.
QUANTITATIVE BEACH RESEARCH
ALONG THE NORTH SEA COAST

Willem Krommenhoek, Ph.D.

1. INTRODUCTION

2. GENERAL FEATURES OF THE NORTH SEA BEACH

2.1 Beach profile
2.2 Ripples and the Reynolds effect
2.3 Wind ripples and backshore phenomena

3. ANATOMY AND GROWTH OF VALVES OF BIVALVE MOLLUSCS

3.1 Anatomy of a valve
3.2 Growth of valves of bivalve molluscs

4. BEACH RESEARCH ON VALVES

4.1 Which side is up and orientation of valves
4.2 Uneven distribution of right and left valves
4.3 Distribution of valves and fragments over the beach zone
4.4 Patterns of erosion

5. PERFORATION OF VALVES

6. FACETING OF VALVES

Literature

North Sea beach

1. INTRODUCTION

Quantitative research on North Sea beaches dates back from the first decades of the 20th century, when in the German literature attention was paid to the orientation of valves of bivalve mollusks on sandy beaches. The aim of this research was to interpret the original position of shell bearing sediments in folded geological formations This was the birth not only of quantitative beach research, but of the description of a variety of taphonomic processes. These processes describe how the remains of organisms are changed before fossilisation and include two categories: processes that cause modification to the skeletal parts, like encrustation or overgrowth of hard skeletal substrates by other organisms; fragmentation or breakage of skeletons; abrasion or the wearing down of skeletons, usually by sand particles in water or air; dissolution and bioerosion or degradation by boring and grazing animals. The second category of taphonomic processes includes processes which affect the relationship among shells in an assemblage such as orientation. After death, skeletal remains like valves are moved by the transporting medium and oriented relative to their hydrodynamic properties.
In modern taphonomy a lot of research has been done on both groups of processes. In this survey I want to introduce some examples of taphonomic research, together with some results obtained on Dutch beaches.

2. GENERAL FEATURES

2.1 Beach profile

beach slope

beach slope with accumulation

The North Sea beaches of Holland are composed of loose sand and are comparatively wide. Most of the sand grains are whitish in color and are made of silica, the fewer darker particles contain iron oxide. In ripples the darker grains, which are heavier, accumulate in the lowest parts and make the ripples clearly visible. Along the coast run flat topped ridges or balls and shallow troughs or lows, of which two to four may be observed in the intertidal zone. The total number of these ridges and troughs depends on the bottom slope. They generally increase both in width and height from the high tide level downwards. The seaward slopes of the balls are usually less steep than the landwards slopes, with the exeption of the uppermost ridge where the reverse condition seems to be more common. The general pattern of balls and lows is rather stable over a period of time. The balls do not form continuous ridges along the length of the beach, but are interrupted by shallow cross channels or outlets, which occur at regular intervals. Strong seaward currents can develop in these outlets during the backwash, which form the so called rip currents. Halfway between the outlets the lows may show a minimum depth, making the water that is pushed over the ball by waves slow away in two different directions. The result of all this is a complex system of currents and seawater movement.

The foreshore, which is characterized by wet conditions of the soil, is the place of ripple marks and the Reynolds effect (see 2.2). The backshore, which is normally quite dry, shows windripples, the development of young dunes, and different phenomena related to aeolian erosion (see 2.2).

2.2 Ripples and the Reynolds effect

low tide with ripples

ordinary current ripples

The position and properties of the subaquateously produced ripple marks in the foreshore are entirely dependent on the relief features. They can be found both in the foreshore, the area exposed at low tide and covered twice a day by seawater, and the backshore, but only during exceptionally high tides. Practically all ripple marks are formed during the ebb stage and the greater part just before the final retreat of the water. Where waves act on a horizontal bottom, symmetric wave ripples will be produced as a product of unidirectionally flowing water. When the waves reach the shore and the water becomes shallower, they become asymmetrical, the shoreward movement of the water particles being stronger than the backward movement. Asymmetrical ripples will be the result, with their steeper sides pointing to the slope. When the foreward movement of the seawater stops, all of a sudden the thin layer of water that is left from swash, percolates into the sand, leaving a clear line of small particles that cannot be taken back downslope by the backwash as this gravity generated flow can develop only where water is present on the beach slope. At the same time the shining beach surface turns dull and whitish.

clear lines of small particles

Ordinary current ripples are abundant on beaches, both in the lows and on the balls. The ripples on the balls and those in the lows often differ in their relative height, which is generally about one-tenth of the wave length. Where the water spreads out in a shallow layer over the smooth seaward slope of a ball, rhomboid ripples are produced. These ripples have a flat rhomb shape, the longer axis of the rhomb indicating the direction of the current. This direction may be marked also by systems of straight groovings. These ripples travel downstream by deposition of material on the lee sides. The necessary conditions for the development of rhomboid ripples are shallow water, rapid flow and a smooth bottom.

outlet with lenguoid ripples

Where ordinary transverse current ripples have been formed and the waterdepth is reduced to a level that the crests of the ripples are uncovered by water, linguoid ripples may develop. The flow of water is now concentrated through gaps and lower places which become widened by erosion. After a certain time the original current ripple is transformed into a new type of regular pattern: that of the linguoid ripple.

forming of linguoid ripples

Like thin whitish lines of deposition mark the end of the water transport by swash, there are also frequently found concentrations of fine shell detritus on the smooth surfaces of the balls. This minute, so called longitudinal ripple marks, with heights not exeeding 1 mm, and with little regularity in spacing are formed either during swash or backwash. The term liniation or striation would be more appropriate.

swash

Walking over the wet foreshore during low tide, one can observe how with each step the sand around our foot looks lighter as if it turns instantaneously dry. Most people think that the weight of our body has pressed the sand particles together. But if this was the case, there would be less space left for the water between the sand grains and the water would have come to the surface next to our foot. So, obviously the body pressure did not compress the sand, but instead has expanded it. This sounds against all reason, but this is what happens: in the wet foreshore the sand particles are piled up in the densest possible way, leaving a minimum of space for the water between the sand grains. Each alteration in this situation will result in an increase of space between the grains, resulting in lowering of the water level. A man's foot will disturb the densest piling up of the grains, increasing the volume between them, as a result of which the water level is lowered. This explains the lighter spot around the foot. This well known, but only by very few beachcombers understood phenomenon, was described and explaine by Osborn Reynolds in 1885.

2.3 Wind ripples and backshore phenomena
On the dry backshore we can observe transport of sand grains by wind. Only from an airspeed of 5 m/s (18 km/h) grains of sand are lifted in the air. Most grains travel within 10 mm of the surface with individual laps of 0.5 - 1.5 m. The height which saltating particles reach depends on its size, wind speed and surface characteristics, With increasing wind speed, grains can be lifted up to one meter high and a few may hit your face as well. When those lifted grains fall back to earth, they carry so much energy that they can make another particle jump, a process known as saltation. This process is the major mechanism of sand movement.

Wind ripples are asymmetric in cross section with windward slopes of about 10 degrees and lee-side gradients of 30-35 degrees. They develop from slight irregularities in the sand surface through a combination of surface creep and saltation. The formation of wind ripples can be nicely observed when a sand loaded wind passes over smooth and moist sand surfaces, especially just below the high tide line. The initial ripples are only small isolated rolls of sand, formed at comparatively irregular interspaces. Freshly deposited grains of sand very soon become moistened by capillary action and stay on their place, forming streaks parallel to the wind.

After an extremely high tide, a gale or after rain, the sand of the backshore is wet as well and will dry up during the next sunny period. Once dry, and in the presence of wind, sand grains will be blown away until they meet an obstacle. However, underneath a valve or other object, drying of the sand takes more time, so the still wet sand grains will stick together while the surrounding particles are blown away, resulting in a mini pyramid. After a while this little tower of wet sand itself is target of sand abrasion and grows thinner and thinner until it collapses under the weight of the shell.

deposition of sand behind object

Another process that can be observed on the backshore is the pattern in which sand grains which are transported by air, come to a halt around a small object on the beach. There is a deposition of sand before the object, but not in contact with it, pruduced by colliding particles that are bounced back, and a deposition in the wind-free area behind the object in the form of a tail figure.

valve on a pyramid of vet sand

Besides ripples, small dunes are also usually present on the backshore. Both longitudinal and transverse dunes can be seen, the last type usually with the typical barchan shape and a few decimeters high. It is in this environment that quantitative research took place. We will now turn to the results of it.

3. ANATOMY AND GROWTH OF VALVES OF BIVALVE MOLLUSCS

3.1 Anatomy of a valve

Most bivalves have two-pieced shells, each piece being called a valve. The two valves are joined together by a flexible chitinous ligament and they close effectively because the upper inner edge of each has a hinge embellished with interlocking teeth and pits. The small teeth below the embryonomic valve or umbo are called cardinal teeth, the longer ones the lateral teeth. Inside each valve are marks left by muscles. The muscles that are used to pull the valves together (adductor muscles) leave adductor scars in the valves. Bivalves with two adductor scars in each valve usually have a clearly visible pallial line connecting them, the muscular edge of the animal's mantle having been attached along it. In most bivalves the pallial line has a distinct embayment, known as the pallial sinus, indicating the former presence of siphons.

The umbo is situated on the dorsal margin and usually towards the anterior end of the shell. Many valves have a lunule, a heart-shaped impression anterior to the umbo, and in some cases there is a escutcheon, an elongate depression on the dorsal margin.

In order to determine whether a valve is a right or left one, the valve is placed with the umbo upwards and the ligament between the umbo and you. In this position the valves show as a right or left one.

3.2 Growth of valves of bivalve molluscs
Valves of bivalve molluscs increase in length and width for a certain period of time in a linear way, meaning that the shape of the valve remains about constant during the growth process. This was established for two species that grow to a length of about 10 cm; the fresh-water bivalve Anadonta cygnaea (N=131) and the marine clam Mya arenaria (N=142). Both species double their width from about 25 mm, when they are 45 mm long, to about 55 mm when 90 mm long, meaning that at all ages the length of the valve is 1.8 - 2.0 times its width. See fig.

When the ratio between length and weight is studied no linear relation is found, as weight is related to surface. In both species the weight of the valve is doubled when its length increases from 45 to 60 mm and again from 60 to about 70 mm, and once more from 70 to 90 mm. Weight at these lengths being resp. 1.5 gr.; 3 gr.; 6 gr.; and 13 gr. From this data it is clear that more weight is added when larger animals grow a certain extra length than when small animals do. Actually growing from 70 to 90 mm in length added 7 grams in weight, that is 0.35 gram per mm; while growing from 45 to 60 mm in length only 1.5 gram weight was added, which equals only 0.1 gram per mm. Middle sized valves growing from 60 to 70 mm in lengh gained 3 grams in weight, which equals 0.3 mm per mm.

For a smaller species like the cockle (Cerastoderma edule) about the same results were found. From a large number of this very common species weight, width and surface of valves was determined. For an estimation of the surface of the valve it was put on a sheet of paper whereupon its outline was drawn. The surface of the projection of the concave valve was measured. Small valves, with a width of 1 cm have a surface of about 1 cm² and weigh about 0.5 grams. Valves with a width of 2 cm have a surface of 2.6 cm² and weigh 0.75 grams; valves with a width of 4 cm have a surface of 9.5 cm² and weigh 5.5 grams. Growing from 1 to 2 cm in width, an average of 0.025 gram of weight was added per mm; during growth from 2 to 4 cm in width, 0.238 gram weight was added annually.

As a general conclusion it can be stated that for species of bivalves with large valves, doubling of the length by growth results in an increase of 8.7 times the intial weight. For species with smaller valves like the cockle, this increase in weight is on the average 7.3 times. As we will see in paragraph 4.3 this feature is very important in distribution of valves over the beach as a result of transport by sea water.

BEACH RESEARCH ON VALVES

4.1 Which side is up and orientation of valves
One of the earliest subjects studied by quantitave beach research was the question how seperate valves of bivalve molluscs are deposited on the seafloor and beach: convex or concave side up? In the German literature of the 1920's the question was raised several times as part of the problem how to interpret the original position of shell-bearing formations (Musschelkalk) in folded geological formations. One of the results of research on this subject was the formulation of Richter's rule: dish-like bodies such as valves of bivalves come to a final rest with the convex side facing upwards. But, as in still water seperate valves always sink with the concave side up when thrown into the water, they obviously have to be turned over by sidewards movements of seawater. Schaefer added to this rule of Richter the observation that once a valve is turned upwards and has thus reached a relative stable position, the valve becomes oriented with the heavy hinge pointing up-current.

shell assemblage

shell bed of Ensis sp.

To see how this rule is appropriate on the Dutch beach at Kwade Hoek Nature Reserve in southwest Holland, in total 1632 valves of Macoma balthica were collected at 11 spots selected over a distance of 3 km. It was found that the percentage of valves found with the concave side up, ranged from 15% to 36%, meaning that 65% to 85% was found with the convex side up. When the average for all 11 spots is taken, the ratio concave side up : convex side up is 25% : 75%. This confirms the classic literature, stating that convex side up valves being three times more frequent than the concave side up ones.

shell bed of bivalves

After the first investigations on how valves are oriented on the beach, a lot of research has been done on processes describing how the remains of organisms are changed before fossilisation. Two categories of these so-called taphonomic processes are distinguished: processes that cause modification of the skeleton or valve, like encrustation or overgrowth of hard skeletal substrates by other organisms; fragmentation or breakage of skeletons; abrasion or the wearing down of skeletons, usually by sand particles in water or in air; dissolution and bioerosion or degradation by boring and grazing animals. Secondly, there are processes which affect the relationship among shells in an assemblage such as orientation. After death, skeletal remains like valves of molluscs, are moved by the transporting medium and oriented relative to their hydrodynamic properties. A lot of research has been done on subjects of what is now called taphonomy. Recently Nagle did experimental work on orientation in a tank and described a basic difference between wave and current orientation of valves. In case of current orientation a majority of shells with an elongate or conical shape point into ar away from the current, depending on the geometry and mass distribution of the shell. In low angle swash zones valves are oriented perpendicular to the ripple marks by incoming and outgoing swash. On high angle beaches a current type of orientation will develop. I looked into this matter as well.

At Kwade Hoek Beach I selected a part of sandy beach with an angle of 10 degrees and with plenty of valves of Ensis, Spisula and Cerastoderma. For 700 valves of Ensis the orientation of the long axis towards the wave crest line was determined by counting valves in sections of 30 degrees. It was found that the majority of valves was oriented in a current type as described by Nagle. The same procedure was done on a slope with a 4 degrees angle. The same pattern was found here, both for juvenile and adult specimens. Obviously, an angle of about 4 degrees is sufficient for this type of orientation.

For Spisula and Cerastoderma Schaefer's statement that a valve becomes oriented with the heavy hinge pointing up-current was checked. For this purpose in an area of a square meter on the 4 degrees angle beach, the number of valves with the orientation "top-up" (top facing landwards), "top-down" (top facing seawards), and "top-sidewards" was counted. 923 Spisula valves and 280 Cerastoderma valves were observed. No prevailing orientation was found, neither for the ones with the convex side up (70%), nor for the ones with the convave side up (30%). Other countings at different places gave the same result. This does not match with Schaefer's statement that the prevailing orientation will be top facing up. However, it might be that swash and backwash currents were too weak at this spot to move stable valves as I did the observations close to the low tide zone.

Other observations include the fact that on top of a shellbed of valves of bivalves the number of concave side up valves is much higher than among valves that are stable positioned on the sand.This fits with Richter's rule that only stable positioned valves are positioned convex side up. On top of the shell bed there is no possibility to anchor the valve with its outer edge in the sediment. Secondly, it was observed that Ensis valves on a shell bed do not show the current type orientation as described above. The current is too much absorbed by the shell accumulation to have enough power to turn the long valves.

4.2 Uneven distribution of right and left valves
It has been known for a long time that separated right and left valves of lamellibranch shells are not found in equal numbers when collected washed-up in beaches. And this does not only apply for valves, but also for flip-flops found washed-up on tropical beaches. Small differences in anatomy, weight and symmetry effect the way almost identical objects are transported. This phenomenon was also described in the early 20th century German literature, like other items as mentioned in paragraph 4.2. A good object to study this phenomenon are valves of the common sand gaper, Mya arenaria, with large valves, the left ones having a tooth-like projection.

On the beach of Kwade Hoek Nature Reserve in southwest Holland I collected at six spots over a strech of 3 km beach over 2000 valves in the summer of 1997. The percentage of left valves found here was highest behind a sand ridge running parallel with the beach for about 1 km and varied from 24% to 28%. From here the percentage of left valves dropped markedly in both direction, resp. to 5-10% and 9-16%.

This counting was repeated in winter. Now 419 valves were collected and the results were quite different. Behind the sand ridge, where the highest percentage of left valves was found in summer, no left valves at all were seen. At the ends of the 3 km collecting strip the highest percentages were observed now: 27% versus 7% in summer, and 7% versus 2% in summer. Alltogether, the percentage of left shells found in summer was 19% versus 47% in winter. It is obvious that the differences are the result of changing currents along the coast.

Finally, in early summer 1999 I visited the beaches of Schiermonnikooog, an island in the North Sea, situated 250 km north of Kwade Hoek. Here 1358 valves of the sand gaper were collected at 7 sites over a 7 km long stretch of beach. At three locations situated behind a sand ridge that was not flooded during high tide, no valves were found. Only on those beaches facing the North Sea directly, valves were present in numbers, the percentage of left valves being 26.7%. An average percentage of 10.6% was found for all spots, versus 12.3% in Kwade Hoek. Also the minimum and maximum percentages found at the different spots in May were close, being resp. 2% and 26% in Kwade Hoek and 3% and 26.7% on Schiermonnikoog.

4.3 Distribution of valves and fragments over the beach zone
Another part of quantitative beach research deals with the distribution of valves over the beach zone. In the 1960's Lever and Thijssen examined the distribution of the cockle over the different beach zones along the Dutch coast. It was found that 71% of the unbroken valves occur in a zone of 8 meters from the low tide line, versus 29% close to the high tide line. For fragmented valves these percentages were resp. 48% and 52%.

I followed in Kwade Hoek Nature Reserve the same procedure and collected valves and fragments in a 35 m wide zone running from the low tide zone up to the high tide level. This zone is about 300 meter long and alltogether 1159 valves and fragments were collected divided as follows: near the low tide level (N=118); 120 m up the beach (N=295); another 250 m in the same direction (N=371); and near the high tide line, just at the foot of the dunes. The percentages of unbroken valves at these different locations was resp. 96.6%, 30.5%, 27%, and 27.3%. For fragments these percentages were resp. 3.4%, 69.5%, 73%, and 72.7%. These results fit in with those of Thijssen, who found 29% of unbroken valves in the upperpart of the zone, versus 27% in this study.

At another occasion unbroken valves of the cockle as well as of Spisula elliptica and Macoma balthica were collected on a strech of beach with a marked slope. Spisula and Macoma are common, with solid, moderate inflated valves, 2.5-3.5 cm wide. In the low tide area 817 valves were collected and 591 in the high tide zone. It was found that the three species were not equally distributed over the beach, the cockle being more frequent in the high tide zone, Spicula and Macoma in the low tide zone. It was also observed that all three species were represented by bigger valves in the high tide zone. The average width of cockle valves being 20.1 mm in the low tide zone, and 23.0 mm in the high tide area at one location, and 23.4 mm versus 26.1 mm at another. When expressed as average weight of valves in the low tide zone compared to average weight in the high tide zone, we found for the cockle 1324 mg versus 2247 mg; for Spisula and Macoma 963 versus 1117 mg; and for other species 806 mg versus 944 mg. In terms of percentage the increase in valve weight is highest for the cockle, but this must be accounted for by the fact that cockles can grow as wide as 5 cm, where the others are restricted to a maximum size of 3-3.5 cm.

When the average size of the cockle increases from 20.1 mm in width to 23.0 mm, it means that at the high tide level valves are 14% bigger (wider) than at the low tide zone. However, as we have seen in paragraph 3.2 this means an increase in weight of 60% and of surface of 27%.

When the average weight of a particle in a valve/fragment mixture was calculated, it was found that particles in the low tide zone had an average weight of 1440 mg, decreasing to an average of 1087 mg for particles in the middle of the zone, and further decreasing to 705 mg for particles in the high tide zone. This means that the average particle weight in the heigh tide zone is half that of particles in the low tide zone.

Looking further into this subject I discovered something else. When cockle valves are divided in three size-classes: large (valves more than 3 cm wide); medium (valves 2-3 cm wide); and small (valves smaller than 2 cm), it was found that near the high tide line half of the unbroken valves (49%) were small ones, versus one fifth in the low tide zone. Because small valves are thin and with relatively little weight, they are easily transported and for that reason account for the significant decrease in particle weight near the high tide zone.

How can all these data be interpreted? Actually, when seawater loaded with sand and debris moves up the shore as swash, the pushing force of the swash will gradually diminish, bringing the majority of valves and fragments to a halt. Only small fragments with little weight will be pushed further on the slope. After a while the swash looses its force completely, resulting in the deposition of all particles. Then after a moment of complete stand-still a lot of water percolates down into the sandy beach, after which gravity pulls the water back, resulting in a backwash that starts at a lower level of the slope and gradually gaining force, not affecting the deposition of the upper swash. At the same time slightly bigger valves will gain more force during the swash as a result of their substantial bigger surface, resulting in a higher place on the slope when deposited. These two factors result in a relative abundance of smaller fragments together with bigger unbroken valves on the higher beach slope.

4.4 Patterns of erosion
Another aspect of taphonomic processes is the erosion of shells and valves after the death of the animal. Fragmentation of the valve starts with fracturing when they are vigorously moved by water. Hollmann described four types of fragmentation of valves, which I recognized on beaches in de Kwade Hoek Nature Reserves. These four patterns are:

(1) The shell breaks into large fragments, like in mussels (Mytilus edulis).
(2) Smaller fragments are chopped off irregularly from all sides until only the hinge and teeth remain, like in oysters (Ostrea edulis)
(3) Fragments break off along growth lines, like in clams (Ensis sp.), gapers (e.g. Mya arenaria), and piddocks (Pholas sp.).In case of Mya arenaria from the left valve only remains the toothlike projection, while the right valve often breaks in tow. This is also the case in piddocks.
(4) The shell cracks along the radial ribs, like in cockles (e.g. Cerastoderma edule). Young cockles and piddocks start eroding on the most concave part of the shell. The small intial opening widens and the shell eventually breaks into pieces.
Other valves, like those of Mactra, Macoma and Donax hardly break. They are relatively thick and homogenous in structure. When eventually damaged, all sorts of fragments can be observed.

When gastropod shells are exposed to impact, the arched whorls and the apertures break first. The further mechanical destruction is determined by the shape and sculpturing of the shell. Gastropods that are frequently found along the Dutch coast like Buccinum undatum and Littorina littorea do have a hard aperture and they erode until finally the entire columella is exposed.

5. PERFORATION OF VALVES
Among beach-found valves of the most prevailing species of bivalves, Spisula elliptica and Macoma nalthica, there is always a fair number with perforations. These perforations are caused by the carnivorous snail Natica sp., which drills holes in the shells of bivalves, thus killing the prey. When the Natica captures its prey, it twists the prey several times in its large extended foot. During this process mucus is secreted around the shell of the prey. The actual drilling process may take as long as 24 hours, and after the shell is punctured the prey is eaten. The location of the hole has been subject of study several times, with different results, depending on the species and collecting spot on the beach. Occasionally valves with two or more holes were reported as well. Also the diameter of the hole varies considerably, being related to the size of the predator. All lierature mention holes being more frequent in right valves than in left ones. This is not surprising as Macoma lies deep in the sandy bottom on its left side with the result that the predator's approach will be from above.

Thijssen observed three facts when studying perforations in valves: (1) perforated valves occur more frequent on top of the beach near the high tide zone, rather than at the base of it; (2) in the high tide zone the diameter of the holes is bigger than at the base of the beach slope; (3) in the high tide zone perforations are situated more closely to the umbo of the valve than they are at the base of the beach slope.

In May 2000 I collected 1755 complete valves of Spisula and Macoma on the Kwade Hoek beach, both in the low- and high tide zones. My results matched nicely with Thijssen's. At the top of the Beach perforated valves are more frequent than in other zones, 22% versus 13% in the low tide zone.

The average diameter of the holes studied by Thijssen was 1.915 mm for left valves and 1.972 mm for right valves, both valves having slighly bigger holes in the high tide zone. As I could not apply the techniques used by Thijssen, I cannot confirm these data.
Finally, judging from observation only, it seems also correct to state that holes in perforated valves in the high tide zone are situated closer to the umbo than they are in the low tide zone.

6. FACETING OF VALVES

In the tropics fine rings of limpet shells can be found which are the result of sand abrasion by moving water during the changing of tide. This faceting as the process is called, can also be observed on European beaches. A nice example can be seen in the cockle, Cerastoderma edule. However, unlike the situation in limpets, the plane of faceting makes an angle of about 60 degrees with the surface on which the valve was lying, which makes it virtually impossible that the abrasion had taken place when the valve was lying with its concave side up. It was Pratje who as early as the 1920s described that many valves were found in an upright position in the North Sea bottom at depths of 8 to 28 meters. In this position, he said, moving sand grains can abrade the top of the shell, producing a facet that is now parallel to the sea bottom plane.

Once washed ashore, a secondary erosion of the valve can take place. On the beach, when the valve is in stable position with the convex side up, abrasion produces a regular hole with polished edge situated on the convex side of the valve. Further rolling and dragging in the surf leads to cracks and punctures in the convex part of the valve, the resulting holes being irregular in shape until finally the valve breaks.

Occasionally we find valves with the original ring faceting and secondary erosion by surf action, resulting in a large hole, and eventually in the loss of the umbonal region, teeth and part of the convex shell.

cockles with broken holes

cockles with broken larger holes

On North Sea beaches I tried to establish the frequency of faceting in the common European cockle. For this purpose I took a sample from a shell accumulation with cockles (Cerastoderma edule); mactras (Spisula subtruncata and Mactra corallina), all being abundant, and razor clams (Ensis directus); rockborer clams (Petricola pholadiformis and Barnea candida); tellins (Macoma balthica and Angulus tenuis); striated venus (Venus striatula); soft shell clams (Mya arenaria) and banded donax (Donax vittatus) which are incidentally found. The sample of 2.5 kg contained 900 gr of cockle valves, the remaining material being mainly valves of mactras. In this accumulation faceting was observed in one every 14 shells, which is 7%.

faceting in cockles

faceting in mactra

In cockle valves faceting is more common than it is in valves of the other species in the shell mixture. From unbroken cockle valves 44% was faceted, versus 33% of the broken ones. The faceted cockle valves could be divided into three groups: those with a diameter of the hole up to 5 mm (=77%); those with a diameter of the hole of more than 5 mm (=13%); and those with a large hole, but broken (=10%).

Literature

Donovan, S.K 1991. The Process of Fossilisation. Belhaven, London.

Hollmann, R. 1968. Ueber Schalenabschliffe bei Cardium edule aus der Koenigsbucht bei List auf
Silt. Helgolander wiss. Meeresuntersuch.18: 109-193, Hamburg.

Lever, J. 1958. Quantitative beach research I. Basteria 22:21-51

Lever, J et al. 1962. Quantitative beach research II. Neth. J. Sea Research, 1: 339-358.

Lever, J. and Thijssen, R. 1968. Sorting phenomena during transport of shell valves on sandy
beaches; studies with the use of artificial valves. Symp. Zool. Soc. London, 22: 259-271.

Nagle, J. S. 1967. Wave and current orientation of shells. J. Sediment. Petrology, 37 (4): 1124-
38.

Pratje, O. 1929. Fazettieren von Molluskenschalen. Palaeontol. Zeitschr. 11: 151-169.

Richter, R. 1942. Die Einkippungsregel. Senckenbergiana, Frankfurt a. M., 25: 181-206

Schaefer, W. 1962. Aktuo-palaeontologie nach Studien in der Nordsee. Waldeman Kramer,
Frankfurt.

Schaefer, W. 1972. Ecology and Paleoecology of Marine Environments. Univ. Chicago Press.

Thijssen, R. 1971. Shells of Macoma balthica on sandy beaches. An example of quantitative
beach research. Thesis. Amsterdam.