Lynceus simiaefacies, Harding, 1941

Fryer, Geoffrey & Boxshall, Geoffrey, 2009, The feeding mechanisms of Lynceus (Crustacea: Branchiopoda: Laevicaudata), with special reference to L. simiaefacies Harding, Zoological Journal of the Linnean Society 155 (3), pp. 513-541 : 519-534

L. SIMIAEFACIES AND ITS

FEEDING MECHANISM

Lynceus simiaefacies , which attains a length of about 5.0 mm, is known only from the Jebel Jihaf, Aden (now part of Yemen). It was found at an altitude of 7100 ft (2164 m) in a temporary pool with a mud bottom and earth sides, fed only by rain. No details of the pool are available except that it was inhabited also by the anostracan Streptocephalus neumanni Thiele, 1904 and that it was completely dry less than a month after the animals were collected. It is suspected that, unlike some habitats frequented by other species of Lynceus , no aquatic vegetation was present. Subsequent collecting of large branchiopods in various parts of the Arabian peninsula has revealed no additional sites, and Thiéry (1996) tentatively suggests that it is endemic to south-western Arabia. The only known material is preserved at The Natural History Museum, London.

Its general appearance and some of its appendages are illustrated by Harding (1941). Although its near spherical form, typical of Lynceus , offers few distinguishing characters, the striking profile of its large head, as seen from in front ( Fig. 13D ) is very distinctive. Protected by thick cuticle, the broad anteroventral region – the rostrum – with its strengthening ribs and ridges suggests that it may be used for pushing into the bottom deposits as the animal forages. In the male the rostrum is square-cut ventrally. The two dark areas of Figure 13D , which appear superficially like eyes, are the sensory fields where, from areas of thin cuticle, arise groups of short sensory setae.

Although Harding gives particulars of the male claspers and outline figures of the second trunk limbs of both sexes, no details of trunk limb armature have ever been recorded. This armature is very striking and is a key feature relating to the animal’s way of life. The gut contents of preserved specimens consist largely of what can be described as fine sand and organic debris. The only recognizable objects seen were a few pennate diatoms. Clearly much inorganic material is scraped from the bottom and L. simiaefacies probably relies on a rapid passage of material through the gut, only a small proportion of which is readily digested.

As in other examined species of the genus, as well as an array of long natatory setae, the multisegmented exopodite of each antenna, but in the female not the endopodite, bears a number of stout spines on its anterior margin. These, one per segment, and lacking on a few segments, taper gradually to a sharp point. There is a more slender spine at the tip of the distal segment. If the exopodite is flexed, these spines project. They possibly serve to obtain purchase, or to anchor the animal, as it employs its scraping limbs, but this is only conjecture. The male bears spines on both rami, which does not negate this suggestion but implies another function in that sex.

In basic form the trunk limbs are similar to those of L. gracilicornis and other species, and it is reasonable to assume that movements of these appendages are essentially the same in all species of the genus, and that observations on L. gracilicornis are a reliable guide to locomotion and food handling in L. simiaefacies , but details can only be checked on living animals. Food is clearly scraped from surfaces by robust scrapers borne on endites 4 and 5 and on the endopodite of the trunk limbs whose presence is an outstanding feature of this species, and is clearly correlated with the nature of the food collected.

Seen from the mid-line of a bisected female ( Fig. 1A ) to which the initial description refers, a set of trunk limbs presents a dense array of spines and setae. Such a view shows the gradation in size of the limbs, largest anteriorly; how the four posterior limbs make up only a small part of the ensemble; how the distal (ventral) endites of the limbs, which have access to the surfaces from which food is collected, bear a numerous array of scraping spines; how, more proximally, the median face of the ensemble is largely covered by a screen of long slender setae (the screening setae), through which numerous postero-dorsally directed spines of the proximal endites can project into the narrow median cage between opposed sets of limbs; and how the gnathobases of the limbs lie in line, one behind the other, along the length of the straight food groove.

Trunk limb armature includes stout spines of several kinds. The heaviest, except for that of the gnathobases, is borne distally except on the last four pairs, which bear stout spines or scrapers on all endites. These scrapers ( Figs 2A–C , 3A, B , 6A ) resemble those whose function is well understood that are employed by a wide range of crustaceans, including many branchiopods, and by certain aquatic insect larvae in which they have evolved convergently. Their role in L. simiaefacies is easily appreciated. The action of a scraper as it is dragged over a surface is self-evident. Examples where the animal was fortuitously fixed when scrapers were dealing with scraped up material are seen in Figure 2B . The scrapers are accompanied by a number of sensory setae.

The two distal endites (4 and 5) and the endopodite of the anterior trunk limbs are elongate, while endites 2 and 3, of different orientation, are broad ( Fig. 4 ). Proportions change gradually along the series. As the limbs diminish in size posteriorly, endites 4 and 5 especially, and the endopodite, become shorter, so that in the last four pairs, which are much smaller than the anterior limbs, all endites tend to be not greatly dissimilar in size and shape. The endopodite remains somewhat longer than the endites.

On trunk limb 1 endites 4 and 5 are armed with very fine scrapers, and the long, slender endopodite which over-reaches them is armed distally with long setae that bear a row of close-set denticles so fine as to be scarcely detectable by light microscopy. Endites 4 and 5 are each armed also with four grid setae. The more proximal endites carry no scrapers, but are armed with an array of long, slender, backwardly directed spines as well as with a row of grid setae.

Trunk limb 2 is similar to trunk limb 1 but its endopodite bears fine scrapers. Endites 2 and 3 each bear a row of long grid setae that are accompanied by long simple spines.

Trunk limbs 3–8 are similar to their anterior companions, but the scrapers of endites 4 and 5 and the endopodite are robust ( Figs 2A–C , 3A, B , 6A ). On endites 2 and, especially, 3 the row of long grid setae is accompanied by spines that, long and slender on trunk limb 2, become progressively shorter and stouter on limbs more posterior in the series. These are variously armed with spinules, often arranged in two rows in the same plane, which converts them into broad structures suitable for pushing material posteriorly and towards the food groove. Several such spines, which indicate their diversity on endite 3 in the middle of the series, are seen beneath typical scrapers in Figure 3B .

Endites 4 and 5 and the endopodite of limbs 9–12 are provided with an array of spines that are in general smaller than those of more anterior limbs, and here they assume a diversity of form that is best appreciated from SEM photographs. Whereas some spines, such as those located distally and dorsally on endite 5 of trunk limb 9, and in a similar position on trunk limb 10 ( Fig. 8A, B ), are basically scrapers, others are modified in other ways and often have two rows of spinules so that the spine can function as a scoop rather than as a scraper. Such spines are well displayed in Figure 8B , in which limbs 9 and 10 have fortuitously twisted in a clockwise direction to reveal their complex armature in face view. That is, their endites are here seen ‘edge-on’.

The scrapers of L. simiaefacies are complex structures whose general form is shown in Figures 2A–C , 3A, B and 6A . Their function, which can be deduced with confidence from their morphology, from what has been seen of trunk limb activity in L. gracilicornis , and from what is known of scraping branchiopods of other orders, is to collect detrital material as they sweep over the substratum. Notwithstanding their minute dimensions they are of robust construction.

Although this probably cannot be observed even in the living animals – scrapers are minute, and the limbs that bear them move rapidly, are largely hidden within the carapace and are just one part of a complex armature of spines and setae – it is evident that each is capable of considerable movement at its articulated base. Simple movements of spines and setae at their basal articulation are common in crustaceans and are often easily permitted by flexible cuticle. More complex movements are possible as a result of elaboration of the cuticular articulation which, as in notostracan branchiopods, may involve a thickened ring of cuticle on which the spine rests. Further elaboration by modifications such as the development of a boss against which the base of the spine can rest during the working stroke, and of a pedicel of flexible cuticle which allows the spine to move with considerable latitude is seen in the notostracan Lepidurus apus (Linnaeus, 1758) (see Fryer, 1988: figs 16–18, 21, 24, 40, 42, 43, 46).

The movements of which the scrapers of L. simiaefacies are capable are evidently refined. In particular it is obvious from preserved material that each is capable of some rotation. For example, in Figure 3A the scrapers lie in the ‘relaxed’ position. The row of denticles that performs the scraping operation here lies more or less parallel to the surface to be scraped. To scrape efficiently some clockwise rotation seems necessary. The denticles will then lie at a less obtuse angle to the substratum and one that is more suited to scraping, and probably ‘lock’ in this position as the limb remotes. They will rotate back to the position seen in Figure 3A as a result of the elasticity of what is in effect the arthrodial membrane of the articulation as the limb promotes and the load on the articulation is withdrawn. There is a projection on the basal margin of each scraper which must slide over a mound on the well-defined cylinder of cuticle on which it rests, and which perhaps locks the scraper in position during the working stroke.

Scrapers are arranged on the endites in groups that make up functional units. Although the orientation of these units when in action is often different from that seen in preserved material, it can be deduced with considerable confidence from morphology and from what has been seen of trunk limb activity in L. gracilicornis . As in a wide range of scraping crustaceans, adjacent scrapers of a unit are often so arranged that the most proximal is short and successively more distal members are progressively longer. This means that there is overlap in their fields of action and that the entire unit scrapes a broad swathe of the substratum ( Fig. 6A ). Moreover, the members of a series often follow a pattern that is common in such devices in which there is a graded succession from coarse to fine scrapers. The coarse scrapers pass over the substratum first and are able to dislodge relatively large and perhaps attached, particles, while that last member or members of the series are not only longer but more finely toothed and comb or sweep up particles left behind by their more coarsely toothed companions ( Fig. 2B ). This can be called the scrape and sweep arrangement and can apply to scrapers borne on a single limb or on successive limbs. However, this is not inevitably the case. Fine-toothed scrapers may precede coarser scrapers during the swing of a limb. This is so on some of the endopodites. Here long, coarsely armed distal scrapers extend posteriorly beyond the range of action of more proximal, sometimes shorter, fine-toothed scrapers. Material swept medially and towards the food groove by the distal scrapers during one cycle of movement is brought to a position from which it can be swept further by the fine-toothed scrapers of the assemblage behind during the next cycle. Both this and the scrape and sweep arrangement can be seen here as distal, coarsely toothed scrapers are followed by even more distally and laterally located fine, comblike scrapers on the same limb.

The distal elements of the endopodite of trunk limb 1 are so finely toothed as scarcely to merit description as scrapers. They over-reach the scrapers of endite 5 – themselves more finely toothed than their counterparts on succeeding limbs – and brush any material collected by them medially and dorsally. Trunk limb 1 clearly contributes less to the collection of food than do succeeding scraper-bearing limbs, but can cope with any material that drifts forward as the animal feeds.

Because the trunk limbs are not flat but are complicated three-dimensional structures, arrays of scrapers do not simply lie in line one behind the other, but are inserted on curved surfaces at different levels. The movement of a limb is not just a stereotyped backward and forward beat: it involves flexion and extension. Moreover, as described below, individual endites are capable of independent movements. The versatility of a scraping/sweeping limb is therefore considerable, and the ability to make fine adjustments by moving endites independently of each other is an advanced condition in a phyllopodous limb. It also demands a high level of co-ordination between adjacent limbs. Impressive co-ordination is apparent even in resting laevicaudatans and must be vastly greater in L. simiaefacies as it collects food from irregular surfaces. Feeding is therefore an even more complicated process than it is, for example, in such ctenopod branchiopods as Sida Straus, 1820 and Diaphanosoma Fischer, 1850 , which have only six pairs of essentially homonomous trunk limbs, of which the sixth pair is greatly reduced, that beat in a monotonously repetitive metachronal rhythm as they extract particles from suspension. It is also more complicated than that of most anostracans with their 11, or occasionally more, pairs of homonomous trunk limbs, at least when they employ them for extracting suspended particles. Even in anomopod branchiopods in which individual limbs of great complexity co-operate with others of very different structure to make an intricate device, versatility of a single limb is essentially limited to the first two pairs and is of a restricted nature. Only the second trunk limb of such anomopods as employ it for scraping food from surfaces rivals the limbs of L. simiaefacies in versatility and even these limbs do not achieve versatility in so many individual parts as do those of the latter.

Movements of the limbs are not confined to swings of the entire appendage. Individual endites, particularly the elongate distal members, and the endopodite, can be flexed. This swings them, and their armature of scrapers and other spines, posteriorly and dorsally, i.e. towards the food groove, and gives subtlety to the movements of their batteries of scrapers, and facilitates the transfer of collected material to the median chamber and ultimately to the food groove. On endite 2 of at least limbs 4–6 there is a fold of cuticle that presumably facilitates some flexing within the endite as the limb moves. It can be seen in Figure 5A but is not apparent in isolated, mounted limbs. The arrangement of the intrinsic muscles responsible in trunk limb 6 is shown in Figure 4 .

The entire limb swings forward and backward in a cycle of promotion and remotion as the animal swims. These, or similar movements when the animal is at rest, or almost so, near the bottom, contribute to food collection. As in locomotion, remotion is the working stroke. These two movements are effected by extrinsic muscles but accentuated or modified by a complex arrangement of intrinsic muscles.

Anteriorly, powerful extrinsic muscles (EP) originating in the trunk insert about halfway along the anterior margin. Their contraction promotes the limb by pulling its anterior region towards the trunk. This is facilitated by a wedge-shaped region near the base of the limb ( RF) on whose inner face the cuticle is thinner than that adjacent to it, and whose outer face appears to be covered over a more extensive surface with thin cuticle. The cuticle of this wedge clearly concertinas, and the angle of the wedge becomes more acute, when the extrinsic promotor muscles contract. The thin outer cuticle presumably also facilitates flexing in this region as the limb remotes and the endites move towards the food groove.

Antagonistic to these promoters is a long extrinsic remoter (ER) that originates in the trunk and inserts on a long ridge-like transverse apodeme (TA) located towards the distal margin of endite 2. More dorsally this muscle bifurcates and gives rise to a smaller extrinsic muscle (GA), whose fibres presumably remain distinct throughout, that ‘abducts’ the gnathobase. Other extrinsic muscles have roles that are less easy to elucidate. Muscle EA, a closely associated narrower muscle (EB) and muscle EC, which inserts more distally, probably assist remotion of the limb in different ways in collaboration with muscle ER. Two smaller muscles (ED), of uncertain function, perhaps make minor adjustments to movements of the limb as a whole.

The endopodite (En) articulates with the corm by a distinct hinge joint. Such a joint is unusual in phyllopodial appendages but is known in the first trunk limb of notostracans, where, however, the cuticle is much thicker. Extension of the endopodite is achieved particularly by means of two well-developed intrinsic extensor muscles (1 and 2) that originate more proximally in the corm. Although these have poor mechanical advantage they are stout, and their load when the endopodite spines participate in scraping is probably seldom great. A very small muscle (3), whose role is uncertain, conceivably helps to flex the endopodite, but has no homologue in adjacent endites. A larger muscle (4) is also problematic. It originates on the margin of the limb and seems not to be involved in moving the endopodite. There appear to be no flexors of the endopodite, and certainly none to fulfil this role in adjacent endites. In a phyllopodial limb it would be no surprise to find haemocoelic pressure responsible for the extension of a trunk limb element, but here it is flexion that appears to lack the necessary muscles – for which such pressure appears unsuitable. Cuticular elasticity is a possible answer.

Endites 4 and 5 (E4, E5) each have an extensor muscle that inserts in a similar position to those of the endopodite. That of endite 5 (muscle 5a) is part of a larger muscle (5) that originates on the transverse apodeme (TA) near the middle of the limb, the function of whose major component is unclear. It may serve as a brace. The extensor (6) of endite 4 also originates on the apodeme. No flexors are present. Endite 3 (E3), which is very different in shape, has a similar extensor muscle (7) to that of endite 4, which suggests that is it capable of a little independent movement, although this is probably small. Muscle 8 superficially suggests it may have a similar function to muscles 5, 6 and 7 and may move endite 2 (E2) slightly, but in fact it originates on the transverse apodeme (TA) and, like its more robust companion, muscle 9, which does likewise, runs obliquely dorsally and anteriorly. These muscles may be concerned with adjusting the shape (curvature) of the limb during remotion.

Apart from being swung by pro- and remotion of the entire limb, the gnathobase (Gn) is capable of independent movements. ‘Abduction’ by extrinsic muscle GA has been noted. The reverse movement is granted by intrinsic muscle 10. A pivot is necessary to permit this and appears to be provided by a region of the gnathobase where a fibrous ‘endoskeleton’ (F) has developed. The two walls of the flattened gnathobase are here united by fibrous threads, and provide a ‘firm’ region exactly where a pivot is needed to allow the muscles to operate antagonistically.

Proximal to muscle 10, and running parallel to it, is a short muscle (11), and distal to muscle 9 and located near the anterior margin of the limb is a short muscle (12) and a much smaller muscle (13). The contraction of these perhaps changes the shape of the limb slightly, and therefore the angle of attack of the armature of the endites. Distal to this and originating on firm cuticle of the corm, muscle 14 runs obliquely proximally, and inserts on the face of the corm. This presumably gives a subtle change in form, or a minute twist, to the corm when it contracts. Muscles to the exopodite leave the corm in its vicinity and it may help to brace the corm as the exopodite muscles contract. The delicate exopodite (Ex) has its own musculature (muscles a–e) that is not concerned with the feeding mechanism.

Although it is not possible to specify with certainty the roles of all its muscles, it is evident that the limb is capable of a variety of movements that can be controlled with considerable precision. Other limbs have a basically similar muscular system but there are minor variants. Thus, endite 4 of trunk limbs 1 and 2, which bears only fine scrapers, is provided with a mere sliver of muscle 6 as a flexor. Even in trunk limb 3 it is more robust. In limb 1 endites 2 and 3 are virtually fused and there is no muscle equivalent to that (7) serving as a flexor in limb 6.

As well as food-handling spines, each of the first eight pairs of trunk limbs is armed with a row of screening setae ( Figs 1A , 5A , 7A, B ). These long, slender, spiniform setae arise in a single close-set row near the posterior margin of endites 2 and 3, are directed posteriorly and incline towards the food groove. Each bears, either on each side or, in some regions, only on the dorsal side, a series of very fine, close-set, spinule-like setules, all of which lie in the same plane. The entire array makes up, on each limb, a fine-meshed screen, the distal part of which overlies, and is functionally continuous with, the proximal region of the screen on the limb behind. The result is a continuous screen which covers the inner face of the proximal endites of the trunk limbs on each side of the body and is a conspicuous feature of an individual bisected longitudinally and viewed from the midline ( Figs 1A , 7A ). A few similar but more widely spaced, more flexible, and more coarsely setose setae arise from more distal endites ( Fig. 1A ). Because the anterior trunk limbs are much larger than the last few pairs, the screen covers most of the series, leaving only the dense array of spines of the posterior limbs uncovered ( Fig. 1A ). These fine-meshed screens, one on each side, make up a median cage between the two sets of trunk limbs.

Towards the proximal end of endite 2 the screening setae arise closer together and incline more steeply dorsally. Here they fan out to cover the distal portions of the gnathobase of their own limb and to some extent of the gnathobase behind it ( Fig. 5A ). Here therefore they effectively seal off a potential dorsal entrance to the inter-limb spaces and ensure that collected food particles are confined to the narrow median chamber. In this region, the dorsal setules are very long while their ventral companions are reduced to short, more robust, minute spinule-like structures ( Fig. 5A ). The long setules of one seta overlie the next dorsal seta, each lying between two spinules which therefore support them, prevent them from being bent so as to touch each other, and maintain the regularity of the grid of minute spaces so formed.

A screen superficially resembles a filter, but its component setae are not filter setae and do not function as such. Each screen merely prevents entry into the inter-limb spaces of particles collected by the scrapers of the distal endites and endopodites and guides them towards the food groove. As seen in Figure 7B , in regions where both rows of setules are long, unlike those of filter setae, which arise opposite each other, they often do so alternately. Similarity to a filter is enhanced by the pilosity of the faces of endites 2 and 3, whose component setules are directed posteriorly and dorsally towards the food groove. These perhaps help to clean the screening setae and ensure that the particles pass to the food groove, but they are not cleaning setules of the kind that clean filters in some filter-feeding crustaceans. Efficient screens are probably particularly important in L. simiaefacies , which must at times inevitably sweep up large accumulations of detritus that could pose problems if they found their way into the inter-limb spaces.

The cage of which the screens are the walls is widest ventrally adjacent to the aperture between the carapace valves, and narrower towards the food groove. The two walls can, however, come together to leave a very narrow gap between them. Only observations on living, feeding animals can reveal how wide is the central space during this process. The spiny armature of endites 2 and 3 protrudes through the screen of the limb in front near the distal ends of its setae but their relative positions inevitably change through the cycle of movements. The armature of endites 2 and 3 of the first trunk limb is of course not covered by the screen.

Material swept posteriorly and dorsally, and towards the midline, by the food-collecting scrapers of the distal endites and endopodite comes within reach of the posteriorly directed stout spines of endites 2 and 3, of which those on endite 3 are often serrated distally. As each limb becomes flexed posteriorly during the working phase of the cycle of movement (remotion), these spines also swing posteriorly and, towards the end of the cycle, ever more dorsally, and push the collected material towards the food groove. Escape laterally into the inter-limb spaces is prevented by the screening setae on each side. Towards the posterior end of the trunk limb series, as limbs diminish in size, the endites become shorter and the distal endites more similar in shape to endites 2 and 3. Especially in the last four pairs the armature of endites 2–5 is highly specialized and is illustrated especially in Figure 8A, B . Here, where screening setae are no longer present, it is clear that material collected and swept posteriorly and towards the food groove by the scrapers of the distal endites of more anterior limbs can be dealt with effectively by the extremely intricate arrays of spines borne by the distal elements of these limbs, which operate on a ‘broad front’. The distalmost scrapers of endite 5 of trunk limb 9, clearly displayed in Figure 8B , and those located on the corresponding endite of limb 10, inevitably help to sweep material towards the food groove as these limbs remote. The rest of the armature of these endites, which is remarkably specialized for this purpose, does likewise. The endite spines of limb 10, seen especially well here, but also of limbs 9, 11 and 12, are widely spread, like the fingers on an extended human hand, and are clearly well suited to pushing accumulations of detrital matter on which L. simiaefacies feeds. The rather similar array present on limb 11 is largely hidden between limbs 10 and 12 in the photograph ( Fig. 8B ). These inevitably push accumulations of food deeper into the median space between the two sets of limbs. Somewhat more proximally these endites are provided with complex arrays of hook-like spines, variously bent, which clearly serve to drag such material further towards the food groove. More proximal endites, shorter than endite 5, bear dense arrays of spines directed obliquely towards the food groove ( Figs 8A , 13A ).

Although the movements of the posterior limbs are much less than those of their anterior companions, and in L. gracilicornis limbs 11 and 12 appear not to participate in the steady metachronal beat of the trunk limb series, and can indeed be used for gripping while the more anterior limbs sweep up material, it is evident from their elaborate armature that in L. simiaefacies they contribute to passing food to the food groove. Indeed, much of the food probably enters the food groove posteriorly. For its handling, the four posterior pairs are indeed highly specialized. Some idea of the complexity of the machinery of which they are a part can be obtained by visualizing the apparatus seen in Figure 8B (with the limb rotated somewhat anticlockwise) separated by a narrow gap from its mirror image on the opposite side.

From that of trunk limb 4 ( Fig. 5A ) to the posterior end of the series, the gnathobases are increasingly heavily armed on the posterior margin with spinules or spines that contribute to the pushing or dragging of material into the food groove. On the four posterior limbs the armature that fulfils this role consists of short, stout spines ( Figs 5B , 8A , 13A, B ).

It is to the food groove, and particularly to its posterior end, that all food particles are swept or pushed. Here they are dealt with by the gnathobases of the trunk limbs whose form is seen in Figures 1B , 5A, B , 6B and 9A . As in many branchiopods, these project into the groove beyond the point at which the limb pivots so that, as the corm of the limb swings posteriorly, its gnathobase swings anteriorly and sweeps material forward along the food groove. The gnathobase of each trunk limb lies within the confines of the food groove. The latter is deep for much of its length, but anteriorly it widens markedly and becomes more shallow in the vicinity of the first two pairs of limbs. This can be seen in Figure 1A where the torn cuticle that lines the groove can be seen to form a shallow basin. Elsewhere its cuticular lining clearly reveals the segmentation of the trunk ( Fig. 1B ) in a way not apparent in most branchiopods, but which is seen also in notostracans. As in that order, the anterior margin of each segmental component overlies that of the segment in front as a transverse fold which offers no resistance to food material being swept forward.

The gnathobases differ in form from limb to limb and diminish in size towards the posterior end of the series. All are suitably armed for sweeping particles anteriorly along the food groove. Each bears two stout denticulated spines that project into the food groove. Except for those on the distinctively shaped first gnathobase these are broadly based ( Figs 1B , 5B , 6B ). On all except the first three gnathobases, they are accompanied by two robust smooth spines, of which, especially in the middle of the series ( Fig. 1B ), the anteriormost is curved so as to be directed posteriorly. Although the denticulate spines clearly play a major role in pushing food material anteriorly, the curved spine must inevitably lift material slightly from the food groove as the gnathobase swings backward on promotion of the limb. There must clearly be some advantage in this: perhaps ensuring that material does not clog the food groove and lifting it towards the denticulate spines in readiness for their next working sweep.

From near the distal extremity of all except the first and last two gnathobases arises a long backwardly directed seta (RS) ( Figs 1A, B , 4 , 5A, B ). These setae lie parallel to the food groove and, for much of the series, extend back sufficiently far to over-reach not only their homologue on the limb behind but also part of that on the next posterior gnathobases ( Fig. 1B ). Those of more posterior limbs, especially of gnathobase 10, are much shorter. These setae are soft, flexible and easily distorted. Each is provided with a double row of long, soft setules that arise, not along the middle of the seta but on the dorsal side, i.e. on the side adjacent to the food groove, towards which they are directed obliquely. These setules are seen in Figure 5B on the dorsal side of the posteriorly directed seta of gnathobase 9, which, because of the rotation of the limb, here appears to be directed vertically.

These setae occupy a position similar to that of a stiffer spine or seta on the gnathobase of various ctenopods ( Cannon, 1933) and, as part of a complex specialization, on that of the gnathobase of the second trunk limb of the anomopod Daphnia O.F. Müller, 1785 and its allies ( Cannon, 1933; Fryer, 1991) and the macrothricid (or ilyocryptid) Ilyocryptus Sars, 1862 ( Fryer, 1974) . In all of these the spine concerned serves to sweep or beat material towards the food groove, and in Daphnia is provided with a fine comb for cleaning an adjacent filter. Although suitably located the setae of L. simiaefacies are not suited to a function that involves sweeping or beating. They are too soft and delicate, and their setules too long and soft for this. It is suggested that, lying as they do in pairs, they form a ‘blanket’ beneath the food groove, which is open ventrally, and may help to prevent material escaping from it during promotion of the limb when the gnathobase makes its return swing. Their flaccid nature probably allows them to bend near their base and allow distal portions to remain covering the contents of the food groove. We suggest that they be termed retaining setae. Their softness and pliability and the orientation of their setules are such that they must easily be pushed aside by particles entering the food groove anywhere along its length. In the position occupied by such a seta on other limbs, trunk limb 1 carries a short, stouter seta.

As well as the heavy armature described, all the gnathobases have an array of longer, more slender, anteriorly directed spines on the anterior margin. Except on the first pair ( Figs 1A , 9A ), where there is just a single row, these are arranged in two rows, one longer than the other, each consisting of about 16 spines on gnathobase 2. The number is gradually reduced in successively more posterior limbs. Appropriate sensory setae are present. The posterior margin of gnathobase 1 bears a few soft setules ( Fig. 9A ). The outer face of the more anterior gnathobases is carpeted with spinules and setules which presumably clean the anteriorly directed spines of the anterior margin of the gnathobases behind.

It is obvious that there can be no possibility of food being carried along the food groove by a current of water as was once suggested was the case in certain branchiopods ( Cannon, 1933). Not only is the food groove largely blocked by the gnathobases and their armature but it sometimes contains material collected by the scraping limbs. This can include fine sand grains. As in all investigated branchiopods that pass particulate matter along a food groove, transport of food particles is clearly an entirely mechanical process.

The food groove deepens from posterior to anterior to accommodate the progressively larger gnathobases. Posteriorly it is narrow and the gnathobases are essentially vertical. In the only sectioned animal – a male cut in transverse section – from about midway along the series it becomes more V-shaped, the V then becoming shallower. Certainly, also in the female, at about the level of the anterior margin of gnathobase 4, it greatly widens ( Fig. 1A ) – an unusual feature in a food groove. This widening is associated with an equally unusual orientation of the maxillule, a robust appendage whose armature ( Fig. 9B ) is not greatly different from that of the not much larger gnathobase of the first trunk limb. This consists of stout denticulate distal spines and a row of finely setose spines that clearly serve as sweepers. Maxillules assist in passing food forward to the mandibles in many branchiopods, the only exceptions being the raptatory Leptodora Lilljeborg, 1861 (Haplopoda) , which either has no maxillules, or in which, as Olesen, Richter & Scholtz (2003) suggest, they are fused to the ‘lower lip’, and the Onychopoda in which they are rudimentary. So far as we can ascertain, in all other branchiopod orders other than the Laevicaudata the armature of the maxillules operates in the vertical plane – like that of the trunk limb gnathobases of Lynceus . In L. simiaefacies the maxillules project obliquely into the widened region of the food groove and are so orientated that their armature operates essentially in the horizontal plane, which gives ample scope to the array of sweeping spines. A similar arrangement clearly prevails in L. brachyurus , of whose mouthparts Sars (1896: plate XIX, fig. 4) gave a very informative illustration more than a century ago, and presumably does so in other members of the genus.

The differences between the arrangement in Lynceus and in the Spinicaudata with which the Laeviacaudata were long associated is particularly striking. The maxillule of the Spinicaudata has as its major element a many-spined lobe that operates in the vertical plane. This carries a subsidiary lobe on its inner face, also armed with spines, sometimes long. This arrangement is illustrated for Limnadia lenticularis (Linnaeus, 1761) by Sars (1896: plate XV, figs 9, 10) and, very clearly, for what is either Cyzicus sibericus or Eocyzicus sibericus by Cannon (1933: fig. 28). Quite apart from the subsidiary lobe, which may bear long spines that operate in the vertical plane, the major element is very different from the maxillule of Lynceus , and reinforces the great differences in both morphology and feeding mechanisms of the two orders.

The mandibles, like those of other species of Lynceus , display many of the attributes of the basic branchiopod plan – a boat-like chitinous skeleton housing the major transverse muscles and providing points of insertion for others originating external to them; pivoting laterally by a pointed articulation, and with the mid-line masticatory surface extending posterior to anterior across a chitinous internal projection ( Fig. 9B ). In all species, however, the molar region of each mandible is dorso-ventrally flattened so their opposed masticatory surfaces are narrow in the vertical, and elongate in the longitudinal, plane ( Fig. 10A, B ). This is indeed a distinctive attribute of the order. Sars (1896) illustrated the orientation of the mandibles of L. brachyurus to each other in a way never subsequently emulated, which also shows their relationship to the maxillules. He also noted, as Richter (2004) was to do independently more than a century later, that they differ from those of ‘other bivalved Phyllopoda’, by which he implied spinicaudatans, ‘and show an approach to the structure characteristic of the Apodidae’ – in modern parlance the Notostraca . The masticatory surface is traversed by a series of vertical ridges. Of these there are 25 ( Fig. 10A–E ), which is considerably more than in other species for which information exists. There are 12 ridges in L. brachyurus ( Kotov, 2000) , 11–13 (or 14) in L. biformis (Ishikawa, 1895) ( Kotov, 2000; Richter, 2004) – illustrated as L. dauricus Thiele, 1907 by Kotov, which, according to Brtek (2002) is conspecific with L. biformis , nine in L. tatei (Brady, 1886) ( Richter, 2004) , and 12 in L. gracilicornis . In all species with few ridges, as particularly well shown in the beautiful SEM photographs of Richter (2004), the ridges make up a row of stout teeth separated from each other by deep troughs. In L. simiaefacies the more numerous ridges, of which the anteriormost and posteriormost are shorter than those in the middle of the series, are inevitably less prominent and instead of troughs there are narrow gaps between the closely associated ridges.

The ridges differ somewhat from those of other species of Lynceus hitherto described. Apart from the posteriormost, which is somewhat separated from the rest ( Fig. 10E ), each has a more or less flat face, each end of which, especially ventrally, is drawn out into a pointed tooth-like extension which projects more or less at right angles to the ridge ( Fig. 10A, B ). Towards the middle of the series the teeth are very distinct and take the form of elaborations of chitin, presumably sclerotized (or even impregnated with minerals) that arise from a distinct socket ( Fig. 10C ). Towards the anterior end of the series some of them are directed somewhat anteriorly ( Fig. 10D ), clearly to facilitate passing material forward. Especially ventrally there is often a much smaller, subsidiary tooth internal to, and located near the base of, the large tooth. The face of each ridge is also provided with a number of much smaller, short, stout, stud-like denticles located mostly near its anterior and posterior margins, and with small, blunt denticles near the middle of each. The isolated posteriormost ridge is the largest and bears a pair of conical projections ( Fig. 10E ). Posterior to this ridge lies a tooth-like conical projection much larger than any element of the more anterior armature.

Differences in the armature of the masticatory surface of the mandible between L. simiaefacies and its congeners are presumably related to differences in the nature of the material handled, but in what way is not clear. Perhaps some of the coarse particles, including minute sand grains, collected by L. simiaefacies could not be dealt with easily by stout teeth separated by deep troughs in which they might be trapped, but this is only conjecture.

Because the denticles of the ridges project towards their partners, it seems unlikely that the mandibles can do much grinding, or indeed that their molar surfaces can ever have close contact with each other. On the other hand, opposed denticles can probably grip relatively large items, perhaps even wider than the band of ridges, and can evidently cope with the fine sand grains of which many are ingested, which it is impossible for a substratum scraper to avoid in desert ponds. The robustness of the denticles is perhaps as much an adaptation to a ubiquitous substratum as to the handling of a particular kind of food. Smaller sand particles can doubtless fit into the gap between the opposed mandibles where the small stud-like denticles offer protection to the cuticular surface. The stout teeth of other species of Lynceus , separated by deep troughs, may be unsuitable for coping with the ubiquitous sand grains. The dorsal and ventral horizontal rows of denticles appear well suited to dealing with such.

To Sars (1896) the masticatory surface of the mandible of L. brachyurus suggested that this species might include more animal matter in its diet than do spinicaudatans, which Lundblad (1920) appeared to accept, and thought that it possibly included smaller crustaceans in its diet, which seems not to be the case. Notwithstanding its rows of stout denticles, L. simiaefacies is clearly a detritus feeder.

Figures

Figure 1. Lynceus simiaefacies. A, bisected female, ventral side uppermost, seen from the midline after removal of the labrum which obscures part of trunk limb 1, showing the median face of the 12 left trunk limbs. Limbs 11 and 12 are very small. The trunk limbs lie more or less as at the end of promotion. Of particular note is that, except on the distal scraper-bearing endites, and at the extreme posterior end, the median face of the limbs is covered by a screen of posteriorly and somewhat dorsally directed setae. The gnathobases of the seven anterior limbs are conspicuous: more posterior gnathobases less easily seen. Note how the cuticle-lined food groove, here torn away over much of its length, widens greatly into a ‘basin’ in the vicinity of the first three gnathobases. Trunk limb 12, detached from the right hand series, is present at bottom right. Except that the gnathobase lies proximally its orientation bears no relation to that in life. Scale bar 500 Mm. B, distal elements of gnathobases 4–6, seen within the confines of the food groove (FG). Note the stout denticulated spines that sweep food material forward (here to the left), the smooth curved spines (SCS) that are directed posteriorly, and the long, posteriorly directed, soft retaining setae (RS), of which the anteriormost extends posteriorly across the two gnathobases posterior to it. The tips of many screening setae, of which those seen anteriorly are clogged with screened out detritus, are readily seen. Note how the cuticle that lines the food groove reflects segmentation. The segments, of which parts of four are represented, increase in length from posterior to anterior. Scale bar 100 Mm. All subsequent figures refer to L. simiaefacies.

Figure 2. A, part of the distal array of scrapers of trunk limb 2. E5, endite 5; En, endopodite. Scale bar 100 Mm. B, a series of scrapers on a single endite (with others beneath). Note how the elaborately articulated scrapers nearest to the observer, which pass over the substratum first, are short and bear stout teeth, while those that do so later are longer and bear a longer row of finer, more comb-like teeth. Scale bar 50 Mm. C, oblique view of adjoining arrays of scrapers. Scale bar 50 Mm.

Figure 3. A, a row of robust scrapers on the endopodite of trunk limb 5, showing the elaborate articulation of each. The scrapers of endite 5 lie immediately beneath it. Scale bar 20 Mm. B, two typical stout scrapers and, beneath them, several broad, more finely denticulate scooping spines and rather similar spines with much finer denticulation, typical of endite 3 of the anterior limbs. Scale bar 20 Mm.

Figure 4. Right trunk limb 6, as seen from the outside, to show its general form and musculature. Only part of the armature of endites 2 and 3 is shown to indicate the arrangement of the grid setae and some of the spines. Only the distal armature of endites 4 and 5 and the endopodite is visible from this side. Further rows of scrapers are located more proximally on the hidden margins of these elements, whose location is indicated by a dashed line in each case. Some of these spines are visible distally on the margin of endite 4, but are omitted to avoid confusion in that region. Likewise, only a few of the gnathobasic setae, of which there are two rows anteriorly, are shown. The delicate exopodite is stippled to show its distinctness from the corm. The long setae with which both its dorsal and its ventral lobes are provided are omitted. The long, cylindrical epipodite is also stippled. Scale bar 250 Mm. EA, EB, EC, assistant extrinsic remotor muscles; ED, probable adjustor muscles; EP, extrinsic promoter muscles; ER, extrinsic remoter muscles; E2–E5, endites 2–5; En, endopodite; Ep, epipodite; Ex, exopodite; F, fibrous endoskeleton of gnathobase; GA, ‘abductor’ muscle of gnathobase; Gn, gnathobase; GS, grid setae; RF, region of flexibility; RS, retaining seta of gnathobase; TA, transverse apodeme; 1–14, intrinsic muscles (for probable roles see text); a–e, muscles of exopodite.

Figure 5. A, proximal region of trunk limbs 4 and 5. Endite 2 (E2) is seen overlying the gnathobases (Gn4 and Gn5) and bearing long screening setae and stout spines directed obliquely posteriorly and towards the food groove. Here some of the armature of the grid setae, e.g. at top right, consists of long dorsal setules that rest on the adjacent seta and are held in place by short, spinule-like ventral setules of that seta. Gnathobase 3 (partly visible) has twisted and reveals its densely setose outer face, which overlies the armature of the anterior margin of gnathobase 4 (revealed by its twisting), which it clearly cleans. The soft, flaccid nature of the long posteriorly directed retaining setae (RS) is obvious from the way they have been distorted during preparation of the material. Scale bar 100 Mm. B, proximal parts of trunk limbs 9 and 10, with part of 11 visible. The robust armature of the gnathobases, which sweep along the bottom of the food groove, is well displayed at the bottom of the photograph, as are numerous, mostly short, spines on the posterior margin of each gnathobase, which push and drag food material towards the food groove. These include straight spines and, as seen on gnathobase 10, curved spines that drag rather than push. The long, posteriorly directed retaining seta of gnathobase 9 is the most posterior such of the series. Because of the angle at which the gnathobase is viewed, it appears to be directed ‘vertically’ and is seen from its dorsal side. More distally the dense array of spines of endite 2 is displayed. To the left the posterior limit of the screen is marked by the tips of several screening setae. The armature of the more distal endites of these limbs is illustrated in Figure 8A and B. Scale bar 50 Mm.

Figure 6. A, endopodite of trunk limb 5 (En5) and adjacent scraper-bearing elements. Anterior (to the left) of it are scrapers of the endopodite and endite 5 of trunk limb 4. The endopodite of trunk limb 5 overlies endite 5 of the same limb. The latter bears a row of scrapers distally: more proximally can be seen some of its armature of often pointed spines. More posteriorly, the posterior margin of the endopodite of trunk limb 6 can just be seen. Note the arrangement of scrapers on the endopodite of trunk limb 5. Proximally they are short. More distally they become progressively longer. Most are stout, but the three distalmost, which are the longest, are finely toothed and over-reach and sweep up material dislodged by their shorter, more robust companions. The distalmost spine in the series is smooth. The short, finely toothed proximal scrapers are suitably located to sweep material brought within reach by distal scrapers of the assemblage in front during the previous cycle. Scale bar 100 Mm. B, details of the posterior region of a gnathobase typical of the middle of the series, showing, posteriorly, one of the stout denticulate spines that sweep material forward, i.e. to the left, along the food groove. Near its base is a sensilla. Between the other two conspicuous elements can just be seen the basal portion of the second stout denticulate spine. This lies more laterally, i.e. nearer to the food groove wall, at a deeper level of focus in the darkness of the food groove. Anterior to the posterior denticulate spine is a smooth spine and, anterior to that, a robust sweeping spine with a reflexed distal portion that bears long spinules. The flexure is probably not an artefact (cf. trunk limb 9, Fig. 5B). The distal part of this spine is partly obscured by the long setules of an adjacent sweeping seta (cf. trunk limb 9, Fig. 5B). Scale bar 20 Mm.

Figure 7. A, the screen that covers the median face of the anterior trunk limbs. The stout spines borne on the underlying endites, which protrude through the screen and push material towards the food groove during remotion of the limbs, are also well displayed. Scale bar 200 Mm. B, details of the screening setae of the type present in many parts of the screen. Scale bar 20 Mm.

Figure 8. A, armature of the four posterior trunk limbs as seen in an individual in which they are fortuitously twisted towards the mid line. Proximally the four gnathobases are seen close together, that of limb 11 being jammed between those of limbs 10 and 12. The more distal armature of this limb is largely obscured by that of adjoining limbs. Being the last of the series, with nothing to obscure it posteriorly, trunk limb 12 clearly displays its gnathobase and endites 2–5. Its endopodite is largely obscured by endite 5 but some of its spinous armature can be seen arising from behind that endite. The arrangement of the armature of the endites of this limb facilitates understanding of that of the three more anterior limbs where, especially in limbs 9 and 10, it is seen ‘full face’ and it is difficult to see where that of one endite ends and that of the next begins. That the more proximal armature of all these limbs consists largely of stout spines, sometimes curved near their tips to enable them to drag material towards the food groove, is readily seen. More distally, especially on limbs 9 and 10, the armature of endite 4 includes many stout hook-like spines (for details see Fig. 8B). The posterior limits of the grid setae, which do not cover the apparatus seen here, are seen to the left (anterior) of the illustration. Scale bar 100 Mm. B, details of the distal endites of trunk limbs 9 and 10 seen at greater magnification. The fifth and fourth endites of limbs 9 and 10, seen face on, have a broad posterior ‘margin’, in each case provided with an armature of highly specialized spines of great diversity. On endite 5 the spines extend somewhat obliquely to facilitate the ‘embracing’ of masses of food material. Those of the anterior margin of limb 9 are scrapers that differ somewhat from orthodox scrapers and have hook-like tips. These perhaps serve to sweep and push rather than scrape surfaces, to which access is less easy than for anterior limbs, but their exact role cannot be defined. More distally and posteriorly are hook-like spines of various types and sizes that clearly drag in collected material or, in the case of very large distal hooks, retain it, and there are more hook-like spines internal to the scrapers. On endite 4, a dense and diverse array of hook-like spines and a few small denticulate spines are well suited to passing material proximally. The distal armature of endite 5 of trunk limb 10, while extending outwards so that the entire ensemble forms a scoop, is quite different. Whereas the distalmost spines can clearly scrape and sweep, their close companions bear diverging marginal rows of denticles, well suited to scooping and sweeping masses of material proximally to a region functionally continuous with endite 4, like which it is armed with a profusion of hooks. The large spiniform spines seen on the extreme right are of trunk limb 12. Much of limb 11 is here hidden between this limb and limb 10. These limbs also pass material proximally. Scale bar 50 Mm.

Figure 9. A, gnathobase of trunk limb 1. Scale bar 100 Mm. B, maxillule; which operates in the horizontal plane. Scale bar 100 Mm. C, skeleton of mandible, dorsal view. Scale bar 250 Mm. Abbreviation: post, posterior.

Figure 10. A, molar region of left mandible in face view. Anterior is to the left. Scale bar 100 Mm. B, the same tilted to show the posterior conical projection. Scale bar 100 Mm. C, the ventral ‘teeth’ of four of the transverse ridges. D, anteriormost ridges. Scale bar 20 Mm. E, posterior end of masticatory region, showing the large isolated posterior ridge of the masticatory region and adjacent ridges. Scale bar 100 Mm.

Figure 13. A, right trunk limb 12 seen from outer (anterior) side. The endopodite is largely obscured by endite 5. Although not visible from this angle, the gnathobase bears two very small, but robust, denticulate spines like its anterior homologues. Scale bar 200 Mm. B, enlargement of A showing detail of the spinous armature of the gnathobase of trunk limb 12, which drags and pushes food into the food groove. C, eggs. Scale bar 100 Mm. D, slightly oblique anterior view of the head of a female. The two eye-like patches are the sensory fields where numerous short setae arise from areas of thin cuticle. Scale bar 1 mm. E, antennule. Scale bar 100 Mm.