Ostrea edulis beds in the central North Sea: delineation, ecology, and restoration

Abstract

Until the late 19th century, extensive beds of flat oyster Ostrea edulis populated the Central North Sea, which have vanished after intensive fisheries. At present, various initiatives are being carried out to investigate the potential to restore this former key species in the area. This historical ecological study contributes by delineating the former oyster bed area and through an assessment of its limits against known gradients in the North Sea. Extensive data from historical maps, texts, and ship-based surveys were used to synthesize our knowledge on the former beds. It was revealed that the area with oyster beds covered ∼6.2% of the total North Sea bottom, with a delineation that could partly be explained by hydrodynamic and temperature gradients. The position and extent of the area are notably different from the area that is used in recent feasibility studies on the restoration of North Sea oyster beds. The offshore oysters lived on muddy sand in relatively cold conditions, and there are several indications that their reproductive rate was low. The apparent disappearance of cold water adapted flat oysters will challenge restoration projects. This study provides indispensable information for the future restoration of flat oyster beds in the North Sea.

distribution, flat oyster, historical ecology, North Sea, Ostrea edulis, restoration

Introduction

Natural beds of European flat oysters Ostrea edulis were formerly abundant across many of Europe’s coasts and shallow shelf seas (Olsen, 1883; OSPAR, 2009; Pogoda, 2019). These beds provided key ecosystem services, allowing the existence of a hard substrate benthic community rich in species (Möbius, 1877; Coen et al., 2007; Houziaux et al., 2011; Kamermans et al., 2018b; Pogoda, 2019). The greatest part of these beds, however, disappeared in the late 19th and early 20th centuries, due to human influences (e.g. Gross and Smyth, 1946; Korringa, 1946, Yonge, 1960; Thurstan et al., 2013). Introduction of the Pacific oyster Magallana gigas in the 1960s and the arrival of the oyster parasite Bonamia ostreae in the 1970s led to a further decline. At present, oyster beds in the North Sea region are formed by the Pacific oyster, now the predominant oyster species. Commercial exploitation of O. edulis still takes place in mixed oyster beds in limited inshore sites, including the Limfjorden in Denmark (Nielsen et al., 2019) and Lake Grevelingen in the Netherlands (Maathuis et al., 2020).

In previous centuries, oyster beds were not only located along the coasts but further offshore at tens of metres depth as well. Extensive, offshore oyster beds were encountered in the central North Sea, in the English Channel, the Bristol Channel, and along the coast of Ireland (Olsen, 1883; Orton, 1926; Houziaux et al., 2008). However, the North Sea oyster beds, along with their associated hard substrate communities, had largely disappeared before World War I, as a result of intensive dredging (De Vooys et al., 2004; Berghahn and Ruth, 2005). Recently, attention for restoration of O. edulis oyster beds in the area has been increasing, much stimulated by the construction of windfarms in the North Sea (Gercken and Schmidt, 2014; Rijksoverheid, 2015; Kamermans et al., 2018a, b) and the establishment of marine protected areas (MPAs) (Schouten, 2018). In order to make informed decisions in future restoration projects as intended by Dutch and German governments (Kamermans et al., 2018b,  Pogoda, 2019), a better knowledge of the distribution of the North Sea’s historical oyster beds would be highly beneficial.

The size and delineation of the former areas within the North Sea that were covered by oyster beds have received limited attention in recent studies. Most restoration studies refer to one specific historical map on the distribution of oysters, drawn by Olsen and published in 1883 (e.g. Gercken and Schmidt, 2014; Kamermans et al., 2018a, b; Pogoda, 2019). However, Olsen (1883) also presented a different delineation of the oyster area in another map in the same work, and provided data that supported the latter in Olsen (1878) and (1885). As a result, the featured map cannot be regarded as an unambiguous source, and locations and limits of the oyster bed area indicated may be questionable. The literature both on the ecology of O. edulis (e.g. Korringa, 1952; Gosling, 2003) and on physical and ecological gradients of the North Sea seabed is extensive. Furthermore, early fisheries trawl surveys were carried out in the North Sea, well over 100 years ago (e.g. Garstang, 1905; Redeke, 1905–1911). Combined, this provides a unique opportunity to combine historical results with modern ecological knowledge.

The present, historical ecological study aims at assessing the former distribution of oyster beds within the North Sea. Historical trawl survey data, combined with the delineations of former oyster beds as drawn in 18th and 19th century maps and texts, are used to draw conclusions on plausible areas formerly covered by oyster beds. The plausibility of the reconstructions is further analysed by studying oyster fishery data and by assessing how O. edulis ecological tolerances fit inside the abiotic and biotic gradients of the North Sea. Subsequently, these ecological factors will be discussed in the context of restoration.

Methods

In addition to recent articles by De Vooys et al. (2004), Berghahn and Ruth (2005) and Houziaux et al. (2008), older British, German and Dutch literature was searched for maps and textual information on the offshore oyster beds in the North Sea. Additionally, data from ship-based research surveys in the North Sea during the late 19th and early 20th centuries were examined, including those onboard the German RV Pommerania in 1872–1873 (Metzger, 1875) and RV Poseidon (Terminfahrten, Fischereifahrten) in 1902–1912 (Schrader, 1911; Stein et al., 1990); the Dutch RV Wodan in 1902–1911 (Redeke, 1905–1911); and the British RV Huxley in 1902–1909 (Garstang, 1905; North Sea Fisheries Investigation Committee, 1909). The Wodan and Huxley research surveys were mainly targeted at fish, in particular, at commercially exploited species, but naturalists also collected data on benthic invertebrates sampled at the majority of locations, recorded onboard in so-called “Invertebrate logbooks”. Not all data collected during the Huxley surveys were published; therefore, for the years 1906–1909, sixteen of the original Invertebrate logbooks, that covered the relevant area, were consulted at the Centre for Environment, Fisheries & Aquaculture Science (Cefas) in Lowestoft, where these are currently held. Rijnsdorp et al. (1996) summarized the survey gears used in the Wodan and Huxley surveys (and the latter are also described in Garstang, 1905). The German surveys were specifically targeted at benthos (e.g. Callaway et al., 2007), and used a variety of sampling equipment (dredges and trawls), but unfortunately without much documented evidence on the durations of dredge or trawl hauls, or on the specifics of nets and mesh sizes used.

Books on historical fisheries (e.g. Benham, 1948, 1955) and information in newspapers were used to obtain more data on oyster bed exploitation and possible losses during the 19th and early 20th century. The online British Newspaper Archive (https://www.britishnewspaperarchive.co.uk/) and the Archive Service of the Royal Library of the Netherlands (https://www.delpher.nl/) were searched, using the search terms “oyster(s)” and “oyster bed(s)”. Additional 19^th-century Dutch and German texts on oyster fisheries were studied in the library of the Royal Netherlands Institute for Sea Research (NIOZ), Texel.

A methodological challenge is the definition of an oyster bed. OSPAR define oyster beds as “beds of the oyster O. edulis occurring at densities of 5 or more per m²” (OSPAR, 2009), whereas Berghahn and Ruth (2005) used a definition based on a minimum of 8 oysters per m². As such quantitative information was not available in the historical literature, all areas named “oyster beds” in historical sources were included here. Positions where large quantities of oysters were caught during research surveys were considered to indicate oyster bed areas.

The reliability of information on oyster bed locations was considered to be lowest in the historical maps, and highest where based on ship-based research survey data. In this study, areas delineating adjacent oyster beds were first determined on basis of the available literature and maps. In a next step, the areas designated in this way were checked against actual observations of oysters found in the historical research survey data. In an attempt to explain the limits of the resulting area, these were then assessed against biotic and abiotic gradients in the North Sea region. This was based on a set of covariates that are known to influence the growth, reproduction, and recruitment of O. edulis (Korringa, 1952, 1957; Yonge, 1960; Moore, 1977): food, temperature, suspended matter, depth, salinity, oxygen saturation, and bed shear stress.

Results

Review of written sources on historical North Sea oyster fisheries

Historical sources examined indicated that flat oyster fisheries in the central North Sea probably commenced in the 18th century and intensified after 1880 (Table 1 and Figure 1). After a period of peak catches from 1885 to 1895, the fishery continued until World War I (Table 1). First concerns about devastation of the oyster beds were expressed by the Dutch oyster expert, P.P.C. Hoek, who in 1894 declared that most oyster beds in the North Sea had disappeared (Het Nieuws van de Dag, 12 January 1894).

Figure 1.

Flat oyster fisheries in the central North Sea. English fishermen sailed on “skillingers” (Benham, 1948), Dutch on “botters” (Venema, 1865), and German on “Finkenwerder Kutter” (Berghahn and Ruth, 2005).

Table 1.

Historical oyster fisheries in the central North Sea.

18th century and first half of 19th century	Dutch, German and Danish oyster fishers were active in shallow waters of the Wadden Sea and the former Dutch Zuiderzee (now Lake IJsselmeer)	Paludanus (1776), De Vooys et al. (2004), and Lotze (2007)
19th century, second half	Occasional fishery from Zoutkamp (July to September) and Texel, probably Oosterend. Fishery from Finkenwärder in Germany (August to October)	Alan (1856), Venema (1865), and Metzger (1871); De Tijd 31 January 1881
Early 1880s—before World War I	Intensive trawling by English fishermen from Brightlingsea and later Grimsby. Up to 90 vessels	Deutsche Fischerei-Zeitung 31 May 1881; Nieuws van de Dag 13 June 1881; Olsen, (1885), Darmer (1894), and Benham (1948, 1955)
After 1890, until at least 1902	Increase in German oyster fishery in central North Sea and at the Helgoländer Bank	Hoek and Kyle (1905), Caspers (1950), Seaman and Ruth (1997), and Berghahn and Ruth (2005)
Early 20th century	Diminishing fishery, oyster beds became exhausted. No restart after World War I	Seaman and Ruth (1997) and Berghahn and Ruth (2005)

During the mid-1800s, catches could be as high as 1000 oysters taken in one haul (Metzger, 1871). Mortenson, a Norwegian researcher who sailed along, counted 90 English vessels in 1889, each of which returned to port laden with ∼25 000–40 000 oysters after 3 weeks out at sea (Deutscher Fischerei Verein, 1890). Based on these figures and the assumption that the fishing season was from August to November (De Vooys, 2001), Berghahn and Ruth (2005) estimated that the total yearly catch by these English fishermen was between 11 and 18 million oysters, but losses at sea indicate that the fishing season was not restricted to these months (Benham, 1955).

According to Hoek and Kyle (1905), German fishermen caught ∼1 million oysters a year from 1891 onwards and ∼2 million oysters a year from 1899 until at least 1902. However, German sources name higher numbers. Local fishermen from Helgoland only caught 2 million oysters in total over the period 1872–1886 at the “Helgoländer Bank” (Caspers, 1950). From 1885 to 1915, fishermen from Hamburg fished west of Helgoland, in the initial years catching 3–4 million oysters per year, but with annual catches after 1895 reduced to 1 million (Seaman and Ruth, 1997).

Dutch and German fishermen sold their oysters directly in Hamburg or Bremen (Möbius 1877; Hoek, 1886; Neudecker, 1990), or laid them out in beds in the Wadden Sea, either to increase their quality (Metzger, 1871; De Tijd 31 January 1881; Hoek, 1883) or to attempt to create new oyster beds (Leeuwarder Courant 23 August 1872; Hoek, 1883). The procedure to lay North Sea oysters in shallow beds in order to increase their taste was also practised in England (The Boston Guardian and Lincolnshire Independent 16 March 1901).

Delineation of oyster beds, based on historical data and maps

Late 19th century

Metzger (1873, 1875) noted that the German research surveys found oysters on mixed muddy-sandy bottoms with shells north of the East Frisian isles and the Dutch coast at 18–23 fathoms (33–42 m) depth, with an optimum between 21 and 23 fathoms (38–42 m). The oyster grounds were reported to extend from a thin stretch southwest of Helgoland, continuing from there in north-western direction until the longitude of the island of Terschelling (∼5½°E). Möbius (1877), Hoek (1886), and Olsen (1878, 1885) agreed with Metzger’s (1873, 1875) depth data. In particular, Möbius (1893) specified a strip of 15–20 km width at 32–42 m depth, starting 20 km southwest of Helgoland and extending in north-western and western direction until the longitude of Texel (∼4¾°E). In addition, the isolated Helgoländer Bank was situated southeast to east of Helgoland, at 25 m depth (Möbius, 1893; Caspers, 1950).

During the 1870s, O.T. Olsen created a simple data form and distributed this to fishermen of Hull and Grimsby, to collect information about their catch and bycatch; he presented the results at a meeting of the British Association for the Advancement of Science in Southampton in 1882. The Warwick & Warwickshire Advertiser & Leamington Gazette of 2 September 1882 reported: “The North Sea abounds of oyster beds of undreamt-of proportions and of astounding prolificacy. This ‘fish farm of Her Majesty’s subjects’ is stated by Mr. Olsen to contain oyster beds of 200 miles (370 km, assuming nautical miles were meant) in length and varying in breadth from 30 to 70 miles (56–130 km). The area represented by these measurements is about ten thousand square miles […], so that if the beds are only sparsely occupied with oysters, both consumers and dredgers have a rare prospect before them. But Mr. Olsen tells us that at one point, extending sixty miles in a north-westerly direction from Heligoland, the oysters lie very thickly on the beds”.

In his Piscatorial Atlas, Olsen (1883) presented two, somewhat different, maps showing the distribution of North Sea oysters. On one of these “the oyster map”, i.e. “Map 50” in Olsen (1883), both a “wider oyster area” is indicated which extends quite far south (i.e. until the coast of Terschelling) and a narrower area inside the former, “where caught in abundance” (here reproduced in Figure 2). On the “North Sea bottom” map, i.e. “Map 2” in Olsen (1883), the area with oysters is larger and extends much further to the north. In addition, both maps in Olsen (1883) also show a separate oyster area at the slopes of the Outer Silver Pit, a deep gully south of the Dogger Bank (Figure 2). Olsen’s (1878) and (1885) texts are in favour of the more northern distribution in “Map 2”. In words of Olsen (1885): “Abreast of Terschelling and off all the islands is fine sand […] Further off, in a depth of 20 fathoms, mud and oysters will be found”. In a section on North Sea oyster grounds, he adds: “It (the bed) stretches across to Terschelling, 25 miles distance from the land. […] The bed extends easterly, passing outside Borkum Reef, down to Heligoland deep”. Furthermore, Olsen (1878, 1885) also mentioned oyster fisheries at Botney Gut, which commenced after 1841.

Figure 2.

Delineation of oyster beds in the central North Sea according to late 19^th-century maps and texts from the following sources: Metzger (1875); Möbius (1877); Hoek (1883); Olsen map 2 and 50 (1883); Olsen (1878), Olsen (1885), Darmer (1894).

Darmer (1894) indicated a more northern extension of the area: the western border approximately followed the 3°45′E longitude between latitudes 53°40′N and 54°45′N, then turned to ENE until 4°E and to the NE until it reached 5°40′E (at ∼55°30′N). From the southwestern point to Terschelling, the area followed the 38 m depth line. The distance from Terschelling to the grounds was 30 miles. In easterly direction, the line continued north of the Bornholm reef grounds, and formed a flat curve towards Helgoland. Darmer (1894) also referred to O.T. Olsen as his source for the position of two smaller oyster beds, one at N 54°9′N, 2°15′E (Outer Well Bank) and the other at 54°55′N, 4°53′E (Figure 2).

Fishermen from Brightlingsea reportedly trawled for oysters at a depth of 20 fathoms (36 m) off “Terschelling Light” (Benham, 1948), the lighthouse on Terschelling. Probably, they also benefitted from the lightvessel Terschellingerbank, which came into service in 1881 and was positioned at 53°33′0″N and 4°54′30″E, in 21 m deep water at low tide (Ministerie van Marine, 1881). The Liverpool Mercury of 16 July 1887 published an article about fishermen from Cleethorpes (Grimsby) who sailed 180–200 miles (290–321 km, provided statute miles were meant) in order to trawl oysters in an area of 200 by 75 miles (321 by 120 km) at 20–30 fathoms (37–55 m) depth. And in 1894, Dutch newspapers (Het Nieuws van de Dag 13 October 1894; Rotterdams Nieuwsblad 15 October 1894) mentioned fishermen from the Dutch island Texel and from England, visiting an extensive oyster bed at 40 nautical miles (74 km) NW of Terschelling. From these reports, we may conclude that the intensive oyster trawling by English fishermen was carried out in the area that is currently still called the Oyster Grounds.

Darmer (1894) mentioned an oyster area that stretches from east to west over 150 nautical miles (278 km), with a width of 30–70 miles (56–130 km), located 25–30 miles (46–56 km) from the coast.

Early 20th century

Several 20^th-century sources from Germany provided maps of locations where their sailing boats fished for oysters (Backhaus, 1906; Lübbert, 1906 in Berghahn and Ruth, 2005; Fischenkarte, 1915 in Gercken and Schmidt, 2014). Moreover, Kyle (1905) depicted a map indicating the oyster area, which is similar to that in Olsen’s (1883) Map 2. Combined, these post-1900 maps from German and Dutch sources are summarized in Figure 3.

Figure 3.

Maps on oyster beds in the central North Sea during the period 1900–1915, based on the following information sources: Kyle (1905); Backhaus (1906); Lübbert (1906); and Anonymous (1915) in Gercken and Schmidt (2014).

The data from the 19^th- and early 20^th-century maps (Figures 2 and 3) were combined to draw the probable oyster area as documented cartographically and in texts at the time, and this has been presented in Figure 5 (line a).

Data from ship-based English, Dutch, and German research surveys carried out between 1872 and 1912 (Figure 4, more details in Methods) generally confirmed the area mentioned in the literary sources and maps. The ratios of (any) oyster pro haul inside the oyster area derived from the maps (Figure 5, line a), were 5/7 for RV Pommerania 1872–1873, 3/27 for RV Poseidon 1902–1912, 11/15 for Wodan 1902–1911, and 14/26 for RV Huxley 1902–1909 (Figure 4). The relatively low oyster catch rates in hauls by RV Poseidon may be attributable to the particular focus on fish in these surveys. In most cases where dredges or Kurre gear were used, the duration of the hauls is unknown; as the survey also used a Grosse (large) Kurre, the Kurre were probably relatively small compared to the beam and otter trawls used on the vessels Wodan and Huxley.

Figure 4.

Observations of flat oysters encountered in trawl and grab surveys by British (Garstang, 1905; North Sea Fisheries Investigations Committee 1909; “Invertebrate logbooks”, see Methods), Dutch (Redeke, 1905–1911), and German surveys (Metzger, 1875; Schrader, 1911; Stein et al., 1990) carried out between 1872 and 1912.

Figure 5.

Delineation of area with oyster beds within the central North Sea, as derived from historical texts and maps (line a) and after inclusion of historical ship-based research survey data (line b).

In particular, larger numbers of oysters were encountered most frequently between latitudes 53°30′N and 55°N, and longitudes 4°E and 7°E. Notably, surveys by RV Huxley recorded many oysters between 54°30′N–55°N, and 4–5°E, to the west of the area derived from historical texts and maps. Moreover, survey results from RV Huxley in the north-west, and from RV Wodan in the southeast, indicated that the historical oyster area was somewhat larger than suggested by the historical texts and maps (Figure 5).

Concluding from all sources mentioned above, Olsen’s “bottom map” (Map 2), although missing some of the northern banks, gave a better representation of the oyster grounds than his “oyster map” (Map 50). Figure 5 shows the contiguous area wherein historical sources depicted or described several oyster beds. The southern border from Helgoland to Terschelling was drawn on basis of the 33–34 m depth line described by Metzger (1875), Möbius (1893), and Hoek (1886), which is close to 20 fathoms (36.5 m) mentioned in Olsen (1885).

Differences in oyster densities within the area remain poorly known. According to Metzger (1873), highest densities were to be found between 38 and 42 m depth, where “1000 oysters could be collected in one haul with a large oyster net”. Moreover, a bed with high-density area of ∼18 km breadth and 40–110 km length was reported to be situated NW or WNW of Helgoland (Möbius, 1877; Olsen, 1885; Darmer, 1894). This area should not be confused with the isolated Helgoländer Bank situated ESE of Helgoland. Intensive fishery by English vessels and later survey data (see late 19th century) confirm that the Oyster Grounds, situated north-west of Terschelling, comprised an extensive high-density area as well.

Historical sources concerning the ecology of the North Sea oysters

Shape and reproduction

The deep-sea flat oysters in the North Sea were found on stable, silt-rich sand, mixed with dead shells; often, they formed clumps containing 3–5 oysters (Möbius, 1877, 1893). The oysters were relatively large, wide, and thick; specimens up to 135 mm width, 118–125 mm length, and 32 mm thickness were no exception (Metzger, 1875). The shells were relatively spherical and light, with much space between the layers (The Liverpool Mercury, 1887). The meat was reportedly tough and salty, with a sharp aftertaste (Möbius, 1877, 1893). According to Orton (1928), deep-sea oysters were probably large and “fat” because they invested less in reproduction as a result of the low water temperature. According to Hagmeier (1916), two-thirds of shells on the Helgoländer Bank were thicker than 31 mm. He suggested a low reproduction frequency at this bank: only 2% of the oysters were “Junggut”, compared to 40% on the shallow beds in the Wadden Sea (Möbius, 1877).

North Sea oyster beds as a larval source for the Wadden Sea

The hypothesis that oyster beds in the Wadden Sea depended on larval import from the North Sea beds has a long history. Metzger (1871) argued that in water of 25 m and less, the tidal current switched from W-E to E-W, depending on the wider current conditions, while at the bottom, the water continually moved from W to E. He suggested that North Sea larvae would not reach the German Wadden Sea coast west of the Helgoland Bight, and that oyster beds in this area therefore would have been man-made. The easternmost area of the Helgoland Bight would have benefitted from the larval input and the oyster beds, and therefore in this area, Metzger argued that oyster beds would have been natural. Möbius (1877), however, who studied the latter area, concluded that the fertility of the oysters was high enough to retain the local population. In 1900, Hoek found oysters at the Meep, a channel in the Wadden Sea located southwest of Terschelling, which were very much alike the North Sea oysters, with their wide, “winged” appearance (Hoek, 1910). This led him to conclude that oysters in the Meep grew from spat derived from either the North Sea beds or the shallower, “rough area” north of Texel and Terschelling. Caspers (1950) postulated that the oyster beds at the Helgoländer Bank required larval import from the coast or from the North Sea beds. Berghahn and Ruth (2005) concluded: “It will probably never be possible to answer the question to what extent European oysters in the Wadden Sea have been dependent on the inflow of larvae from the huge oyster stocks in the North Sea”.

Epifauna

Möbius (1893) published lists of epifauna inhabiting O. edulis beds in the Wadden Sea, at the Helgoländer Bank, and at the “deep-sea” oyster beds of the central North Sea. Samples from the deep-sea beds were collected at five different locations by the Pommenaria survey on 23 and 24 August 1872, north of Terschelling; 80 invertebrate species were recorded. Tesch (1910) reported 17 species found on or closely associated with flat oysters on the Oyster Ground. Möbius’ (1893) and Tesch’ (1910) species data are summarized in a Supplementary Table S1. Möbius (1893) was aware that his list was far from complete, but concluded the epifaunal diversity to be higher in the central North Sea oyster beds than on the Helgoländer Bank, and even more so than in the Wadden Sea oyster beds. About 40% of the 67 benthic species may be classified as associated with hard substrates (Supplementary Table S1). Most of these species can still be found on the Cleaver Bank, a mixed-bottom area south of the Dogger Bank investigated recently (van Moorsel, 2003). As Korringa (1954) recorded over 250 species on flat oyster shells in the Eastern Scheldt, we may suspect that the epifauna was richer than these incomplete data suggest.

Factors associated with the general distribution of O. edulis

Several ecological factors were associated with the delineation of the oyster area. The dominant factors for growth in bivalves are feeding conditions and temperature (Grant et al. 1990; Gosling, 2003).

Ostrea edulis is an active filterer which feeds on phytoplankton (especially diatoms), bacteria, detritus, and suspended organic matter (Korringa, 1952; Yonge, 1960). In addition to plankton in the water column, O. edulis profits from resuspended organic material from the bottom (Grant et al. 1990; Gosling, 2003).

Oysters grow during the warmer months of the year; in the Eastern Scheldt, a minimum temperature of 15°C was needed for growth (Korringa, 1952). Optimum temperatures for growth are higher, and oysters in Danish waters showed an optimum filtration efficiency at 20°C (Newell et al., 1977). Flat oysters from coastal areas are tolerant to cold. British, Dutch, and Danish flat oysters are able to survive several weeks at −1.5°C (Korringa, 1952). Anecdotal information seems to indicate that the deep-sea oysters from the North Sea were more sensitive to low temperatures: most of the North Sea oysters laid on a local bed near the Wadden island Texel were frozen to death in a severe winter, while almost all the local oysters survived (De Tijd 31 January 1881).

Korringa (1957) illustrated that various European O. edulis populations differed in the seawater temperature at which they commenced breeding, with oysters from the Mediterranean, France, England, and the Netherlands spawning when seawater temperatures reached 15–16°C and normal larval development beginning at 17.5°C. Deep-sea oysters in the English Channel and in the Firth of Forth (Scotland) lived in water that seldom exceeded 15°C. Yonge (1960) assumed that these oysters released their spawn at 12–13°C. The oysters in the North Sea also lived in relatively cold water, and seawater at the Helgoland Bank only reached 16°C in warm summers during the years 1872–1906 (Hagmeier, 1916). Under cold water conditions, oyster larvae developed slower and settled later, decreasing their survival rates (Korringa, 1957).

Growth is also negatively influenced by a prolonged exposure to a high quantity of suspended matter: increase in suspended sediment leads to a decrease in filtration rate (Korringa, 1952; Hutchinson and Hawkins 1992; Moore, 1977) and reduced growth (Grant et al., 1990). Spat fall is sensitive to the abrasive effect of the combination of current and sand (Moore, 1977).

Other factors, mentioned in literature to affect flat oyster growth and distribution, are unlikely to have been limiting in the central North Sea area before 1900: (i) depth, given that the species’ normal distribution range is up to 80 m deep (Laing et al., 2005), (ii) salinity, and (iii) oxygen saturation of the seawater, which was mostly above 90% (Gehrke, 1916), well within the tolerance ranges of flat oysters.

Conditions in the North Sea between N 53° and 56°

The area within the North Sea area where the deep-sea oysters occurred lies east of the Dogger Bank and west of Denmark. The depth increases gradually from the islands of the Wadden Sea to ∼50 m. At depths shallower than 30 m, where the influence of tidal currents or waves is large, the bottom consists of sand. In deeper water, muddy sand dominates. Deposition of organic material coincides with that of bottom particles: Creutzberg et al. (1984) found a strong correlation between organic material (plankton, bacteria, and detritus) and mud content at the Frisian Front. New mud is transported and deposited into the area, originating from the eastern coasts of England and from the Elbe river in Germany, bringing in nutrients that favour algal blooms.

In this area, mean bottom temperatures in the years 1997–2002 ranged from 5 to 7°C in winter, depending on location, and in summer from north to south ranged from 14 to 18°C (ICES, 2008). When summer stratification takes place, this occurs in water of 30 m depth and deeper. Below the thermocline, the seawater seldom reaches 16°C (Witbaard and Bergman, 2003; van Leeuwen et al., 2015). Summer stratification will be discussed further in the next paragraph.

Factors associated with historical distribution of O. edulis in the North Sea

The flat oyster area inferred from historical data lies in an area of muddy sand (Figure 6a), nowadays inhabited by a macrobenthos cluster dominated by the ophiuroid (brittle star), Amphiura filiformis (e.g. Meyer et al., 2018). In this section, ecological factors that may explain the historical distribution of flat oysters in the same area are synthesized.

Figure 6.

Delineation of the historical flat oyster bed area, overlaid on maps showing different abiotic gradients in the North Sea. (a) Sediment mud content, calculated as the relative fraction with grain sizes <63 μm (or mud %); adapted from Bockelmann et al. (2018). (b) Mean wind and tide bed shear stress, 1997–2017; adapted from Wilson and Heath (2019). (c) Mean summer bottom temperature (°C), 1997–2002; adapted from ICES (2008). (d) North Sea hydrodynamic regions; adapted from van Leeuwen et al. (2015), with addition of the term Transitional East introduced by Capuzzo et al. (2018).

The southern and eastern limits of the oyster area coincide with the present transition zone from muddy sand at depth, to sand nearer to the coast (Figure 6a). As Olsen (1885) described a change from fine sand to mud and oysters at 20 fathoms depth (from Terschelling eastward), it may be assumed that this transition zone is relevant to the southern border of the oyster bed area. Creutzberg and Postma (1979) studied the same region and found this transition zone generally located close to the 30 m isobath. Sand and silt are kept in suspension by tidal currents (Creutzberg and Postma, 1979) and by kinetic energy of waves; in this area, Creutzberg and Postma (1979) found that a maximum surface velocity <0.9 knots (0.46 m/s) coincides with the distribution of silty elements. Oysters do not occur on bottoms in motion, and sandy bottoms are not suited for a successful spat fall (e.g. Möbius, 1877; Laing et al., 2005). Thus, the historical flat oyster beds were in an area characterized by relative low mean bed shear stress (Figure 6b).

The northern limits of the historical flat oyster area appear to be associated with the summer bottom temperature (Figure 6c). The summer thermocline is responsible for low bottom temperatures that decrease in northward direction (Elliott and Clarke, 1991) (Figure 6c). The importance of summer water temperatures near the northern fringe of the oyster’s range is discussed by Spärck (1950) and Korringa (1952). As the isotherms lie in a southwest–northeast direction, the summer temperature may also account for the western border of the oyster area.

The largest part of the historical oyster area lies in a region where summer stratification varies between years (Figure 6d; van Leeuwen et al., 2015), the so-called “Transitional East” area in Capuzzo et al. (2018). The resulting low temperatures could account for the low reproductive investment of the oysters suggested by Orton (1928; par. 3.3.1). Although during summer stratification bottom water oxygen values on the Central Oyster Grounds may drop to 5.2 mg/l for short periods (Greenwood et al., 2010), this value is still well above the minimum required for oyster growth (0.5 mg/l; Kamermans et al., 2018b).

Discussion

This study has documented the extensiveness of central North Sea flat oyster beds during the late 19th century and the early 20th century, in an area covering ∼35 350 km² (representing 6.2% of the North Sea). At the time, the oyster beds yielded rich catches which were, however, largely depleted before the outbreak of WWI. Contemporary sources documented that a rich epifauna was associated with these oyster beds.

The historical records on oyster sizes and the formation of clumps consisting of 3–5 oysters fixed together (Möbius, 1877, 1893) are in line with the description of Houziaux et al. (2011) of beds consisting of “large amounts of old specimens (>20 years), cultch (old valves) and new settlers which probably necessitate decades or centuries to build up”. A broad range of historical sources indicate that the oyster fishery between 1880 and 1914 was so intense that the beds rapidly lost their commercial value. Natural recovery of oyster beds was prevented by the major expansion of large-scale trawl fisheries in the North Sea during the late 19th and early 20th centuries (Alward, 1911; Callaway et al., 2007; Engelhard, 2008) and the high trawl footprint since (Engelhard, 2005; Eigaard et al., 2017). The associated epifauna lost the hard substrate it depended upon, and species associated with oyster beds disappeared. This is likely to have had many ecosystem consequences, including the loss of any holdfast for the egg cases of skates and sharks, a lack of suitable hiding or nesting places for a range of crab and other invertebrate species, and a loss of structured, stable habitat as required for many benthic fish species, especially the younger life stages (Callaway et al., 2007; Sguotti et al., 2016). The impact of the altered food situation, reduced three-dimensional structure, and potential changes in visibility to larger demersal fishes including commercial species is uncertain.

The disappearance of the North Sea oyster beds may in turn have led to reduced recruitment of larvae into the Wadden Sea beds, a notion that has been discussed for almost 150 years (Metzger, 1871; Berghahn and Ruth, 2005). Following the disappearance of the oyster beds, decades of trawling for food fish have further reduced the number of long-lived, slow-growing benthic invertebrate species, such as the bivalve species, Aequipecten opercularis, Arctica islandica, and Modiolus modiolus (Callaway et al., 2007).

Reproduction of O. edulis still takes place in the southern North Sea and eastern English Channel, suggesting that there is potential for natural recovery of flat oyster beds in the southern North Sea (Kerckhof et al., 2018). However, whether the same oysters could be used to repopulate the central North Sea is questionable, given the historical evidence that the “deep-sea” North Sea oysters reproduced at low temperatures (Korringa, 1957, 1969), and that the populations which historically lived in waters not exceeding 15°C, do no longer exist (e.g. Thurstan et al., 2013; Fariñas-Franco et al., 2018). Ostrea edulis reared in Scottish, Dutch, and Danish aquaculture settings, and Norwegian oysters collected for Dutch restoration projects, tend to live in seawater that reaches at least 17°C in summer.

The debate as to whether genetic homogenization may have led to a loss in adaptability of European oyster populations, as first suggested by Gross and Smyth (1946) and Korringa (1969), has led to various genetic studies. Vera et al. (2016) concluded that there is still significant genetic differentiation within North Sea populations, with the presence of regional genetic structure and potential for local adaptation of O. edulis; the authors suggested caution when transplanting individuals, especially between distant geographical regions. To minimize the temperature challenge, it is advisable to start with restoration in the central North Sea in the southern part of the former area, from fairly nearby sources.

In its broad outline, this study has highlighted an area within the North Sea where projects with the intention to recover the seabed with O. edulis would be expected to have the greatest likelihood of being successful. It is of note, however, that hydro-climatic conditions in the North Sea have changed, principally since the late 1980s, especially with regard to sea surface temperature and primary production, and hence, food supply (Kröncke et al., 2011; Capuzzo et al., 2018). Thus, conditions may have changed to an extent that the area which historically favoured oyster bed production may not be representative of the present situation. Wilson & Heath (2019) found an increase of bed shear stress by over 20% in the eastern North Sea since 1971; this increase suggests that conditions favourable for flat oysters may now be positioned somewhat further from the coast.

Further uncertainty lies in the impacts of intensive trawling on the seabed sediments, as availability of suitable substrate material is crucial for oyster larvae settlement (Smyth et al., 2018). In this context, it is remarkable that several locations which Kamermans et al. (2018a) considered less suitable for self-sustaining O. edulis beds due to the sediment being too silty lie in close proximity to locations where at the beginning of the 19th century, surveys by RV Huxley observed large quantities of oysters on sandy ground. Moreover, the sediment composition in the oyster bed area may also have changed as a result of the disappearance of the oysters themselves. The filtering capacity of oysters does not only improve water clarity, but also changes sediment characteristics in reef areas (Southwell et al., 2017). In present and future windfarm areas, further changes in local seabed conditions may be expected due to the drag from the foundation structures, the resulting bed shear stress, and other factors yet to be quantified (Carpenter et al., 2016). Thus, by and large, while the delineation of the historical flat oyster beds gives us an indication where restoration might be feasible, due consideration should be given to the present hydrodynamic and seabed characteristics (see also Kamermans et al., 2018a).

Several recent studies on flat oyster restoration have referred to Olsen’s (1883) “Oyster Map”. An outline of the historical oyster bed area further from the German and Dutch coast affects the outcome of these studies. Although conditions in the North Sea changed (Kröncke et al., 2011), we believe that the gradient associated with the historical southern boundary is marked enough to be taken into account in current site selection. Gercken and Schmidt (2014) recommended restoration of offshore oysters in shallower waters (<20 m depth), as temperatures required for reproduction are reached earlier at these locations. While agreeing on the importance of temperature, our results suggest that conditions in these shallow, sandy areas characterized by high bed shear stress are less favourable for flat oyster restoration. Pogoda (2019) reported on restoration trials in the Natura 2000 site, Borkumer Riffgrund, at depths of 20–40 m. Olsen (1885) stated that the oyster beds, at the time, extended well beyond Borkum Reef. However, the probability exits that his informants avoided the coarser grounds in this area. In the study by Kamermans et al. (2018b), clear criteria for site selection are provided which generally concur with the descriptions of abiotic factors provided in the present study. There was, however, some disagreement in that Kamermans et al (2018b), based on model calculations and Olsen’s (1883) historical “Oyster Map”, concluded that areas with sea bed shear stress <0.6 N/m² are suitable for flat oyster development. The delineation of the oyster bed area in the present study, overlaid on the same bed shear stress map is suggestive of an “oyster area tolerance limit” of <0.3 N/m². Again, the present study suggests higher restoration potential in the more northerly part of the area, but it must be added that much smaller offshore O. edulis beds also occurred in the south near the coast of the province Zeeland. Research in nearby Belgian waters (Houziaux et al., 2008, 2011) found that oyster beds occurred in sandy gravel fields.

Natural restoration of oyster beds is only conceivable after formation of a stable substrate facilitated by a long-term succession or through provision of hard substrate. After the realization of the importance of historical baseline data for setting long-term goals, the OSPAR Commission included North Sea O. edulis beds as a priority marine habitat for protection in European MPAs (OSPAR Commission, 2012; Fariñas-Franco et al., 2018). In Germany, the Sylt Outer Reef Natura 2000 site (5314 km²) includes parts of the east of the historical flat oyster bed area. Moreover, the Government of the Netherlands is in planning stages for three MPAs, to be located in the Central Oyster Grounds (2000 km²), on the Frisian Front (1600 km²), and southeast of the Frisian Front (400 km²). All of these areas include parts of the historical oyster bed area. In each of these cases, the European Commission was asked to approve management measures aimed at restoring seabed integrity (Schouten, 2018). Provided effective measures are taken, new opportunities for a natural succession towards oyster beds may arise.

Olsen’s (1883) “Piscatorial Atlas”, including the many maps describing North Sea fishing grounds, probably prevented that biologists and managers were unaware of the former oyster beds, as was the case in South Australia (Alleway and Connell, 2015). Nevertheless, the map led to an underestimation of the former extension of the beds; likewise, along the Scottish North Sea coast, the former extensiveness of oyster beds was only recently apprehended (Thurstan et al. 2013; Fariñas-Franco et al., 2018). While Olsen’s (1883) maps are relatively well known and accessible, the present study is an example of a search in historical data that yields more detailed information about former populations (Thurstan et al., 2013; Engelhard et al., 2016; Fariñas-Franco et al., 2018), information that enables researchers to introduce long-term perspectives to research and management advice.

Supplementary data

Supplementary material is available at the ICESJMS online version of the manuscript.

Acknowledgements

This study was facilitated through the ICES Working Group on the History of Fish and Fisheries (ICES WGHIST). We are indebted to the library facilities of the Centre for Environment, Fisheries & Aquaculture Science (Cefas), Lowestoft, and the Royal Netherlands Institute for Sea Research (NIOZ), Texel. Special thanks to Mary Brown (Cefas) and Sven Ballenthin (Forshungsbibliothek Gotha) for their help with rare historical data. Finally, we wish to thank the two anonymous reviewers for their critical reading and helpful comments.

Data availability statement

The data underlying this article will be shared on reasonable request to the corresponding author.