra Larva SEM1.pct
Lir bryoniae Larva SEM2.pct
Phb cambii Larva SEM1.pct
O galii Larva SEM1.pct

As the general shape of the mouth hooks, also the asymmetry can be seen as an adaptation to the life in a leaf mine (Dempewolf, 2001). This conclusion was based on the observation that those species with symmetrical mouth hooks (mainly in Agromyza) usually produce full depth mines, leaving aside only some parts of vascular bundles. Leaf miners with asymmetrical mouth hooks are found to cause thinner mines and are restricted to certain cell layers.
That may be illustrated by a picture of Cerodontha denticornis ( Cer denticornis Larva SEM2.pct). The cephalic region of a leaf-mining agromyzid larva normally is laterally flattened and in apical direction gradually narrowing. This shape is intensified by the position of the mouth hooks, especially by the one in the highest position. Moreover, the single apical hook also facilitates the maggot's ability to handle tough vascular bundles because it can concentrate its strength on one mouth hook. That may be of some significance because even if the larvae mainly feed on the mesophyll they cannot always avoid to cross vascular bundles and other lignified tissues (see Scheirs et al., 1997, Scheirs et al., 2001). The asymmetry of mandibles is also present in non-leaf-mining agromyzids but according to my own phylogenetic research it is a legacy from leaf-mining ancestors. The most ancient feeding habit in Agromyzidae is leaf-mining (Dempewolf, 2001, see Evolutionary History).

Above the mandibles some structures of the facial mask are located ( Mel lappae ceph nat.pct). They are the main sensual organs of the larva A albipennis Larva1 SEM.pct. For more information on larvae of cyclorrhaphan Diptera see Courtney et al., 2000, Ferrar, 1987, Teskey, 1981. Some notes on the bionomics can be found in the part on Diptera within the Order Diptera in the Higher Taxa module.

The segmental borders are normally covered with small denticles, which enable the larva to creep in the mine and exert pressure on the leaf tissue and thus on the operating mouth hooks. These areas are called locomotion welts or creeping welts:

A albipennis larva4 SEM.pct
Lir bryoniae larva SEM4.pct
Phb cambii Larva SEM2.pct
Cer denticornis Larva SEM4.pct
O alliariae Larva SEM2.pct

They are mainly developed at the sides of agromyzid larvae, but several groups of species have locomotion welts surrounding the whole body. The size and shape of the locomotion welts and their denticles depend on the larval environment. Denticles are smaller in stem mining species whereas the largest denticles are found in leaf miners of soft and ephemeral leaves, which exert the weakest pressure on the larval movements.

The larval spiracles often contain characters that can be used to identify species or species groups. Agromyzid larvae have got two pairs of openings of the respiratory system. They are situated on the first thoracic and last abdominal segments respectively. Both anterior and posterior spiracles are usually strongly projecting above the larval body. The projecting parts of the spiracles consist of some spiracular bulbs whose number, shape and distribution can vary considerably among the species. Each spiracle can have 3-60 separate bulbs. A close view reveals that each bulb show a minute central spiracular opening responsible for gas exchange. The openings are dot-shaped in the anterior and slit-shaped in the posterior spiracles.
Within the selection of species treated in the CD-ROM the most primitive condition of posterior spiracles is present in Agromyza albipennis ( A albipennis Larva3 SEM.pct). Here, the three large bulbs are of equal size, the opening slits are flanked by spiracular processes, which are reduced in any other agromyzid genera. Other varieties are shown on the following pictures:

Li pusilla p spiracle.pct
Cer denticornis Larva SEM5.pct
Cer incisa Larva SEM.pct
Chr milii Larva SEM3.pct
Mel lappae Larva1 SEM.pct

Between the spiracular openings and the trachea lies the felt chamber, functioning as filter and as protection of the tracheae (Jones and Kim, 1988). Each opening is surrounded by a bulb-like terminal cavity of the felt chamber, they are frequently referred to as bulbs or spiracular bulbs in the literature and also in the texts of this CD-ROM.

The puparium of higher Diptera consists of the sclerotized integument of the third instar larva enclosing the pupa ( L bryoniae puparia.pct). The true pupa within is exarate.
Several former larval characters such as the spiracles and parts of the locomotion welts are still visible on the surface of the puparium (e.g. Tropicomyia Puparium.pct). Yet the position of the spiracles changes during the pupariation, especially the anterior ones move to the tip of the body. The facial mask and sometimes parts of the segmental margins are retracted. Even the larval cephalopharyngeal skeleton is concealed under the puparial cuticle and can be dissected. Although it is usually somewhat deformed (e.g. as in the figure: Ag albipennis ceph.pct) it can be used for identification of agromyzid subgroups.
The color of the puparium normally is brownish, sometimes yellowish transparent or black. Within species, sometimes striking color variation due to seasonal dimorphism was observed (Nowakowski, 1962).
The anterior part of the puparium forms a cap separated from the remaining part by puparial cleavage lines. It can be opened by the emerging adult fly by expanding its ptilinum (see External adult morphology). Usually the cap splits in two parts with the upper one containing the former anterior spiracles and the lower part where the former larval cephalopharyngeal skeleton is attached to. Within the Tribe Ophiomyiini (Ophiomyia, Melanagromyza, Hexomyza, Kleinschmidtimyia, Tropicomyia) the upper part is further divided by a longitudinal suture puparia Tropicomyia/phytomy.pct.

Agromyzid larvae have always three different stages, which differ in size and shape. Some general differences are only briefly discussed because the last larval instar is taxonomically the most important. Some more detailed descriptions are available in (Dempewolf, 2001). The second instar widely resembles the third one, it is only smaller and sometimes more slender. Occasionally the number or shape of the mouth hooks can be different (e.g. Agromyza albipennis and probably also some other grass mining Agromyza). These differences normally cause no practical problems if the population density is high enough to allow the comparison of several specimens and mines.
The very tiny first larval instars are more characteristic than the second instars. Most sense organs, the locomotion welts and spiracles appear atrophied. The anterior spiracles in particular are under a light microscope virtually invisible and for a long time regarded as absent in all cyclorrhaphous Diptera. Recently two minute holes on the surface, representing a single opening for each spiracle, were discovered using scanning electron microscopy (SEM) (Kitching, 1976, Dempewolf, 2001).

The shape of the cephalopharyngeal skeleton of first larval instars hardly resembles the later ones, for example the subdivision of the upper arm of the basal sclerite is not developed. Similarly, the lateral shape of the mandibular complex appears very different. A closer view using the SEM reveals in several groups the arrangement of mouth hooks to be similar to the one in later instars. However, very peculiar forms occur in several groups. For example in Chromatomyia and most Phytomyza the mouth hooks form a tube, through which probably the food is ingested. In two cases in Agromyza I found first instar larvae with asymmetrical mouth hooks that are in later instars replaced by symmetrical mandibles (more information in Dempewolf, 2001).