Table of Contents
- Summary & Introduction
- Epifaunal Worm Tubes on Lower Lias Ammonites – Results
- Epifaunal Worm Tubes on Lower Lias Ammonites – Detailed Observations and Interpretation
- Epifaunal Worm Tubes on Lower Lias Ammonites – Discussion
- Epifaunal Worm Tubes on Lower Lias Ammonites – Conclusions and References
4. Detailed observations and interpretation
Here we confine discussion to two alternative categories; attachment to a juvenile, living ammonite, which involved the worms reacting to the growth of the ammonite shell, and attachment after the ammonite died.
4.1. Attachment to and growth on a juvenile living ammonite
The most common association previously reported by Lange (1932), Schindewolf (1934) and Buys (1973), and also among our material, is of a single worm tube that attached initially on one side of the ammonite, reached the periphery, grew along the venter in the same direction as the ammonite and was eventually overgrown itself by the ammonite (Figs 2, 8). We regard this as the ‘standard pattern’. We assume that the worm tube grew at the same rate as the ammonite. This is a reasonable assumption since the modern serpulid Spirobranchus giganteus lives entirely within modern corals (at least ten different species) and keeps pace with the growth of the various coral colonies it inhabits (see, for example, Marsden and Meeuwig, 1990). Thus, it is possible to reconstruct the growth of both the worm tubes and the ammonites (Fig. 8). This shows that in the living orientation of the ammonite (Trueman, 1941, fig. 14d; Jacobs and Chamberlain, 1996, fig. 5) the site of initial attachment was close to 6 o’clock (Fig. 8A) and that subsequent growth appears to have maintained the aperture of the worm tube as close to this position as possible.
Fig. 8. A-D, Schlotheimia (Schindewolf, 1934, pl. 2, fig. 4b) in likely life orientation to illustrate the position and orientation of the worm tube as it and the ammonite grew. Note that although the ammonite shell grew in a clockwise direction (as seen from the left side) any point on the shell surface effectively rotates anticlockwise because the ammonite aperture remains in a horizontal position throughout the ammonite’s growth. The growth stages are: A, at the initial attachment of the worm (indicated by the arrow); B, at the end of its growth in the umbilical seam; C, at the end of its growth across the outer whorl and D, the adult ammonite. Note that a growth increment (heavy line) that would extend the worm tube to the lowest point on the venter (v) subtends the angle a o v, whereas the same growth increment in the umbilical seam subtends a much larger angle (a o us). Worm tubes growing in the umbilical seam grew faster relative to the ammonite than those on the venter. Note also that the worm tube started to cross the outer whorl when it reached a point at 6 o’clock with respect to the ammonite. Further growth in the umbilical seam would have been less ideal for feeding. A-D are based on the assumption that the worm tube grew at the same rate as the ammonite. Under this assumption the position of the aperture remained close to 6 o’clock with respect to the ammonite throughout growth. The illustrations are mirror images of the original ammonite to simulate clockwise rotation with growth. On the original specimen the worm tube attached to the right side, not the left as shown.
On this assumption it is possible to consider how the worm would benefit from having its aperture at 6 o’clock. Ammonites are assumed to have swum backwards as modern Nautilus does (see Jacobs and Chamberlain, 1996 for a review of ammonite buoyancy and swimming). Nautilus inhales water into its mantle cavity, passes it over the gills and expels it through a muscular siphon, thus producing a reaction, which drives the animal through the water. Since the siphon points forward when relaxed, Nautilus normally swims backwards. Modern serpulid worms are filter-feeders (e.g., Fauchald and Jumars, 1979) with two feeding structures (branchia), which they deploy on either side of the aperture for feeding. Each branchium bears ciliated filaments either arranged in a circle or a spiral. In still water the ciliated filaments create currents that pass from the outside of the circular branchia inwards or from below (the widest part of the spiral) upwards in spiral branchia. Thus, as normally deployed in front of the worm tube aperture, the ciliated currents would approach from behind the aperture in both types of branchium. Hence, with the worm tube aperture at 6 o’clock the currents generated by the ammonite swimming backwards and by the worm’s branchia would be aligned (Fig. 9).
Fig. 9. Camera lucida drawing of a Promicroceras shell in life orientation seen from the left side to illustrate the orientation of currents resulting from an ammonite swimming (solid arrows) and those generated by the worms’ branchia (broken arrows). The currents only align with the worm tube aperture at 6 o’clock (6) with respect to the ammonite’s normal orientation enabling the worm to feed more effectively. In Promicroceras the distance along any radius (R) from the origin (o) to the umbilical seam (us) is almost the same as that from the umbilical seam to the venter (v). 3, 9 and 12, correspond to positions on a clock face.
At all other positions the currents would not be aligned. So, for example, at 3 or 9 o’clock they would be at right angles to each other, whereas at 12 o’clock they would be directly counter to each other. Furthermore, even the most sluggish ammonite would have swum at a speed that would generate much stronger currents than the worm’s branchia (0.6 mm per sec at 20o C; Gray, 1928). Thus, the 6 o’clock position would enable the worm to filter the maximum volume of water and, presumably gather the greatest quantity of food. Even so, when the worm tube grew across a whorl its aperture was not in the ideal orientation (Fig. 8C). Clearly the worms could survive in less than ideal situations, perhaps because modern serpulids are not permanently attached to their tubes and presumably the fossil worms could twist within the tube to minimize the angle between the two currents. We believe this is the best explanation of the ‘standard pattern’ and that departures from it require individual consideration.
At least three consequences derive from the ‘standard pattern’. Modern serpulids attached to sloping surfaces grow upwards (e.g., Seilacher, 1960) and some grow in association to form reefs (e.g., Chapman et al., 2007). However, in crossing a whorl the fossil serpulids grew downwards. Thus, it seems serpulids do not respond to gravity or light, but to currents. In most cases growing up a sloping surface will expose a modern serpulid to stronger currents. Equally, under our interpretation of the ‘standard pattern’ the fossil serpulids were attempting to take advantage of stronger currents. Dietl et al. (2000) recorded serpulid worms with tube apertures predominantly pointing backwards on the horseshoe crab Limulus, which would have been able to take advantage of currents generated when the limulids moved forwards.
Secondly, a possible reason for serpulids growing on ammonites was that the worms could take advantage of scraps of food released when the ammonite fed. This seems unlikely for three reasons. Firstly, feeding currents generated by the worms would have run forwards and carried food particles away from the branchia. Secondly, modern serpulids effectively filter very small particles (Dales, 1957; Davies et al., 1989). It is doubtful if a feeding ammonite would generate many particles in this small size range. Finally, if ammonites fed in shoals and competed for food as many marine predators do, it is highly unlikely any would have fed without swimming rapidly to avoid competitors. The chances of any worm having been able to filter out microscopic particles generated by the ammonite feeding were very small. Interestingly, Bailey-Brock (1976) reported serpulids on the ventral surfaces of two species of slipper lobsters with the apertures directed towards the lobsters’ mouths, but suggested that this orientation was influenced by feeding currents generated by the lobsters’ mouth parts, not by the worms stealing the lobster’s food.
Finally, our interpretation of the ‘standard pattern’ is only valid if the ammonites actively swam backwards. Thus, it tends to confirm previous interpretations of both shell orientation and swimming activity of Promicroceras (e.g., Trueman, 1941; Jacobs and Chamberlain, 1996) and, by implication, other ammonites.
4.2. Attachment and growth only after the death of the ammonite
LYMPH 2010/35 (Fig. 10) is a damaged specimen of Echioceras with almost half the ammonite missing and has at least 5 worm tubes attached on the right hand side, but none on the left. The worm tubes meander randomly across the right side of the ammonite shell and show no preferred direction of growth or orientation with respect to the ammonite (Fig. 10A). One Polymorphites (LYMPH 2010/66) also has possibly three worm tubes meandering on the left side and venter (Fig. 11). We conclude that in both cases the worms attached post-mortem. The tube labelled ?3 (Fig. 11) appears to be a collapsed agglutinated tube. If so, whatever animal built it, the ammonite must have been on the sea floor at the time to allow the tube builder to pick up sediment grains. Below this worm tube a small example of the inarticulate brachiopod Discinisca occurs (D, Fig. 11B). These two examples are locally significant, as conditions suitable for benthic epifauna seem to have been relatively rare during the deposition of the Charmouth Mudstone Formation in Dorset and Devon.
Fig. 10. A, left and B right side of LYMPH 2010/35, a wholly septate Echioceras. Note that there are no worm tubes attached to the left side, but possibly 5 (a-e) on the right side. The worm tubes show no preferred orientation with respect to the ammonite growth direction. (a) follows the umbilical seam in the direction of ammonite growth, but eventually overgrows (b). (b) follows the umbilical seam, but counter to the ammonite growth. (c) grew in a tight U-turn between (a) and (b). (d) grew counter to the ammonite growth and may be the same tube as (e). The preserved part of (e) follows a U-shaped path in the grooves between the ammonite ribs. The random patterns of worm tube growth on one side of the ammonite both suggest post-mortem growth of all worm tubes. Scale bar = 10 mm.
Fig. 11. LYMPH 2010/66. A, left side and B, venter of Polymorphites from bed 118 of the Belemnite Marls Member, Charmouth Mudstone Formation, just west of Golden Cap, Dorset, showing possibly three worm tubes. Worm tube 1 crosses the left side (A) and climbs onto the venter (B). Worm tube 2 performs a U-turn round one of the ribs (A) and then grows across the ribs of the penultimate whorl. Possible worm tube 3 is the meandering impression on the venter (B) to the left of ?3. D is a minute brachiopod Discinisca on the venter of the ammonite between worm tubes 1 and ?3 (B). Scale bars = 5 mm.