S39.5: Predation pressure by birds on mussels

Forschungszentrum Terramare, Schleusenstr. 1, D-26382 Wilhelmshaven, Germany, fax 04421 944 199

The only mussel species found both subtidally and intertidally in northern temperate climates is the Blue Mussel Mytilus edulis which accounts for a high proportion of the benthic biomass. Mussels grow on soft sediments if they can find attachments, or on hard substrates, where they live submerged or regularly inundated by sea water (Gosling 1992). Their chief avian predators are Eiders Somateria mollissima which either dive or dabble for them, and Oystercatchers Haematopus ostralegus and Herring Gulls Larus argentatus, which feed on them while they are exposed at low water.

Predation pressure on the Blue Mussel, hereafter referred to as the mussel in this review, can be measured as the loss to predators of mussels and of mussel biomass per unit area in a defined time interval. Attempts to measure these quantities have been directed by questions such as (a) how much does predation pressure by birds on mussels contribute to the energy flow in a coastal or estuarine ecosystem (e.g. Baird & Milne 1981, Asmus et al. 1998, Nehls et al. 1998), (b) do birds deplete the prey population significantly over a winter (Zwarts & Drent 1981), (c) what is the effect of predation by birds on a mussel bed in the course of a whole year (Egerrup & Hoegh Laursen 1992), (d) what is the importance of mussels as a food for shorebirds (Faldborg et al. 1994), (e) how great a reduction of intertidal area, or other environmental changes reducing intertidal mussel populations, would cause a drop in bird numbers in response to a decreased carrying capacity (Goss-Custard et al. 1995a; b; Goss-Custard et al. 1996; Goss-Custard & Willows 1996), (f) if mussel populations decrease and mussel-feeding bird populations increase, what effect does predation by birds have on the development of the mussel populations (Hilgerloh & Herlyn 1996; Hilgerloh 1997; Hilgerloh et al. 1997), and (g) if musssel fisheries increase, will enough remain for birds or should fishery quotas be limited (Laursen 1987)?

Predation pressure by birds on mussels has been studied in three different areas: (a) as a contribution to the energy flow in an ecosystem, (b) as a factor which has an effect on the prey population, and (c) as a factor which is relevant for the carrying capacity of a site for birds if intertidal land is lost or other environmental changes occur which reduce the availability of food. The aim of this review is to summarize progress in research in these three areas.

The objective of biological research is to improve the understanding of nature. Progress can be achieved by discovery of new facts, by new explanations of facts which are already known, by new definitions of terms, by new methods, by new or modified evaluation concepts, or by models which generate predictions which can be disproved. According to Mayr (1982, 1997) most progress is achieved not by the discovery of new facts but by the emergence of new definitions and ideas which have to be tested by new measurements and observations. In the research on predation pressure, most progress is likely with respect to evaluation concepts or predictive models, which require a high level of understanding of the relationship of the incorporated variables.

Predation pressure cannot usually be measured directly in the field; it has to be inferred from different variables which are measured in the field. However many observations and measurements are performed, there will never be enough to document the precise extent of predation pressure. Therefore, models have to be developed which make assumptions, implying simplifications. As a consequence their predictions will be imprecise and the models need to be refined in the light of increasing knowledge. It is this area of research that is likely to show the most progress. The models will be influenced by the background of the scientists who develop them; this may cause controversies (Mayr 1997).

First I shall review different approaches to evaluating predation pressure on mussels and then discuss results relating to the main research areas.

In order to evaluate this, data have been collected in two different ways: by exclosure experiments and by observations of the behaviour of birds foraging on the mussel bed. Exclosure experiments allow direct measurements of the effect of predation by birds, but observations can be used only to infer the predation pressure.

In general three different treatments have been applied simultaneously: the first excludes only birds, the second birds and crabs and the third is the control. By comparison of densities and biomass at the beginning and end of the experiment, the decrease in numbers and biomass can be measured directly. Experimental designs have differed. Egerrup & Hoegh Laursen (1992) used three large cages, each of 2 x 1 m with three replicates for each treatment whereas I used cages of 1 m² with 7 replicates (Hilgerloh unpubl.). The larger cages have the advantage that the development of the mussel population within the cages can be monitored better, but the larger number of smaller cages allow the variability on the mussel bed to be taken into account more effectively. In a different type of exclosure experiment the mortality of mussels was determined by counting empty shells and differentiating between various causes of death (Worral & Widdows 1994). Exclosure experiments have been applied successfully in rocky environments (Paine 1974), but have drawbacks in soft substrata because of sedimentation effects (Virnstein 1978). The other problem is that sometimes a mussel bed is so inhomogenous that it is difficult to generalise.

In this approach, certain parameters of behaviour are used to infer the predation pressure. Two essential parameters are the number and percentage of birds of each species feeding on the mussel bed. Data are collected by scan sampling at regular time intervals (e.g. every 30 minutes during the period of low tide). The average proportion of mussels in the food of each species and the average feeding rates are inferred from observations of single birds, and the average exposure time of mussels during the study period is estimated. Analyses of faeces or regurgitated pellets provide information on the sizes of mussels selected. In the laboratory, the ash free dry weights (AFDW) of different sizes of mussels are determined to allow calculation of the biomass eliminated. The development of the mussel population during the study period is determined at regular time intervals by sampling and evaluation of abundance and biomass (e.g. Zwarts & Drent 1981, Hilgerloh et al. 1997). The model for the evaluation of the predation pressure and its effect on the mussel population integrates the measures of the average numbers of feeding birds, the proportion of the exposure time during which they feed, the absolute time the mussel bed is exposed, the feeding rate, the percentage of mussels in the diet, the sizes of mussels preferred by each bird species and the average biomass per size class. As a result, the number of mussels and the biomass taken per day and unit area was calculated for each relevant mussel-feeding bird species during the study period (Hilgerloh et al. 1997).

From the sampling on the mussel bed, the growth of mussels, production and losses per unit area are known for the sampling intervals. The losses to birds can be compared with the production and the total losses, in order to evaluate the effect of predation by birds on the development of a mussel population (Hilgerloh et al. 1997). The innovation here is that studies on both the production and elimination of mussels are included in a single model. However, the model requires further refinement, e.g. by a finer time-scale.

If the study area includes several mussel beds, a different approach is necessary than if it includes only one. The number of birds of the relevant mussel-feeding species in the study area is taken as a base (e.g. Faldborg et al. 1994; Hilgerloh 1997). Monthly mean values are taken from counts which are performed at fortnightly intervals. The average daily consumption of a bird of each species is estimated according to literature sources (Aschoff & Pohl 1970; Kersten & Piersma 1987). The proportion of mussels in the food of a bird species varies regionally (e.g. Laursen 1987, Nehls 1991; Hilgerloh 1996, in press). Either the maximal possible food consumption is calculated on the assumption that 100% of the diet is Blue Mussels (Faldborg et al 1994) or calculations are made on the basis of local studies of diet to provide estimates of food consumption in the study area (Hilgerloh 1997).

The assessment of the biomass and abundance of the mussel populations is done by mapping the limits of the mussel beds and by sampling on typical mussel beds.

The evaluation of large areas can give only rough estimates of the effect of predation pressure by birds on the mussel beds. However, the approach is adequate to gauge the situation in areas of this size and is more likely to be generally applicable than the results of a study on a single mussel bed with its own special conditions. This approach includes the dynamics of the development of the mussel populations and the assessment of the elimination of mussels by birds in relation to production and total elimination in the course of a year (Hilgerloh 1997). In general, there is a lack of data which document the annual production and elimination.

Mathematical models require a much greater effort for their establishment than simple models. However, they have the advantage that they can take into account the dynamics of natural processes. In the evaluation of the effect of predation pressure by birds on the development of a mussel population they can include for example the process of mussel growth (Goss-Custard & Willows 1996), and can even simulate decisions birds have to make in nature (Goss-Custard et al. 1995a; b; c).

While simple models have the disadvantage that they are deterministic and make many simplifications, the most advanced mathemathical models have the disadvantage that they need many data, which are very hard to obtain in the field. When they are not available, the scientist is obliged to make assumptions. Mathematical models allow more accurate predictions than simple models.

In the models so far developed, competition for food is regarded as the main reason for density-dependent mortality or emigration (Goss-Custard et al. 1995b, Goss-Custard & Willows 1996). One model, which makes predictions of the effects of habitat loss on winter mortality, is based on the interaction between the density-dependent mortality rate in winter caused by competition for food and a density-dependent production rate in summer (Goss-Custard et al. 1995). The model predicts the total population size following a loss of winter habitat leading to intensified competition. The critical value which has to be determined is the threshold density at which winter mortality becomes density-dependent. It is the moment at which the population size starts to be affected as habitat is gradually removed.

In another model, which is also based on competition between birds for food as the main cause for mortality and emigration, responses of both mussel and Oystercatcher populations to environmental changes caused by anthropogenic factors are determined (Goss-Custard & Willows 1996). Simulations predict the effect of the removal of feeding areas on the survival rates of Oystercatchers, especially when the quality of the areas that remain declines.

Attempts have also been made to solve the problem that models set up in one environment may give unreliable predictions when applied to another. Based on the interaction between mussels and Oystercatchers the generalised model is made up of functions that are derived from basic properties of the organisms themselves. It includes a sub-model of the Oystercatcher and another of the mussel, both of which are made to interact. The model predicts responses of populations of both species to combinations of environmental changes that have not yet occurred. This model has not yet reached its final state. So that the predictions relating to new circumstances can be made with increasing confidence, refinements are needed. For example, the foraging decisions made by the birds may have to take more fully into account their body condition, their risk of being taken by a predator or of becoming parasitized, and rely less on the simple principle of maximizing current intake rate (Goss-Custard & Willows 1996). The value of the whole approach depends on how well the behaviours or properties of the animals concerned are understood. In principle, the model can be applied to other predator-prey systems within estuaries, as for example Oystercatchers and Cockles (Cerastoderma edule) or Knots (Calidris canutus) and mussels.

Energy flow is sometimes expressed as the consumption efficiency, the energy intake at trophic level n as a proportion of the net productivity at trophic level n - 1. It is an index of the pressure of predators on their prey community (Krebs 1985). In the Ythan estuary, the consumption efficiency of birds preying on mussels was 71% (Baird & Milne 1981). In other words predation by birds removed 71% of the biomass production.

Recent studies have shown a similar high consumption efficiency of 90% for stable mussel beds in Königshafen (Nehls et al. 1998) and of 81% for a mussel bed in the Hobo Dyp (Egerrup & Hoegh Laursen 1992) and of a similarly high percentage again in the Ythan estuary (Raffaelli et al. 1990). This is quite different from the situation on the tidal flats of Spiekeroog with its decreasing mussel stocks (Herlyn 1996), where the consumption efficiency was only 12 and 29% in two different years (Hilgerloh 1997). On an unstable mussel bed in the tidal flats of Spiekeroog (Hilgerloh et al. 1997), the consumption efficiency was 33% in autumn and winter. In the Baltic Sea a totally different situation exists: in a mussel population which was always submerged hardly any predation occurred (Kautsky 1981).

The overall consumption efficiency of birds on all invertebrates on tidal flats is comparatively low, 9% in the area of the islands of Sylt and Römö (Asmus et al. 1998). Thus, the high consumption efficiency on stable mussel beds is exceptional.

Mussel populations may differ with respect to age structure, growth rate and predator species and numbers feeding on them. As study sites often are selected in response to local problems of nature conservation, the situations investigated may differ with respect to several environmental factors. Thus, it is not possible to compare the results of different studies in order to measure the effect of one variable.

The biomass taken by birds depends on the characteristics of the mussel population, the predatory species, the number of feeding birds and the exposure time. Thus, it can vary considerably. According to exclosure experiments on the Danish tidal flats, Oystercatchers and Eiders eliminated 116 g AFDW m^-2 (Egerrup & Høegh Laursen 1992), while on the Dutch tidal flats Oystercatchers alone eliminated 48 g AFDW m^-2, according to direct observations (Craeymeersch et al. 1986). Eiders consumed 360 g AFDW m^-2 in Königshafen/Schleswig-Holstein according to Nehls et al. (1998). In a regional study in the Danish Wadden Sea 300 g AFDW m^-2 emerged as a potential consumption, assuming that 100% of the food of all three predator species – Eiders, Oystercatchers and Herring Gulls – consisted of mussels and that eiders did not feed on mussels which live submerged or on hard substrates (Faldborg et al. 1994). According to another regional study, in the Wadden Sea of Lower Saxony, the same three bird species consumed a total of 32 and 71 g AFDW m^-2 in two different years (Hilgerloh 1997). On a newly settled mussel bed in the same area Oystercatchers and Herring Gulls consumed 71 g AFDW m^-2 in 5 1/2 months (Hilgerloh et al. 1997). Predation pressure differs according to the available size classes and the predatory bird species, as birds are size selective.

To establish the significance of predation by birds for the development of the mussel populations, it is necessary to know what proportion of the total losses is consumed by birds. On the tidal flats of Lower Saxony, in a situation with a high total elimination, predation by birds accounted for only 7 and 15% of total elimination in two different years (Hilgerloh 1997). A similar relationship was found in the same area on a newly settled mussel bed, which disappeared during the first winter. On this mussel bed, 16% of the total loss of biomass in autumn and winter could be attributed to predation (Hilgerloh et al. 1997). These values are lower than in Denmark but similar to those for the Lynher estuary, SW England. In Denmark, predation accounted for 52% of elimination (Egerrup & Høegh Laursen 1992) or for 64% on stable mussel beds (Faldborg et al. 1994). From exclosure experiments in the Lynher estuary, Worral & Widdows (1984) concluded that birds accounted for 16% of total elimination. The small proportion of total elimination accounted for by birds on the tidal flats of Spiekeroog can be explained by the high total elimination.Thus, in an area of decreasing mussel stocks, as in the study area of Lower Saxony, predation was an elimination factor of minor importance, while on stable mussel beds on Danish tidal flats, predation accounted for a high proportion of biomass loss. However, because of the size selectivity of birds, not all mussel size classes suffer to the same extent from bird predation.

Mathematical models have predicted that when habitat is gradually removed the Continental subpopulations of Oysterctachers will be more severely affected by density-dependent winter mortality than the Atlantic subpopulations. The authors emphasize that field studies on winter habitat loss in migratory bird populations should first test whether bird density has already reached the critical threshold level, i.e. whether some birds are already dying through competition for food (Goss-Custard et al. 1995). The model, including sub-models for both mussels and Oystercatchers, revealed that a reduction of the feeding habitat by 60% would reduce the population of Oystercatchers by 50% (Goss-Custard & Willows 1996). If, along with the gradually decreasing habitat, the quality of mussels decreased at the same time through loss of the best habitats, then the population of Oystercatchers would be halved by the time the habitat was reduced by 40%. If a reduction of 60% took place, only 20% of the population would be left (Goss-Custard & Willows 1996). Further predictions for other environmental changes are in preparation.

Considerable progress has been made in the development of models for the evaluation of bird predation on mussels and of its effect on mussel populations. In future, refinements of the models are to be expected. However, there is a lack of inter-regional comparisons based on a common approach. In order to be able to compare different regional studies, biologists responsible for surveys of mussel beds should be aware that data on production and total elimination in the course of a year are needed for the evaluation of predation effects on the mussel populations. Based on the new approach for the evaluation of the predation pressure on a single mussel bed, comparisons between mussel beds with different properties are needed. The lack of basic data on the feeding behaviour of some mussel-feeding bird species mitigates against the improvement of mathematical models. However, more predictions for different environmental changes ought be possible with the models presently available.

Aschoff, J. & Pohl, H. 1970. Der Ruheumsatz von Vögeln als Funktion der Tageszeit und der Körpergröße. Journal für Ornithologie 111: 38-47.

Asmus, H., Lackschewitz, D., Asmus, R., Scheiffarth, G., Nehls, G. & Herrmann, J.-P. 1998. Transporte im Nahrungsnetz eulitoraler Wattflächen des Sylt-Rømø Wattenmeeres. In: Gätje, C & Reise, K. (eds): Ökosystem Wattenmeer – Austausch-, Transport- und Stoffumwandlungsprozesse. Berlin; Springer: 393-420.

Baird, D. & Milne, H. 1981. Energy flow in the Ythan estuary Aberdeenshire, Scotland. Estuar Coast Shelf Sci 13: 455-472.

Craeymeersch, J.A., Herman, P.M.J., Meire, P.M. 1986. Secondary production of an intertidal mussel (Mytilus edulis L.) population in the Eastern Scheldt (SW-Netherlands). Hydrobiologia 133: 107-115.

Egerrup, M. & Høegh Laursen, M.L. 1992. Aspects of predation on intertidal Blue Mussels (Mytilus edulis L.) in the Danish Wadden Sea. C.M./ICES K 25: 1-20.

Faldborg, K., Jensen, K.T., Maargaard, L. 1994. Dynamics, growth, secondary production and elimination by waterfowl of an intertidal population of Mytilus edulis L. Ophelia Suppl 6: 187-200.

Goss-Custard, J.D., Clarke, R.T., Briggs, K.B., Ens, B.J., Exo, K.-M., Smit, C., Beintema, A.J., Caldow, R.W.G., Catt, D.C., Clark, N.A., Le V. dit Durell, S.E.A., Harris, M.P., Hulscher, J.B., Meininger, P.L., Picozzi, N., Prýs-Jones, R., Safriel, U.N. & West. A.D. 1995a. Population consequences of winter habitat loss in a migratory shorebird. I. Estimating model parameters. Journal of Applied Ecology 32: 320-336.

Goss-Custard, J.D., Clarke, R.T., Le V. dit Durell, S.E.A., Caldow, R.W.G. & Ens, B.J. 1995b. Population consequences of winter habitat loss in a migratory shorebird. II. Model predictions. Journal of Applied Ecology 32: 337-351.

Goss-Custard, J.D., Caldow, R.W.G., Clarke, R.T. & West, A.D. 1995c. Deriving population parameters from individual variations in foraging behaviour. II. Model tests and population parameters. Journal of Animal Ecology 64: 277-289.

Goss-Custard, J.D., West A.D., Clarke, R.T., Caldow, R.W.G. & le V. dit Durell, E.A. 1996. The carrying capacity of coastal habitats for oystercatchers. In: Goss-Custard, J.D. (ed) The Oystercatcher. Oxford: Oxford University Press: 327-351.

Goss-Custard, J.D. & Willows, R.I. 1996. Modelling the responses of mussel Mytilus edulis and oystercatcher Haematopus ostralegus populations to environmental change. In: Greenstreet, S.P.R. & Tasker, M.L. (eds) Aquatic predators and their prey. Oxford; Blackwells Scientific Publications: 73-85.

Gosling, E. (ed) 1992. The mussel Mytilus: ecology, physiology, genetics and culture. Developments in Aquaculture and Fisheries Science 25. Amsterdam; Elsevier:589pp.

Herlyn, M. 1996. Zur Bestandssituation der Miesmuschelbänke des niedersächsischen Wattenmeeres. Mitteilungen der Norddeutschen Naturschutzakademie 1, Schneverdingen: 56-61.

Hilgerloh, G. 1996. Miesmuscheln (Mytilus edulis) in der Nahrung der Eiderenten (Somateria mollissima) auf Spiekeroog und Langeoog. Acta Ornithoecologica, Jena/Germany 3.3: 239-243.

Hilgerloh, G. 1997. Predation by birds on blue mussel Mytilus edulis beds of the tidal flats of Spiekeroog (southern North Sea). Mar. Ecol. Prog. Ser. 146: 61-72.

Hilgerloh, G. in press. Die Nahrung der Eiderenten im Nationalpark Niedersächsisches Wattenmeer. In: Umweltbundesamt & Nationalparkverwaltung Niedersächsisches Wattenmeer (eds) Umweltatlas Vol. II (Ostfriesisches Wattenmeer und Hohe Weg Watt, Neuwerker und Wurster Wattenmeer). Ulmer Verlag, Stuttgart.

Hilgerloh, G., & Herlyn, M. 1996. Auswirkungen der Prädation durch Vögel auf eine junge Miesmuschelbank (Mytilus edulis). Verh. dtsch zool Ges. Fischer Verlag, Stuttgart 89: 306.

Hilgerloh, G., Herlyn, M. & Michaelis, H. 1997. The influence of predation by herring gulls (Larus argentatus) and oystercatchers (Haematopus ostralegus) feeding on a newly established mussel (Mytilus edulis) bed in autumn and winter. Helgoländer Meeresunters. 51,2: 173-189.

Kautsky, N. 1981. On the trophic role of the blue mussel (Mytilus edulis L.) in a Baltic coastal ecosystem and the fate of the organic matter produced by the mussels. Kieler Meeresforsch Sonderh 5: 454-461.

Kersten, M. & Piersma, T. 1987. High levels of energy expenditure in shorebirds; metabolic adaptations to an energetically expensive way of life. Ardea 75: 175-187.

Krebs, C.J. 1985. Ecological efficiencies. In: Krebs, C.J. (ed) Ecology – The experimental analysis of distribution and abundance. Third edition. New York; Harper Collins: 633-641.

Laursen, K. 1987. Consumption of mussels by Eiders in the Danish Wadden Sea. In: Tougaard, S. & Aasbirk, S. (eds) Proc 5th intern Wadden Sea Symp, The National Forest and Nature Agency and the Museum of Fishery and Shipping, Esbjerg, Denmark: 239-245.

Mayr, E. 1982. Die Entwicklung der biologischen Gedankenwelt. Vielfalt, Evolution und Vererbung. Berlin; Springer-Verlag: 766pp.

Mayr, E. 1997. This is Biology. The science of the living world. Cambridge; The Belknap press of Harvard University Press: 327 pp.

Nehls, G. 1991. Bestand, Jahresrhythmus und Nahrungsökologie der Eiderente, Somateria mollissima, L. 1758, im Schleswig-Holsteinischen Wattenmeer. Corax 14: 146-209.

Nehls, G., Hertzler, I., Ketzenberg, C. & Scheiffarth, G. 1998. Die Nutzung stabiler Miesmuschelbänke durch Vögel. In: Gätje, C & Reise, K. (eds): Ökosystem Wattenmeer – Austausch-, Transport- und Stoffumwandlungsprozesse. Berlin; Springer: 421-435.

Paine, R.T. 1974. Intertidal community structure: experimental studies on the relationship between a dominant competitor and its principal predator. Oecologia 15: 93-120.

Raffaelli, D., Falcy, C. & Galbraith, C. 1990. Eider predation and the dynamics of mussel bed communities. In: Barnes, M. & Gibson, R.N. (eds) Trophic relations in the marine environment. Aberdeen Univ. Press: 157-169.

Worral, C.M. & Widdows, J. 1984. Investigation of factors influencing mortality in Mytilus edulis L. Mar Biol Lett 5:85-97.

Virnstein, R.W. 1978. Predator caging experiments in soft sediments: caution advised. In: Wiley, M.L. (ed) Estuarine interactions. London; Academic Press: 261-273.

Zwarts, L. & Drent, R.H. 1981. Prey depletion and the regulation of predator density: Oystercatchers, Haematopus ostralegus, feeding on mussels Mytilus edulis. In: Jones, P. & Wolff, W.J. (eds) Feeding and survival strategies of estuarine organisms. New York; Plenum Press: 193-216.