Penetration of Water Into Blind-Ended Capillary Tubes and Its Bearing on the Functional Design of the Lungs of Soldier Crabs Mictyris Longicarpus

Soldier crabs, Mictyris longicarpus, are small to medium, round-bodied crabs (males typically reach 14 g and 2.5 cm carapace width) living on sheltered intertidal sand-flats around the coasts of Australia (McNeill, 1926). They are remarkable for their habit of emerging en masse at low tide to ‘march’ across the sand-flats in huge ‘armies’ as they systematically rake the surface layers of sand for food (Cameron, 1966; Quinn, 1983). As a soldier crab feeds, water (pumped from the branchial chambers) is used to separate food particles from sand (Cameron, 1966; Quinn, 1980, 1983, 1986).

Soldier crabs appear to be obligate air-breathers, obtaining over 90 % of their oxygen requirement via sophisticated lungs developed within their branchial chambers (Maitland, 1987; Farrelly and Greenaway, 1987). The lungs are complex structures formed from a regular series of branching finger-like airways which penetrate the deep vascular parenchyma lining the branchiostegites (lateral extensions of the carapace covering the gills) and portions of the epibranchial septa (membranous partitions which partially separate the gills below from the lungs above; Quinn, 1980; Maitland, 1987; Farrelly and Greenaway, 1987). In this paper we describe plastic casts of the lung and examination with a scanning electron microscope which reveal that the lung consists of a system of nonanastomosing, blind-ended, air-filled capillary tubes. The gill compartments have been found always to contain water (Quinn, 1980,1983, present study). This water may reach the airways of the lung as the crab changes its orientation. Since oxygen diffuses about 10000 times more slowly through water than through air (Dejours, 1981), the lung will cease to function effectively by aerobic respiration if substantial amounts of water enter it. The question we address, therefore, is how an air-filled lung exposed to water in such close proximity does not flood. To this end, we develop a model which simulates the architecture of the soldier crab’s lungs and use it to derive formulae to describe the extent to which water at various tidal depths may penetrate a system of non-anastomosing, blind-ended, air-filled capillary tubes. The non-anastomosing feature of the lung design also facilitates the shedding of the lung cuticle at each moult.

Adult male soldier crabs, Mictyris longicarpus, were collected at low tide on intertidal sand-flats around Botany Bay in Sydney, Australia. Crabs were dug out of their sand retreats (about 30 cm below the surface) and were kept for up to a week in fresh damp habitat sand exposed to natural light and ambient temperatures (20−30°C).

Lung tissue was observed with a scanning electron microscope (Cambridge S4-10). For these observations, the lung tissue was fixed in 10% buffered formaldehyde, dehydrated, critical-point-dried, and gold/palladium coated before examination.

To minimize problems arising from tissue shrinkage and the like, dimensions of the lung airways were measured directly from plastic replicas of freshly removed lung tissue. Measurements were made with a Wild M5 binocular microscope with an eyepiece graticule and camera lucida.

The plastic replicas were made by pouring Batson’s no. 17 corrosion compound (kit 7349, Polysciences, Warrington, PA, USA; shrinkage is less than 1%, Bennett, 1988) over the luminal surface of freshly removed branchiostegites (on which most of the lung is formed). A thinned solution of the compound was used consisting of 4 ml of monomer base, 1ml of methylmethacrylate, 1ml of catalyst and four drops of promoter. Before polymerization commenced, the preparation (compound plus lung) was evacuated to 8 kPa for a few seconds to remove trapped air from the compound and the lung airways. Full atmospheric pressure was then restored to force the compound into the evacuated lung airways. After setting, the cast was split from the calcified shell of the branchiostegite and immersed in a macerating solution of 20% NaOH (w/v) plus 0.3 mol l⁻¹ EDTA. After repeated washing and maceration, the last traces of cuticle were removed by immersing in full-strength ‘Domestos’ (a household bleach) prior to washing and drying. Some casts were coated with gold/palladium and photographed under the scanning electron microscope.

Vestibule diameter was measured in eight male crabs whose live body masses spanned the range 1.1g to 10.5 g. Excised branchiostegites (right side only) were dipped in a 1 % Methylene Blue solution for 2 min to delineate clearly the perimeter of each vestibule. Sixty vestibules covering the posterior and middle regions of the lung were drawn from each crab with the aid of a camera lucida at 64× magnification. From these drawings, the average vestibule diameter was calculated for each individual.

The volume of water held within the branchial chambers of non-feeding soldier crabs was determined as described by Maitland (1990).

As is well known, surface tension forces determine whether a liquid will wet a solid. For the case of a liquid lying on a solid in the presence of a vapour, the surface tension forces are γ_LV, γsv γsL-Each force is normal to the line that is the boundary between the vapour (V), the liquid (L) and the solid (S). Surface tension is sometimes defined as the force per unit length of the boundary. The forces are, respectively, tangential to the liquid/vapour surface, the solid/vapour surface and the solid/liquid surface. The angle between γ_LV and γ_SL is the contact angle, θ. The contact angle between sea water and the surface of soldier crab cuticle was measured from a 5μl drop of sea water placed on the surface. Surfaces for which θ was measured were the luminal surfaces of excised lungs (the inside surface of the branchiostegite, see Fig. 2A) and the external smooth carapace cuticle located between the branchial chambers on the dorsal surface of the crab. The drop of water was photographed in profile and θ was measured from 30 × enlargement prints on both sides of the drop and the average value was found for three drops (three crabs: 1 drop on the dorsal surface of each crab and 1 drop on the lung surface of each crab).

Soldier crabs live on sheltered sand-flats between the neap tide levels. Prior to being submerged at high tide, crabs bury themselves beneath the sand (Fig. 1) in a ‘corkscrew’ fashion. This behaviour traps a pocket of air within a cavern 4−5 times the volume of the crab (Fig. 1A). Crabs digging with their right side travel downwards in a clockwise spiral, excavating sand from the floor of the cavern and plastering it onto the roof. In this way the trapped pocket of air is taken below to a depth of between 10 and 30 cm, or to the level of the water table (Fig. IB). The air cavity remains intact beneath flooded sand. The process is reversed (sand removed from above and placed below) when crabs emerge again at the next low tide (Cowles, 1915; McNeill, 1926). Laboratory observations made during the study reported herein on crabs burrowing against a glass-sided tank, and field observations made during the excavation of crabs from the sand, confirmed the reports of Cowles and McNeill. At high tide in the Sydney area, crabs within their air caverns may be covered by up to about 2 m of water (Marine Services Board of New South Wales Tide Tables).

Although the data presented below apply primarily to the lung lining the branchiostegites, the lung formed on the epibranchial septum is essentially similar, although less well developed (Maitland, 1987; Farrelly and Greenaway, 1987).

Vestibules penetrate a short distance into the lung before subdividing into primary, secondary and tertiary respiratory airways oriented at right angles to the branchiostegite. Plastic casts of these airways provide a clear picture of their architecture (Fig. 2B). The air columns terminate against the branchiostegite as blind-ended tubes 35−80 gm in diameter (Fig. 2B). Each vestibule leads into its own discrete system of airways and there are no anastomoses between one vestibular airway system and the next (Fig. 2B,C). Air columns are spread evenly across the surface of the branchiostegite (Fig. 2D).

In seeking a physical basis to explain these observations we note that, while it is common knowledge that water readily penetrates a capillary tube that is open at both ends, water penetration into a complex system of non-anastomosing, blind-ended airways or tubes (Fig. 3), which may be used to simulate a crab lung, has not been considered. In essence, it is the compression of air trapped in the blind-ended tubes that produces a force to oppose that of surface tension and thus effectively prevent the tubes from flooding. This explanation may be convincing in the case of a crab that maintains a normal upright posture within the burrow, but an active crab makes many changes of orientation. During these changes, with water lying on the surface of the lung, trapped air might be expected to travel as a bubble along a lung ‘capillary’ and thus escape from the lung airways to leave those airways flooded as water replaces the air. However, we show below that this expectation is unlikely to be realized.

Sea water was found to wet the external carapace cuticle of the solder crab, with a contact angle of 41° (40.8±3.3°; mean±s.D.; N=6). However, 0 was 83° (82.8±1.7°; mean±s.D.; N=6) for sea water in contact with lung cuticle. This increase in 0 for lung cuticle could be due either to the presence of a waterproofing substance or it could result from a cavitation or ‘bubble’ effect caused by the air trapped within the vestibules beneath the drop of water. Rough or cavitated surfaces are known to increase contact angles (McIver and Schurch, 1987). For the present study, we chose to take θ for lung cuticle to be 41° because we are uncertain as to the true cause of the increase in θ for lung cuticle and because this is a ‘worst-case scenario’.

Although the detailed branching topology associated with each vestibule within the lung varies both within and between animals, the overall topology is essentially similar. For purposes of the calculations to follow, this branching topology can be simplified as shown in Fig. 3. The length of the vestibule (region 1) is b and its radius is r_Y. Five airways (the range was found to be from 3 to 9), each of length l₂ and radius r₂, lead from the vestibule. From each of the airways of region 2 arise two (or sometimes three) airways of length l₃ and radius r₃. Finally, each tube of region 3 leads into three (or more) blind-ended tubes of length /₄ and radius r₄ (region 4 of Fig. 3). To generalize, we denote a region by N, the number of branches in each region by n_N, length by l_N and radius by r_N. If M is the number of regions present, the possible values of N are 1,2,.., M. In large crabs (over 9g), M may be 5.

Of course, the tubes do not have a uniformly circular cross section and the walls are not uniformly cylindrical, but the data are adequate for the purposes of estimating depths H₁, H₂ and H₃ (the tidal depth H which will totally flood regions 1, 2, 3, respectively). Using Fig. 3 with values for n as given in Table 1, and the measured values for tube lengths and radii with θ=41°, equation 5 gives the following values: H₁=2.72m, H₂=11.2m, H₃=23.21m, when each region is fully flooded (that is, when x₁=x₂=x₃^=:0). In the study area (Botany Bay, Sydney), the highest recorded tide during the year 1986-1987 was 2 m (as noted above). Clearly, therefore, under normal circumstances, water cannot penetrate very far into the soldier crab lung vestibules, even at high tide.

Some of the soldier crabs brought into the laboratory attempted to moult so it was possible to study how the cuticle lining the lungs is shed. The nonanastomosing blind-ending architectural design of the lung appears to be important, and, like pulling fingers out of a glove, the old cuticle is withdrawn from the lung (direct observation).

By necessity, if the lung were made of anastomosing tubes, the cuticle would need to have specialized breakage zones incorporated within the branches to enable the cuticle to be shed. Such breakage zones were not found in the soldier crab lung. In general, arthropod respiratory structures avoid this problem by utilising non-anastomosing designs. Thus, insect tracheae (Wigglesworth, 1972), isopod pseudotracheae (Hoese, 1983) and tick spiracular structures (Pugh et al. 1988) are all invaginated, non-anastomosing, tree-like structures.

Moulting allows an arthropod to grow. In soldier crabs, moulting also allows the lungs to be maintained in a healthy condition. Some crabs, particularly older individuals (increased intermoult interval), were found with vestibules blocked with sand grains. However, these could be removed along with the old cuticle at the next moult.

We thank David Sandeman for support while D.P.M. was in receipt of a University of New South Wales Dean’s Postgraduate Scholarship. A.M. thanks W. Dawber for useful discussions. We also appreciate the helpful comments made by the reviewers.