The contribution of seagrass beds (Zostera noltii) to the function of tidal flats as a juvenile habitat for dominant, mobile epibenthos in the Wadden Sea
Since the substantial loss of subtidal eelgrass (Zostera marina L.) in the 1930s, seagrass beds in the Wadden Sea are limited to the intertidal zone and dominated by Z. noltii Hornem. This study deals with the effect of vegetated tidal flats on quantities of mobile epifauna and proves empirically the function of seagrass canopies as a refuge for marine animals remaining in the intertidal zone at ebb tide. Drop-trap samples were taken in the Sylt-Rømø Bight, a shallow tidal basin in the northern Wadden Sea, on vegetated and unvegetated tidal flats during July and August 2002, and during the entire growth period of Z. noltii from May to September in 2003. The species composition in Z. noltii and bare sand flats showed minor differences since only two isopod species (Idotea baltica and I. chelipes) occurred on Z. noltii flats exclusively. Juvenile shore crabs (Carcinus maenas L.), brown shrimps (Crangon crangon L.) and common gobies (Pomatoschistus microps Krøyer) were also found abundantly on bare sand flats. However, the results showed significantly higher abundances and production of these dominant species on vegetated tidal flats. Additionally, the analyses of faunal size classes indicated higher percentages of small individuals in the seagrass bed during the entire sampling period. Despite drastic diurnal fluctuations of dissolved oxygen at low tide, faunal density in the residual water layer remaining in seagrass canopies at ebb tide was found to be consistently higher than that found in artificially created tide-pool units. Although species composition of mobile epifauna did not basically differ between vegetated and unvegetated tidal flats, Z. noltii beds are considered to contribute quantitatively to the function of tidal flats, as an extended juvenile habitat for some of the most important species of the Wadden Sea food web.
The shallow waters of the Wadden Sea are generally considered as important juvenile habitats contributing to the development of species stocks and faunal composition of the North Sea (Zijlstra 1972; Beukema 1992). For fishes and epibenthic macrofauna, this is widely attributed to habitat diversity by topographical properties, such as extended shallow-water areas, tidal flats and gulleys, and to salinity gradients predominating in river estuaries (Zijlstra 1978). The role of biogenic habitats is barely considered, although they represent the majority of the few complex bottom structures within an ecosystem dominated by soft sediments. While seagrass habitats received major attention in the neighbouring Baltic Sea region (Boström and Bonsdorff 1997, 2000) and at the Swedish Skagerrak (Pihl and Rosenberg 1982; Pihl Baden and Pihl 1984), it received less attention in the Wadden Sea. This might be due to major seagrass loss during the 1930s when subtidal Zostera marina beds became extinct in the Wadden Sea area, and these have failed to recover until today. A further seagrass die-off began in the early 1970s, when intertidal Z. noltii beds declined drastically in the central (German) and in the southern (Netherlands) part of the Wadden Sea (Den Hartog and Polderman 1975; Reise 1994). Only little attention has been paid to the ecological functions of the remaining intertidal Z. noltii beds in the northern Wadden Sea (Germany/Denmark). Whereas subtidal seagrass beds are generally known to act as important refuges for several fishes and invertebrates, increasing faunal diversity and species richness (e.g. Orth et al. 1984; Heck et al. 1995), the faunal composition harboured by Z. noltii beds in the intertidal zone of the Wadden Sea is not thought to differ from that found on unvegetated tidal flats (Den Hartog 1983). However, previous studies showed that the presence of intertidal seagrasses promotes meiofaunal abundance (Hellwig-Armonies 1988), and potentially increases food availability for endo-and epibenthic organisms by acting as a sink for organic matter (Asmus and Asmus 2000). Due to grazing water fowl (Ganter 2000; Nacken and Reise 2000) and harsh weather conditions during winter, the occurrence of dense Z. noltii beds in the Wadden Sea is limited to the summer months (Schanz and Asmus 2003). Thus, habitat functions of vegetated tidal flats are temporary during the course of the year. From June to October, “young of the year” juvenile stages (0-group) of brown shrimps (Crangon crangon), common gobies (Pomatoschistus microps) and shore crabs (Carcinus maenas) use tidal flats intensively as refuges, avoiding predation pressure in the subtidal zone (Klein Breteler 1976; Kuipers and Dapper 1984; Del Norte-Campos and Temming 1998). High densities of juvenile brown shrimps can be found additionally in small tide pools and puddles (Berghahn 1983; Hinz 1989), outlasting the period of tidal emersion. Resources of residual water might allow organisms to avoid tidal migration and to remain in the intertidal zone during the ebb tide.
Dense Z. noltii beds may provide additional ebb-tide refuges by retaining water on the tidal flats, trapped between the leaves. However, oxygen contents within the canopy might decrease at night due to plant respiration and might induce unsuitable conditions for epifaunal organisms.
The aim of this study is to demonstrate the impact of vegetated tidal flats on individual densities of dominant epibenthic species that are considered as most important within the food webs of the Wadden Sea.
This investigation is based on the following hypotheses: (1) the presence of Z. noltii beds in the intertidal area promotes density and biomass of dominant, 0-group epibenthos during the vegetation period. (2) Z. noltii beds provide extended ebb-tide refuges for juvenile mobile epibenthos by retaining water on the tidal flats, trapped between the leaves.
Materials and methods
Vegetated and unvegetated sampling sites were located within the same tidal level on the intertidal area on the east coast of Sylt (Fig. 1).
Shoot density of Z. noltii was determined monthly from May to September 2003 in six replicate test areas. A frame of 25×25 cm was randomly placed on the seagrass canopy. Shoots were counted on sub-units (156 cm2) and seagrass density was extrapolated to 1 m2.
Sampling of mobile epibenthos
For quantitative sampling of mobile epibenthos on inundated tidal flats, a portable drop-trap (modified after Pihl Baden and Pihl 1984) was used in the daytime during the main vegetation period of Z. noltii from July to September 2002, and during the whole vegetation period from May to September 2003. Six random replicates were taken monthly in Z. noltii beds and adjacent bare sand flats on following days during slack water (1 h before high tide to 1 h afterwards). The drop-trap consisted of a 1.20-m-high aluminium frame with a sampling area of 0.5 m2 and was mounted at the bow of a small boat powered by an outboard engine. Before sampling, the engine was turned off and after 2 min of drifting, the trap was dropped. It was emptied for epifauna using a hand-net (1 mm mesh size) attached to a 1.5-m-long handle. After six hauls through the sediment surface and water column, the trap was considered as empty, since previous studies showed no relevant increase of individual numbers with further hauls (Polte et al. 2005). Samples were taken to the laboratory and preserved by freezing in seawater. Organisms were determined to species level, counted and divided into size classes of 5- to 10-mm steps respectively, according to total body length (fishes and shrimps) or carapace width (crabs). From each sample, the 0-group individuals of dominant species were dried separately at 80°C and ashed at about 500°C for 14 h to estimate biomass in ash-free dry weight (AFDW).
Production intervals from July to August and from August to September were pooled to demonstrate habitat specific differences of faunal production during the main seagrass vegetation period.
Artificial tide pools
Simultaneously with faunal sampling, physical parameters were measured within artificial tide pools and the canopy-water layer at low tide during the day and night. Water samples were taken to the laboratory and analysed for salinity (AUTOSAL salinometer model 8400A Guildline Intr.). Oxygen content and temperature were measured in situ by a Microprocessor Oxymeter (OXI 196, WTW). Salinity samples, temperature and oxygen measurements were taken for each of the six tide-pool replicates and six randomly chosen areas within the canopy-water layer.
Abundances of dominant species were calculated to 1 m2 and presented as arithmetic mean±standard error (SE). Total abundances were calculated as the sum of species abundances within each replicate. For annual comparison of faunal distribution, high-tide abundances of July and August were pooled. Results were statistically analysed by means of 1-way analysis of variance (ANOVA), followed by a Tukey’s honest significance (HSD) multiple comparison test (tide-pool experiment). Previously, data were tested for homoscedasticity of variances using Cochran’s test and transformed by (log+1) if necessary to fulfil the requirements of ANOVA. Statistical significance was assumed if the P-value was <0.05. Statistical analyses were conducted using STATISTICA (StatSoft, Tulsa, Okla.).
The shoot density of Z. noltii was 1,088 m−2 in May and increased up to 12,000 shoots m−2 during August. Within the semi-diurnal tidal cycle, investigated intertidal areas were exposed for 4–6 h. While unvegetated tidal flats were completely exposed to air except for several sinks and soft-bottom tide pools, Z. noltii beds retained water between the leaves, forming extended water resources during low-tide periods. The mean depth of these water layers was 6.2 cm (±0.12, n=6). Tidal phase and predominant wind directions generally influenced the extension of the water layer, but are not considered to affect the experiment, since experimental units were sampled during the same day.
Species composition of mobile epifauna
The mobile, epibenthic fauna on investigated tidal flats was dominated by shore crabs (Carcinus maenas), brown shrimps (Crangon crangon), gobiid fish (P. microps) and isopods (Idotea spp.). These species represented the entire composition of mobile epifauna, except for gastropods (Hydrobia spp., Littorina spp.), which were not involved in this study. Isopods were the only group restricted to vegetated areas, whereas all decapod crustaceans and the gobiid fishes were common in Z. noltii beds, as well as on bare-sand flats. Isopods were composed of the species Idotea baltica (96%) and I. chelipes (4%). The mean abundance of both Idotea spp. found in the seagrass habitat was 64.3 (±6.4) individuals m−2 in July 2003. The following results focus on the quantitative distribution of Carcinus maenas, Crangon crangon and P. microps generally found on both vegetated and unvegetated tidal flats.
High-tide abundance in different intertidal habitats
One-way ANOVA analysis of mean (±SE) abundance of dominant epifaunal species in seagrass and bare sand habitats during Zostera noltii growth period 2003. Significance level (P) set at 0.05; degree of freedom (df) is 1 for each month
Most recruits of Crangon crangon captured on tidal flats in May and June were less than 10 mm in body length. The percentages of individuals belonging to the <15-mm cohort increased in both habitats from July to September, whereas the occurrence of the <20-mm cohort was more pronounced on bare-sand flats.
Recruitment of P. microps took place in June, when the percentage of the <10-mm cohort was higher on tidal flats without vegetation while individuals caught in the Z. noltii bed were larger. During July and August, the majority of individuals were grouped into the <20-mm cohort in both habitats. At the end of the investigation period in September, most of the gobiid fish caught in vegetated and unvegetated tidal flats were in the cohort <30 mm. Smaller individuals were limited to the seagrass bed.
Significance level (P) of one-way ANOVA between epifaunal production on vegetated and unvegetated tidal flats. Pooled intervals from July to September 2003 showed a significantly higher production of Carcinus maenas, Crangon crangon, Pomatoschistus microps and species as a total within the seagrass habitat
Total epifaunal abundance within Zostera noltii beds during flood and ebb tide
A total of Crangon crangon, Carcinus maenas and P. microps showed that a mean of 67 (±3.0) individuals m−2 remained in the seagrass bed at low tide. At high tide, total abundance showed a mean of 167 (±7.2) individuals m−2. As a result, the high-tide abundance exceeded the low-tide abundance significantly by a factor of about 2.5 (ANOVA, P<0.001, df=1, F=48.41).
Individuals of mobile epibenthos were consistently more frequent in the CWL than in artificial tide pools located in both habitats (Fig. 7). ANOVA of abundances m−2 in experimental tide-pool units and the CWL showed significant effects during the low-tide period in the day (P<0.01, df=2, F=10) and night time (P<0.001, df=2, F=28). While the utilization of tide pools by mobile epifauna in vegetated and unvegetated habitats was influenced by diurnal variations, abundances were higher and more stable in the CWL (Tukey’s test, P<0.05). During the day, abundances did not differ between sand pools and seagrass pools (Tukey’s test, P=0.4), whereas at night faunal density was significantly higher in seagrass pools (Tukey’s test, P<0.01).
Sampling of mobile epifauna was limited to daytime. A number of studies dealing with distribution of mobile epifauna in different habitats describe a diurnal variability of abundances and often an increased faunal activity during the night (Summerson and Peterson 1984; Sogard and Able 1994; Mattila et al. 1999).
Previous investigations at the particular sampling sites showed no significant diurnal variations of abundances (Polte et al. 2005) at a certain tidal level. Thus limitation of sampling to daytime flood is not considered to influence general distribution patterns of dominant species in different intertidal habitats.
Worldwide subtidal seagrass beds are considered to represent important nursery areas for fishes and invertebrates. These areas provide shelter from predation by increasing habitat complexity (e.g. Virnstein et al. 1983; Orth et al. 1984; Hindell et al. 2000). The Wadden Sea area as a whole is generally thought to act as a nursery area for crustaceans and fishes (e.g. Zijlstra 1972; Van der Veer et al. 2001). However, the reliable identification of nursery areas requires the investigation of certain criteria (Beck et al. 2001). Several studies classify nursery areas according to their contribution to the production of a species population (review in Heck et al. 2003). In numerous coastal waters, seagrass beds show a major contribution to production (Fredette et al. 1990; Valentine and Heck 1993). In the Sylt-Rømø Bight, the production of benthic macrofauna harboured by habitats such as seagrass meadows and mussel beds is considered to be low compared to the productivity of the entire system, according to the small percentages they cover of the entire tidal basin area (Asmus and Asmus 2000). Since the intertidal zone of the Wadden Sea might be classified as an extended juvenile habitat as a working-term, the role of Z. noltii beds is difficult to isolate, especially when the faunal composition does not basically differ from that of unvegetated tidal flats, and the number of juveniles contributing to species’ adult populations is not attributed to single habitats but to an entire coastal system. Shallow-water zones generally represent important refuges from fish predation (Ruiz et al. 1993; Paterson and Whitfield 2000; Linehan et al. 2001), and promote faster growth due to higher temperatures (Gibson et al. 2002). Thus the habitat function of intertidal seagrass beds might be superimposed by the distribution of juveniles along vertical depth gradients in shallow coastal waters (Jenkins et al. 1997).
However, during their first benthic stages, the production of all species was significantly higher within the seagrass bed compared to unvegetated sand flats. These results indicate that intertidal seagrass beds may quantitatively support the function of the intertidal zone as juvenile habitat. Additionally, drop-trap sampling of proximate seagrass and sand flats indicates that abundances of epifauna are increased because the residence of juveniles in the intertidal zone is seasonally matched with the growth period of Z. noltii. Despite high temperatures recorded in the intertidal zone in 2003, clear inter-annual differences in individual numbers in July/August 2002 and 2003 were limited to gobiid fishes and did not affect distribution patterns of individuals on vegetated and unvegetated tidal flats. From May to September 2003, a seasonal effect on epifaunal distribution patterns was pronounced, which is suggested to be linked to settlement strategies of species on tidal flats. In June, before the seagrass beds reach their maximum shoot density (Schanz and Asmus 2003), abundances of Crangon crangon and P. microps did not differ compared to unvegetated tidal flats. However, Carcinus maenas reached the highest density in this particular period. Since Crangon crangon and P. microps enter the tidal flats as benthic post-larvae from the subtidal zone, colonization of different intertidal habitats occurs secondarily as a post-settlement process. In contrast, settlement processes of Carcinus maenas are inherently affected by seagrasses that provide suitable settling structures for megalopae stages (Moksnes 2002).
The distribution patterns of mobile epifauna on vegetated and unvegetated tidal flats were characterized by the dominance of individuals smaller than 5 mm during the whole vegetation period until September. The percentage of larger specimens was higher on unvegetated sediments. In general, this size-class distribution was maintained by all investigated species. As they all are important predators significantly influencing benthic communities on tidal flats (Beukema 1992; Baird et al. 2004), the shift of habitat type is supposed to be due to predation success of the foraged juveniles. Vertical size-class distributions, as well as non-significant Crangon crangon abundances between vegetated and unvegetated areas in September, indicate that utilization of bare-sand habitats applies at later juvenile stages.
Role of Zostera noltii beds as ebb-tide refuge
Although the total abundance of mobile epibenthos within the seagrass bed was significantly higher at high tide, a considerable number of individuals were found in the layer of residual water retained by the seagrass canopy (CWL) during ebb tide. Since migration to subtidal areas with the fading tide might expose juvenile epifauna to predation (Kneib 1987; Gibson et al. 2002; Amara and Paul 2003), the CWL is an important refuge that allows a permanent residence in the intertidal zone. However, Van der Veer and Bergman (1986) showed that physical stress caused by oxygen deficiency within residual water resources in the intertidal zone leads to an increased tidal migration of 0-group plaice towards deeper tidal gullies at night. Accordingly, the utilization of intertidal habitats might be influenced by diurnal fluctuations of the oxygen climate. Surprisingly, the results showed consistently higher epifaunal abundances in the CWL compared to experimentally created tide pools, although the oxygen level dropped drastically in the CWL at night. In fact, at night, epifaunal abundance within tide pools in different habitats and the CWL was reciprocally proportional to the oxygen content in the particular environment (Fig. 7). At low tide during the day, the oxygen content within pools on bare-sand flats and pools inside the seagrass bed showed no differences. However, at night, the oxygen content in sand pools did not drop as drastically as it did in seagrass pools or the CWL although the water temperature was similar. The decreasing oxygen content is attributed to plant respiration (e.g. Pedersen et al. 1998; Greve et al. 2003) rather than respiration by associated fauna because in the daytime no correlation between faunal abundance and oxygen content could be found at the particular investigation units. Since sampling success in tide pools might be promoted by accumulation of animals on a limited submerged area, the high faunal density found within the spacious CWL is even more noticeable. As a result, diurnal fluctuations of oxygen contents are not considered to influence the function of intertidal seagrass canopies as ebb-tide refuges for mobile epibenthos.
Species composition of dominant, mobile epibenthos is not necessarily related to seagrass cover in the intertidal zone of the Wadden Sea. Only Idotea spp. were closely associated with Z. noltii habitats. However, the presence of Z. noltii beds significantly increases abundances and production of juvenile epifauna on the tidal flats during the main growth period of the plants. Additionally, the residual water layer within dense Z. noltii beds provides extended ebb-tide refuges for dominant epifaunal species, despite hypoxic conditions occurring in emerged seagrass beds at night. As a result, intertidal seagrass beds contribute to the function of the intertidal zone as a juvenile habitat for mobile epifauna in the northern Wadden Sea.
Sincere thanks are given to A. Hintz, P. Götz, T. Zeilinger and B. Kürten for their enthusiastic assistance in the field. We gratefully acknowledge Dr. R. Asmus for all the support during this study, and J.Godbold for her help in correcting the English. Thanks are extended to two anonymous referees for valuable comments on the manuscript.
- Baird D, Asmus H, Asmus R (2004) Energy flow of a boreal intertidal ecosystem, the Sylt-Rømø Bight. Mar Ecol Prog Ser 279:45–61Google Scholar
- Beck MW, Heck KL Jr, Able KW, Childers DL, Eggleston DB, Gillanders BM, Halpern B, Hays CG, Hoshino K, Minello TJ, Orth RJ, Sheridan PF, Weinstein MP (2001) The identification, conservation, and management of estuarine and marine nurseries for fish and invertebrates. BioScience 51:633–641Google Scholar
- Berghahn R (1983) Studies of flatfish and brown shrimp (Crangon crangon) from tidal flats of the Wadden Sea following their transition to a bottom-dwelling mode of life. Helgol Wiss Meeresunters 36:163–181Google Scholar
- Beukema JJ (1992) Dynamics of juvenile shrimp Crangon crangon in a tidal-flat nursery of the Wadden Sea after mild and cold winters. Mar Ecol Prog Ser 83:157–165Google Scholar
- Boström C, Bonsdorff E (2000) Zoobenthic community establishment and habitat complexity—the importance of seagrass shoot-density, morphology and physical disturbance for faunal recruitment. Mar Ecol Prog Ser 205:123–138Google Scholar
- Crisp DJ (1984) Energy flow measurements. In: Holme NA, McIntyre AD (eds) Methods for the study of marine benthosGoogle Scholar
- Den Hartog C (1983) Structural uniformity and diversity in Zostera-dominated communities in Western Europe. Mar Technol Soc J 17:6–14Google Scholar
- Fredette TJR, Diaz RJ, Montfrans J van, Orth RJ (1990) Secondary production within a seagrass bed (Zostera marina and Ruppia maritima) in lower Cheasapeake Bay. Estuaries 13:431–440Google Scholar
- Gibson RN, Robb L, Wennhage H, Burrows MT (2002) Ontogenetic changes in depth distribution of juvenile flatfishes in relation to predation risk and temperature on a shallow-water nursery ground. Mar Ecol Prog Ser 229:233–244Google Scholar
- Greve TM, Borum J, Pedersen O (2003) Meristematic oxygen variability in eelgrass (Zostera marina). Limnol Oceanogr 48:210–216Google Scholar
- Heck KL Jr, Abele KW, Roman CT, Fahay MP (1995) Composition, abundance, biomass and production of macrofauna in a New England estuary: comparisons among eelgrass meadows and other nursery habitats. Estuaries 18:379–389Google Scholar
- Heck KL Jr, Hays G, Orth RJ (2003) Critical evaluation of the nursery role hypothesis for seagrass meadows. Mar Ecol Prog Ser 253:123–136Google Scholar
- Hellwig-Armonies M (1988) Mobile epifauna on Zostera marina, and infauna of its inflorescences. Helgol Wiss Meeresunters 42:329–337Google Scholar
- Hinz V (1989) Monitoring the fish fauna in the Wadden Sea with special reference to different fishing methods and effects of wind and light on catches. Helgol Wiss Meeresunters 43:447–459Google Scholar
- Jenkins GP, May HMA, Wheatley MJ, Holloway MG, (1997) Comparison of fish assemblages associated with seagrass and adjacent unvegetated habitats of Port Phillip Bay and Corner Inlet, Victoria, Australia, with emphasis on commercial species. Estuarine Coastal Shelf Sci 44:569–588CrossRefGoogle Scholar
- Kneib RT (1987) Predation risk and use of intertidal habitats by young fishes and shrimp. Ecology 68:379–386Google Scholar
- Kuipers BR, Dapper R (1984) Nursery function of Wadden Sea tidal flats for the brown shrimp Crangon crangon . Mar Ecol Prog Ser 17:171–181Google Scholar
- Orth RJ, Heck KL Jr, Montfrans J van (1984) Faunal communities in seagrass beds: a review of the influence of plant structure prey characteristics on predator-prey relationships. Estuaries 7:339–350Google Scholar
- Pedersen O, Borum J, Duarte CM, Fortes MD (1998) Oxygen dynamics in the rhizosphere of Cymodocea rotundata. Mar Ecol Prog Ser 169:283–288Google Scholar
- Pihl Baden S, Pihl L (1984) Abundance, biomass and production of mobile epibenthic fauna in Zostera marina (L.) meadows, western Sweden. Ophelia 23:65–90Google Scholar
- Reise K (1994) Changing life under the tides of the Wadden Sea during the 20th century. Ophelia 6:117–125Google Scholar
- Ruiz GM, Hines AH, Posey MH (1993) Shallow water as a refuge habitat for fish and crustaceans in non-vegetated estuaries: an example from Chesapeake Bay. Mar Ecol Prog Ser 99:1–16Google Scholar
- Schanz A, Asmus H (2003) Impact of hydrodynamics on development and morphology of intertidal seagrasses in the Wadden Sea. Mar Ecol Prog Ser 261:123–134Google Scholar
- Sogard SM, Able KW (1994) Diel variation in immigration of fishes and decapod crustaceans to artificial seagrass habitat. Estuaries 17:622–630Google Scholar
- Summerson HC, Peterson CH (1984) Role of predation in organizing benthic communities of a temperate-zone seagrass bed. Mar Ecol Prog Ser 15:63–77Google Scholar
- Valentine JF, Heck KL Jr (1993) Mussels in seagrass meadows: their influence on macroinvertebrate abundance and secondary production in the northern Gulf of Mexico. Mar Ecol Prog Ser 96:63–74Google Scholar
- Van der Veer HW, Bergmann MJN (1986) Development of tidally related behaviour of a newly settled 0-group plaice (Pleuronectes platessa) population in the western Wadden Sea. Mar Ecol Prog Ser 31:121–129Google Scholar
- Virnstein RW, Mikkelsen PS, Kalani DC, Capone MA (1983) Seagrass beds versus sand bottoms: the trophic importance of their associated benthic invertebrates. Fla Sci 46:363–381Google Scholar
- Zijlstra JJ (1972) On the importance of the Wadden Sea as a nursery area in relation to the conservation of the southern North Sea fishery resources. Symp Zool Soc Lond 29:233–258Google Scholar
- Zijlstra JJ (1978) The function of the Wadden Sea for the members of its fish-fauna. In: Dankers N, Wolff WJ, Zijlstra JJ (eds) Fishes and fisheries of the Wadden Sea. Final report of the section “Fishes and fisheries” of the Wadden Sea Working Group. Report 5. Stiching Veth tot Steun aan Waddenonderzoek, Leiden, Netherlands, pp 20–25Google Scholar