Biodiversity & Conservation

Beds of dead maerl

Dead maerl beds


Beds of dead maerl
Distribution map

Dead maerl beds recorded (dark blue bullet) and expected (light blue bullet) distribution in Britain and Ireland (see below)


  • EC_Habitats
  • UK_BAP
  • OSPAR

Ecological and functional relationships

The biodiversity and ecological structure of maerl beds is summarised by Birkett et al. (1998). Grall et al. (2006) used carbon and nitrogen isotope analysis to examine the trophic relationships within maerl beds in the Bay of Brest. The biodiversity of maerl beds (and some ‘dead’ maerl beds) is shown by Birkett et al. (1998), Cabioch (1968), Hall-Spencer (1998), Kamenos et al. (2004b), Kamenos et al. (2004c), Rostron (1985). Live and dead maerl beds can support similar communities (see section 2 above). However, any species or ecological functions likely to be specific to live maerl are omitted.

Dead maerl provides a substratum for the attachment of epiflora and epifauna, and a range of interstices for mobile epifauna, and shallow burrowing infauna. The variable and open structure of the maerl sediment can also provide good oxygenation at depth and allows many species to burrow deeply into the maerl substratum, while other deep burrowing species (e.g. mud shrimp) can also occur.

  • Primary productivity is provided by epiphytic macroalgae or microphytobenthos (e.g. benthic diatoms) growing on the maerl thalli at the surface of the bed, together with deposited phytoplankton and particulate organic materials.
  • Photosynthetic macroalgal epiphytes are likely to include red algae e.g. Gracilaria spp., Ceramium spp., Polyides rotundus, Dictyota dichotoma, green algae e.g. Ulva spp., Cladophora spp. and brown algae e.g. Chorda filum, although the actual species present will vary between sites.
  • Filter feeders can be divided into those that feed from the water column, those that feed within the surface layer of the bed (epifaunal) and those that feed at the interface between the surface of the bed and deep sediments, e.g. the burrowing bivalves.
  • The community may be dominated by a large number of filter feeding species including sponges (e.g. Scypha ciliata, Suberites spp. and Halichondria panicea); hydroids (e.g. Obelia geniculata); anemones (e.g. Metridium senile); polychaetes (e.g. the fanworm Sabella pavonica, the parchment worm Chaetopterus variopedatus, and keel worms Pomatoceros spp.); decapods (e.g. Pisidia longicornis); molluscs (e.g. Pecten maximus, Crepidula fornicata); bryozoans (e.g. Bugula spp., Scrupocellaria scruposa); echinoderms, and sea squirts (e.g. Ascidiella aspersa, Botrylloides leachii).
  • Epifaunal and infaunal surface deposit feeders include polychaetes (e.g. Notomastus latericeus) and some crustaceans (e.g. Apseudes latreilli and Athanas nitescens) (Grall et al., 2006). Burrowing polychaetes also feed on organic material in the finer sediments present within the bed.
  • Surface grazers include chitons and gastropods (e.g. Bittium reticulum and Gibbula cineraria) that feed on benthic diatoms, biofilms and young macroalgae and hydroids growing on shells, pebbles and maerl thalli.
  • Predators include demersal fish, star fish (e.g. Asterias rubens), crabs (e.g. Liocarcinus spp. and Cancer pagurus), and gastropods (e.g. the common whelk Buccinum undatum) as well as infaunal carnivorous polychaetes (e.g. the eunicids) where present.
  • Omnivores and scavengers include gastropods (e.g. Nassarius spp., Buccinum undatum), crabs and starfish.
Grall et al. (2006) concluded that the majority of the biomass (in Bay of Brest live maerl beds) was represented by interface filter feeders, small carnivores and epifaunal deposit feeders. Nunn (1993) and Hall-spencer (1998) point out that maerl bed can host a diverse number of infaunal bivalve molluscs, which can represent an extremely high biomass (Hall-Spencer et al., 2003).

Seasonal and longer term change

Birkett et al. (1998) noted that there was considerable variation in maerl flora and fauna even within the same bay system (e.g. Galway Bay or Baie de Morlaix) and that algal abundance changed between winter and summer. In the Mediterranean, the maerl bed epifloral diversity doubled over summer. In Galway Bay, maerl bed algal diversity increased in summer, in part due to greater stability of the bed surface in summer. Comparison of the two sites showed that the increase in algal cover was greatest at the shallow site (5 m) than at the deeper site (10 m). Prostrate macroalgae may stabilize the maerl beds in summer by the formation of stolons and secondary attachments (Birkett et al., 1998). Where they occur, byssus forming bivalves (e.g. mussels, flame shells) may also stabilize the surface of the bed. Where present, the gravel sea cucumber (Neopentadactyla mixta) migrates deeper within the bed in winter months where is remains in a state of torpor until spring (Smith & Keegan, 1985). In areas exposed to wave action, and especially winter storms, the surface of the bed may be mobilized, and form ripples and mega-ripples (Keegan, 1974; Birkett et al., 1998; Hall-Spencer 1998; Hall-Spencer & Atkinson, 1999). Storms and wave action may result in a reduced epiflora and epifauna, and provide space for more opportunistic epiflora (Birkett et al., 1998), but is unlikely to affect deep burrowing fauna (e.g. some bivalves and mud shrimp) (Hall-Spencer 1998; Hall-Spencer & Atkinson, 1999). In the long term established maerl beds are known to be extremely old, and carbon dating suggest that some beds may be between 4000 and 6000 years old (Birkett et al., 1998). Bosence & Wilson (2003) calculate the maximum age of the St Mawes Bank, Falmouth, to be 4000 years.

Habitat structure and complexity

Nodules of maerl provide a loose, open structure with numerous interstices of varying size for a wide range of species to dwell and feed within, while also providing a hard substratum for attachment by epiflora and epifauna. Maerl beds usually include dead shell within the maerl matrix that provides additional substratum. However, maerl substratum is highly variable in terms of depth of the maerl bed itself, patches of different substratum (gravels, sands, muds etc.) within the bed, and the nature of the underlying sediment. As shown above, Grall et al. (2006) identified a number of different microhabitats and trophic levels. Overall, it is the loose open and heterogeneous structure of the bed that provides the diversity of microhabitats and hence diversity of the biological community (Birkett et al., 1998; Kamenos et al. 2003; Grall et al., 2006). The open structure also allows water flow to penetrate deep into the bed allowing for species to occur at considerable depth within the maerl matrix. While dead maerl does not provide the structural heterogeneity as live maerl, it was similar to gravel habitats, and so more open than other sediments. For examples, in Scottish beds, rayed artemis Dosinea exoleta was recorded at a maximum depth of 42 cm, the razor shell Ensis arcuatus at 48 cm, and the gaper clam Mya truncata at 56 cm (Hall-Spencer, 1998). The mud shrimp Upogebia deltaura can form burrows down to 68 cm (Hall-Spencer & Atkinson, 1999), while the Neopentadactyla mixta can spend winter at depths of 60 cm in maerl (Smith & Keegan, 1985). The structural complexity is augmented by the presence of tube-building species, such as the mud shrimp and burrowing anemones (Cerianthus lloydii) and parchment tube worms (Chaetopterus variopedatus), whose burrows allow oxygenated water to penetrate the bed and also provide habitats for interstitial and commensal species (e.g. Mysella bidentata). Dead maerl is presumed to be more fragmented, as is impacted (dredged or extracted) ‘dead’ maerl. Therefore, it is likely to have a lower heterogeneity than pristine maerl beds (Grall et al., 2006) but nevertheless still support a diverse fauna and flora (Rostron, 1985; Hall-Spencer 1998; Birkett et al., 1998).

Productivity

Dead maerl beds lack the primary productivity of the maerl itself. Grall et al. (2006) estimated that macroalgal productivity of epiflora was 135 g/m2 in the Bay of Brest, and recorded 70 g (AFDW) or 11000 individuals per m2 for macrofauna (small invertebrates) and 100 individuals per m2 (11.3-34.8 g AFDW/m2) for megafauna (large invertebrates). In Galway Bay, Bosence (1979; cited in Birkett et al., 1998) reported invertebrate abundances ranging from 10 individuals per 0.25 m2 of Hiatella arctica to 270 individuals per 0.25 m2 for Bittium reticulum. However, the productivity of individual beds is likely to be variable.

Recruitment processes

Dead maerl is by definition unable to recruit or replenish. Passive migration of live maerl nodules or dead maerl fragments is possible from adjacent beds (if present) due to storm or wave activity, or via attachment to macroalgae moved by wave and currents (Birkett et al., 1998). A diversity of species groups may be found in maerl beds (live or dead), and recruitment in a number of example species groups and species is given below.
  • Vadas et al. (1992) reviewed recruitment and mortality of early post settlement stages of benthic algae. Grazing, canopy and turf effects were the most important factors determining recruitment and settlement but that desiccation and water movement may be as important for the early stages. He indicated that recruitment is highly variable and episodic and that mortality of algae at this period is high. Chance events during the early post settlement stages are therefore likely to play a large part in survival.
  • The propagules of most macroalgae tend to settle near the parent plant (Holt et al., 1997; Norton, 1992; Schiel & Foster, 1986). For example, the propagules of fucales are large and sink readily and red algal spores and gametes are immotile. Norton (1992) noted that algal spore dispersal is probably determined by currents and turbulent deposition (zygotes or spores being thrown against the substratum). For example, spores of Ulva sp. have been reported to travel 35 km and Phycodrys rubens travel 5 km. The reach of the furthest propagule and useful dispersal range are not the same thing and recruitment usually occurs on a local scale, typically within 10 m of the parent plant (Norton, 1992). However, many of macroalgae also have heteromorphic life histories that include a microscopic gametophytic or sporophytic stage that may itself be more tolerant (or less, depending on species) of environmental change and function in part like a ‘seed bank’.
  • Guillou & Sauriau (1985) investigated reproduction and recruitment in a Venus striatula population in the Bay of Douarnenez, Brittany. There were 2 periods of spawning activity, one in the spring and then again in late summer. The larvae undergo planktotrophic development, metamorphosis occurring 3 weeks after fertilization (Ansell, 1961; cited in Guillou & Sauriau, 1985). There were 2 periods of recruitment, one at the end of the spring and the second in autumn. The mean life span was 5 years and the maximum 10 years. No evidence was found to suggest that recruitment patterns for the other venerid bivalves differed significantly.
  • Dauvin (1985) reported that the oval venus Timoclea ovata (studied as Venus ovata) recruitment occurred in July-August in the Bay of Morlaix. However, the population showed considerable pluriannual variations in recruitment, which suggests that recruitment is patchy and/or post settlement processes are highly variable. Olafsson et al. (1994) reviewed the potential effects of pre and post recruitment processes. Recruitment may be limited by predation of the larval stage or inhibition of settlement due to intraspecific density dependent competition. Post settlement processes affecting survivability include predation by epibenthic consumers, physical disturbance of the substratum and density dependent starvation of recent recruits. Hence, venerid bivalve recruitment is probably unpredictable and sporadic.
  • Hydroids are often the first organisms to colonize available space in settlement experiments (Gili & Hughes, 1995). Hydroids that lack a medusa stage, release planula larvae which swim or crawl for short periods (e.g. <24hrs) so that dispersal away from the parent colony is probably very limited (Gili & Hughes, 1995). However, sea beard Nemertesia antennina releases planulae on mucus threads, that increase potential dispersal to 5 -50 m, depending on currents and turbulence (Hughes, 1977). Few species of hydroids have specific substrata requirements and many are generalists capable of growing on a variety of substrata. Hydroids are also capable of asexual reproduction and many species produce dormant, resting stages that are very resistant of environmental perturbation (Gili & Hughes, 1995). However, it has been suggested that rafting on floating debris (or hitch hiking on ships hulls or in ship ballast water) as dormant stages or reproductive adults, together with their potentially long life span, may have allowed hydroids to disperse over a wide area in the long term and explain the near cosmopolitan distributions of many hydroid species (Gili & Hughes, 1995).
  • Sponges may proliferate both asexually and sexually. A sponge can regenerate from a broken fragment, produce buds either internally or externally or release clusters of cells known as gemmules which develop into a new sponge, depending on species. Most sponges are hermaphroditic but cross-fertilization normally occurs. The process may be oviparous, where there is a mass spawning of gametes through the osculum which enter a neighbouring individual in the inhalant current. Fertilized eggs are discharged into the sea where they develop into a planula larva. However, in the majority development is viviparous, whereby the larva develops within the sponge and is then released. Larvae have a short planktonic life of a few hours to a few weeks, so that dispersal is probably limited and asexual reproduction probably results in clusters of individuals.
  • Echinoderms are highly fecund; producing long lived planktonic larvae with high dispersal potential. However, recruitment in echinoderms is poorly understood, often sporadic and variable between locations and dependent on environmental conditions such as temperature, water quality and food availability. For example, the heart urchin Echinocardium cordatum recruitment has been recorded as sporadic, only occurring in 3 years out of a 10 year period (Buchanan, 1967).
  • The mating system of amphipods is polygynous and several broods of offspring are produced, each potentially fertilized by a different male. There is no larval stage and embryos are brooded in a marsupium, beneath the thorax. Embryos are released as sub-juveniles with incompletely developed eighth thoracopods and certain differences in body proportions and pigmentation. Dispersal is limited to local movements of these sub-juveniles and migration of the adults and hence recruitment is limited by the presence of local, unperturbed source populations (Poggiale & Dauvin, 2001). Dispersal of sub-juveniles may be enhanced by the brooding females leaving their tubes and swimming to un-colonized areas of substratum before the eggs hatch (Mills, 1967). However, amphipods are generally mobile and liable to recruit from the surrounding substratum.
  • The tube building polychaetes, e.g. Pygospio elegans and the sand mason worm Lanice conchilega, generally disperse via a pelagic larval stage (Fish & Fish, 1996) and therefore recruitment may occur from distant populations, aided by bed load transport of juveniles (Boström & Bonsdorff, 2000). However, dispersal of some of the infaunal deposit feeders, such as Scoloplos armiger, occurs through burrowing of the benthic larvae and adults (Beukema & de Vlas, 1979, Fish & Fish, 1996). Recruitment must therefore occur from local populations or by longer distance dispersal during periods of bed-load transport. Recruitment is therefore likely to be predictable if local populations exist but patchy and sporadic otherwise.
  • Mya arenaria demonstrates high fecundity, increasing with female size, with long life and hence high reproductive potential. The high potential population increase is offset by high larval and juvenile mortality. Juvenile mortality reduces rapidly with age (Strasser, 1999). Strasser et al. (1999) noted that population densities in the Wadden Sea were patchy and dominated by particular year classes. Therefore, although large numbers of spat may settle annually, successful recruitment and hence recovery may take longer than a year. Recruitment of shallow burrowing infaunal species can depend on adult movement by bed-load sediment transport and not just spat settlement. Emerson & Grant (1991) investigated recruitment in Mya arenaria and found that bed-load transport was positively correlated with clam transport. They concluded that clam transport at a high energy site accounted for large changes in clam density. Furthermore, clam transport was not restricted to storm events and the significance is not restricted to Mya arenaria recruitment. Many infauna, e.g. polychaetes, gastropods, nematodes and other bivalves, will be susceptible to movement of their substratum.
  • Ascidians such as Ascidiella scabra have external fertilization but short lived larvae (swimming for only a few hours), so that dispersal is probably limited. Ascidiella scabra has a high fecundity and settles readily, probably for an extended period from spring to autumn. Svane (1988) describes it as ‘an annual ascidian’ and demonstrated recruitment onto artificial and scraped natural substrata. Eggs and larvae are free-living for only a few hours and so recolonization would have to be from existing individuals no more than a few km away. It is also likely that Ascidiella scabra larvae are attracted by existing populations and settle near to adults (Svane et al., 1987). Fast growth means that a dense cover could be established within about 2 months. Where neighbouring populations are present recruitment may be rapid but recruitment from distant populations may take a long time.
Mobile epifaunal species, such as echinoderms, crustaceans, and amphipods are fairly vagile and capable of colonizing the community by migration from the surrounding areas, probably attracted by the refugia and niches supplied by the maerl. In addition, most echinoderms and crustaceans have long-lived planktonic larvae with high dispersal potential, although, recruitment may be sporadic, especially in echinoderms. Many of the burrowing bivalves have planktonic larvae, potential wide dispersal but high larval mortality resulting in sporadic recruitment.

Time for community to reach maturity

No information was found concerning the time taken for the ‘dead’ maerl communities to reach maturity, and where recovery has been examined, no distinction between ‘live’ and ‘dead’ was made. However, several studies examined undredged, fallow and dredged sites.

De Grave & Whitaker (1999) compared a dredged (extracted) maerl bed with one that been left ‘fallow’ for six months in Bantry Bay, Ireland. They noted that the dredged bed had significantly fewer molluscs than the fallow bed, but significantly more crustaceans and oligochaetes. Hall-Spencer & Moore (2000a, 2000b) examined the recovery of a maerl community after scallop dredging in previously un-dredged and dredged sites in Scotland. In comparison with control plots, mobile epibenthos returned within one month; fleshy macroalgae within six months; the abundance of the tube anemone Cerianthus lloydii was not significantly different after 14 months; other epifauna (e.g. Lanice conchilega and Ascidiella aspersa) returned after 1-2 years; but some of the larger sessile surface species (e.g. sponges, the plumose anemone Metridium senile, the horse mussel Modiolus modiolus and flame shell Limaria hians) exhibited lower abundances on dredged plots after four years. Deep burrowing species (mud shrimp, large bivalves e.g. Mya truncata and the gravel sea cucumber Neopentadactyla mixta) were not impacted and their abundance changed little over the four year period. Hall-Spencer et al. (2003) note that long lived (>10 years) species (e.g. the rayed artemis Dosinia exoleta) can occur at high abundances in maerl beds but that the sustainability of stocks are unknown at present. Hall-Spencer (2000a) noted that there was no significant difference between controls and experimentally dredged sites after 1-2 years at the sites previously subject to dredging.

Overall, it appears that most of the maerl related community could develop within five years, although long-lived and/or large sessile species (e.g. bivalves, anemones, and sponges) would take longer.

Additional information

None

This review can be cited as follows:

Tyler-Walters, H. 2013. Beds of dead maerl. Marine Life Information Network: Biology and Sensitivity Key Information Sub-programme [on-line]. Plymouth: Marine Biological Association of the United Kingdom. [cited 19/12/2014]. Available from: <http://www.marlin.ac.uk/habitatecology.php?habitatid=999&code=2004>