Building with Nature BwN Guideline Environments Project phases Governance BwN Knowledge base
BwN Building Blocks BwN Toolbox Pilots and cases BwN Knowledge

Log in

In the Netherlands, the potential impacts of sand mining activities on populations of Sandwich Terns often form an important topic in Environmental Impact Assessments (EIAs). Sand mining causes an increase in water turbidity, which may affect populations of visually hunting birds such as terns. 
The quantification of ecological effects in EIAs is mostly done by deterministic modeling of cause-effect chains; a highly conservative approach with an unknown uncertainty margin. A probabilistic approach to the quantification of the possible ecological effects may be an alternative. 
The objective of this case study is to explore how and to what extent a probabilistic approach can be applied to the quantitative modeling of the potential effects of sand mining on tern populations. As an example, the probabilistic methodology is worked out for the effects of a fictitious dredging project on a population of Sandwich Terns.

    General

     

    Title: Adaptive Monitoring Strategies in sand mining / dredging; Knowledge Topic Sandwich Terns – a probabilistic analysis of the ecological effects of dredging.
    Location: Two fictitious dredging projects in the Dutch coastal zone, located within the foraging area of Sandwich terns, at a distance of 7.5 km from their breeding site.
    Date: The large fictitious project takes 8 years to complete and the small fictitious project 2 years.
    Partners: This case study was developed by DHV & IMARES
    Costs: not relevant
    Abstract: A quantitative cause-effect chain on how tern populations may be affected by sand mining activities makes it possible to model the impact of dredging on the tern population. Using a probabilistic approach, the impact of two fictitious different dredging scenarios on the number of breeding terns is modeled.
    Topics: Coastal development, Ecological studies, Ecology – birds, Environmental impacts, Environmental management, Environmental monitoring, Environmental protection, Impact assessment, Marine mining, Project Management, Risk assessment, Risk management, Water quality

    Project Objective

    The objective of this case study (van Kruchten & van der Hammen, 2011) is to explore how a probabilistic approach can be applied for the quantitative modeling of the potential effects of sand mining on tern populations. As an example, the probabilistic methodology is worked out for the effects of a fictitious dredging project on a population of Sandwich Terns (Sterna sandvicensis).

    Project Solution

    To assess the impact of an activity on the environment, investigators often deal with limited knowledge about possible effects. Consequently assumptions have to be made. In general, the foundation of the methods that were used in the past to estimate impacts on tern populations quantitatively is very limited.

    probabilistic approach gives insight into the probability of occurrence of certain impacts. This prevents accumulation of worst case assumptions yielding unrealistic negative impact estimates of an activity. Not only does a probabilistic approach enable more realistic impact estimates, it also quantifies the uncertainty of this impact. Moreover, it shows the influence of natural dynamics on the uncertainty margins. This also contributes to one of the basic steps for generating Building with Nature development & design solutions: understanding how the system functions (see BwN guideline).

    Governance context

    Prior and during marine construction activities like dredging, proponents need to ensure authorities that the dredging activities are performed in an environmentally acceptable manner. For most dredging projects an Environmental Impact Assessment (EIA) is made, often specified and required by law or regulation. An EIA regularly contains an assessment of the potential effects of the project. In the Netherlands, the potential impacts of sand mining activities on tern populations often form an important topic in dredging-related EIAs. Probabilistic analyses as demonstrated below can be used to assess potential effects of dredging activities on the Sandwich tern.

    Probabilistic Approach

    Approach

    This section describes a probabilistic approach to the quantitative modeling of the potential effects of sand mining on tern populations. The method described herein is mainly applicable in the planning and design phases of a dredging project.

    Cause-Effect Chain

    In order to model the impact of dredging activities on Sandwich Tern populations, a literature search was carried out to find out how dredging could affect these populations. This literature search led to the cause-effect chain that is shown in Figure 1. In this chain the relations that are expected to influence the population size are visualized.

    In short, the cause effect chain starts from an increase of fine sediment concentrations caused by dredging, which leads to an effect on the population size through a decrease in breeding success. The main steps involved leading to this reduction are (orange boxes in Figure 1).

    1. sand extraction by dredging activities causes an increase of the silt concentration in the water column,
    2. the increase of the silt concentration causes an increase in turbidity of the water,
    3. the increase in turbidity reduces the catchability of fish by terns,
    4. the reduced catchability leads to an increase in time needed to catch enough food,
    5. if the available time is a limiting factor, the amount of food brought to the chicks is reduced,
    6. if food intake of the chicks is reduced, the breeding success of the terns decreases.
      Van Kruchten & van der Hammen (2011) explain the steps in the cause-effect chain in more detail.

    Modelling the cause-effect chain

    Monte Carlo analysis

    The cause-effect chain for the impact of dredging activities on tern populations contains several uncertain and variable factors. As these uncertainties may have a significant influence on the ultimate impact, a probabilistic model approach is taken.
    First, the relationships between the different elements in the cause-effect chain are made quantitative (see sections below). The resulting chain of equations was used in a Monte Carlo analysis, simulating the impact of dredging on the tern populations a large number of times (e.g. 1000 times), each time with a different selection of the stochastic inputs. This analysis results in a probability distribution of the change in population size due to the dredging activities.
    The following sections describe the relationships between the different elements in the cause-effect chain.

    From sand extraction to water transparency

    Water transparency is mainly influenced by the fine sediment that is released by the dredging activities, the background concentration of suspended particulate matter (SPM) and the concentration of phytoplankton. To model the impact on tern populations, the increase of the SPM-concentration during the feeding period of tern chicks within the foraging area of the tern population is relevant.
    Background SPM-concentrations and phytoplankton concentrations show large fluctuations, caused mainly by changing weather conditions. Because of these fluctuations, the SPM and phytoplankton concentrations are taken as stochastic inputs to the model.

    From total suspended matter to water transparency

    Advanced numerical models can be used for modeling the reduction of the water transparency by sand mining activities and phytoplankton concentrations. In the present demonstration case, however, the following simple empirical relation is used (Suijlen & Duin, 2001), assuming light transparency to be a good measure for visibility:

    with
    d = light attenuation coefficient (m -1 )
    SPM = concentration of suspended particulate matter (g/m 3 )
    Chl-a= concentration of Chlorophyll-a (μg/m 3 )

    Water clarity, or visibility, is usually measured with a Secchi disk. The following empirical relation for the southern North Sea, determined by Visser in 1970, is used to convert the light attenuation coefficient K d to the Secchi disk transparency S (Baptist & Leopold, 2010):

    with
    S = Secchi disk transparency (m)

    From water transparency to hunting efficiency

    Turbidity is expected to reduce the number of prey fishes per unit time that can be caught by terns, which in its turn is a function of the ‘findability’ of the prey fish and the capture efficiency of a hunting Tern. 
    Baptist & Leopold (2010) studied prey capture probability as a function of water transparency for foraging Sandwich Terns. Prey capture probability is defined as the probability that an attempt to catch a fish succeeds. A logistic optimum curve shows the relationship between transparency and prey capture probability:

    In Figure 2, the solid line presents the results of the logistic regression, the dotted line gives the 95% confidence interval. The histograms on the top and bottom respectively give the number of caught prey and missed prey and the dots give the average probability for each histogram class (Baptist & Leopold, 2010).

    'Findability' of prey

    Unfortunately no information is available on the relation between turbidity and 'searching-time'. It seems reasonable to assume, however, that searching-time increases with increasing turbidity. The relative increase of the searching-time is assumed to equal the relative increase of the number of prey capturing attempts.

    Number of prey that can be caught per day

    The relative increase of the number of attempts and the searching time is necessary for estimating the number of fishes that can be fed to the chicks. For estimating this number the following additional information is necessary (also see Van Kruchten & van der Hammen, 2011):

    • The number of fishes that can be caught under optimal transparency conditions. Three options were used: 1) worst case approach: assuming that the amount of fish that can be caught per day at optimal conditions is exactly the number of fishes that is necessary for the adult terns, plus the number that is necessary for feeding the chicks in order to achieve an optimal breeding success; 2) conservative approach: assuming such a capture rate at optimal conditions, that any decrease of the capture rate leads to a lower breeding success; 3) empirical approach: estimating the capture rate at optimal conditions from measurements.
    • The amount of fish adult terns need for food per day.
      This depends the time terns spend flying per day, the digestability of fish, the energy content of fish, the average fish mass and the Basal Metabolic Rate (BMR). Based on available literature it was calculated that each adult has to catch 43 fishes per day (see van Kruchten & van der Hammen, 2011). This number is based on: an average mass of 4 g per fish; 33 fishes per day for adult consumption; and 10 fishes per day per adult for the chicks (of which 30% gets lost, so 14 fishes are fed to the chicks per day (7 per adult)).
    • How much time per day is available for foraging. For a population that breeds at a distance of, for example, 7.5 km from the foraging area, it would mean that less than 8 hours can be spent to foraging.

    From amount of food brought to the nest to breeding success

    Although food conditions are often suggested as a limiting factor for breeding success, a clear relation between breeding success and the amount of food provided has not yet been found. Stienen & Brenninkmeijer (1998) measured the breeding success and the amount of food provided. This data was used to derive the relation between provided number of fishes and breeding success (Figure 3). Further research is needed on the breeding success if during short periods of time hardly any food is provided to the chicks.

    Population dynamical model

    A decrease of the breeding success of terns is expected to result in a decrease of the population size. A method for modeling this change of the population size is the Leslie-matrix, which is a discrete, age-structured model for population ecology. For an explanation of this method is referred to Van Boven & Schobben (1993). They developed a population-dynamical model for the Sandwich Tern in the Netherlands. The values of the survival rates were adjusted in this case in such a way that on average a constant population size is modeled by the Leslie-matrix. There are no clear indications that the survival or reproduction rates of terns in the Netherlands be influenced by the population density. Because of this, no density-dependent effects are incorporated in the Leslie-Matrix.

    Case Studies

    Case Study Results

    This section describes the results of the demonstration case: the probabilistic methodology is worked out for the effects of a fictitious dredging project on a population of Sandwich Terns.
    The impact of two different dredging scenarios is considered:

    1. Far field effects of large dredging project, which takes several years and;
    2. Far field effects of a (relatively) small, short duration dredging project.

    Two different approaches to the fish capture rate are elaborated:

    1. the conservative approach, assuming such a capture rate at optimal conditions, that any decrease of the capture rate leads to a lower breeding success;
    2. the empirical approach, estimating the capture rate at optimal conditions from measurements. (If sufficient data are available, a probability density function of the capture rate at optimal conditions can be estimated. In this case, implicitly also the natural variation of prey availability can be taken into account).

    Conservative approach

    The capture rate at optimal conditions is estimated at 8 fishes per day provided to the chicks. Any decrease of this number would directly lead to a decrease of the breeding success. Subsequently, the input variables of the population dynamical model are adjusted in such a way that the tern population in the reference situation is constant (on average).

    For dredging scenario 1, the impact on the number of breeding pairs was modeled for a range of background turbidities in the fictitious foraging area, see Figure 4 (van Kruchten & van der Hammen, 2011). The different probability density functions of SPM- and chlorophyll-a concentrations and the set of input variables for the different model runs are provided by van Kruchten & van der Hammen (2011). The figure shows the probability (y-axis) that the impact that will occur in reality is less (more positive value) than the possible impact X at the x-axis. A negative value at the x-axis means a decrease of the population size. For example: the results for C1 (the scenario with the highest average SPM concentration,23.0 mg/l ) show that the probability is 81% that the decrease of the number of breeding pairs will be less than 20%. Consequently, the probability of occurrence of a decrease of 20% or larger is 19%. For C4 (the scenario with the lowest average SPM concentration, 9.3 mg/l) the probability of occurrence of a decrease of 20% or more is close to zero. The Appendix explains in more detail how to read these graphs.

    As expected, the impact of dredging is larger at a foraging area with higher background turbidity. While the average Secchi disk transparency is 1.0 m in area C1, it is 1.9 m in area C4. At a transparency of 1 m, an increase of turbidity leads to a stronger decrease of capture probabilities than at a transparency of 1.9 m (see Figure 2).

    In addition, the impact of dredging scenario 2 on the number of breeding terns has been modeled by using the conservative approach (see Figure 5 and Figure 6). Figure 5 compares the results of dredging scenario 1 (pink line) and dredging scenario 2 (blue line); probability distribution for the relative change of the number of breeding pairs for background turbidity like C1: average SPM concentration of 23.0 mg/l; Figure 6 gives the same for background turbidity like C4: average SPM concentration of 9.3 mg/l (van Kruchten & van der Hammen, 2011).

    The results of dredging scenario 2 at low background turbidity (9.3 mg/l) show a relatively large probability of positive effects. This can be explained by the low background turbidity and the short period over which turbidity is increased by the dredging activities. On average the background turbidity (location C4) is lower than the optimal turbidity; the water is mostly too clear for optimal foraging conditions. If this is the case, an increase of silt concentration has a positive effect on the catchability of prey (see Figure 7 for TSM-concentrations < 20 mg/l and Chl-a concentrations < 20 μg, based on the equations of Suijlen & Duin (2000), Visser (1970) and Baptist & Leopold (2010), (van Kruchten & van der Hammen, 2011)). Only in years in which silt- and chlorophyll-a concentrations are substantially above average, an increase of turbidity due to dredging activities can have a negative effect on catchability. The probability that such background turbidity conditions (extreme for C4) occur simultaneously with the impact of the dredging, which is only noticeable during two years in scenario 2, is quite low. This explains why the probability of occurrence of positive effects is relatively large for this specific case.

    Capture rates based on measurements

    As long as no observations on capture rates or the amount of provided food are available for the possibly affected tern population, the precautionary principle requires using the conservative approach. In the present subsection the impact on tern populations has been modeled as if measurements of capture rates were available.

    Stienen & Brenninkmeijer (1998) observed that on average 13.4 fishes were provided to the nest per day. The optimal capture rate was adjusted to such a value that for dredging scenario 1 and background turbidity like C2 (16.3 mg/l), the average number of fishes provided was 13.4 per day. Figure 8 shows the probability distributions for the relative change of the number of breeding pairs under those conditions (dredging scenario 1, background turbidity like C2 (16.3 mg/l)) (van Kruchten & van der Hammen, 2011). The figure also shows the influence of the uncertain parameter values of the relation between food provided and breeding success; where relation 1 differs in parameter setting from relation 2.

    Alternative presentation of the impact of dredging

    In previous subsections the impact of dredging on tern populations was defined as the relative decrease of the number of breeding pairs, compared to the number of breeding pairs in the reference situation (without dredging).
    Also in the reference scenario a decrease of the number of breeding terns is possible. The results of van Kruchten & van der Hammen (2011) show that the probability that the population size will decrease autonomously over a period of 35 years is about 50% . As a consequence of dredging activities (increasing turbidity) the probability distribution for the population development will change (e.g. chance that the population will decrease autonomously becomes larger than 50%). The Secchi disk transparency in the reference situation was 1.34 m on average. If the transparency decreases for two successive feeding periods (year 1 and 2) to (constantly) 1 m, this hardly changes the probability distribution for the population development over 35 years. In case of a Secchi disk transparency of 0.8 m, the probability on a decrease of the population increases significantly, compared to the probability in the reference situation.

    Discussion

    Discussion

    To assess the impact of an activity on the environment, investigators often deal with limited knowledge about possible effects. Consequently assumptions have to be made, where the accumulation of worst case assumptions leads to unrealistically pessimistic estimates of the impact. A probabilistic approach will give more realistic estimates and also quantifies the uncertainty of this impact.
    The case on Sandwich terns, as described in van Kruchten & van der Hammen (2011), shows that it can be difficult to make a reliable assessment of possible impacts of dredging activities on a tern population, because:

    • most of the research reported in the literature is descriptive and does not quantify the links in the impact effect chain; as a result, many of those links remain quantitatively uncertain;
    • some of the relations in the chain that could not be derived from measured data, they were made probabilistic with a given probability distribution as they could not be based on measured data;
    • some pragmatic (and conservative) assumptions had to be made, by lack of information; consequently, the uncertainty margin of the expected impact on the tern population might differ from what the model results indicate;
    • the connection between "prey availability" and "time to catch enough food" had to be based on very limited data and limited background knowledge. As a consequence, the estimated relationship is highly uncertain. Therefore, it is recommended to collect more data on these connections in order to narrow down the uncertainty margins in the model results.

    Improving the modeling of impacts on tern populations

    Further research on the following topics is recommended to improve the model of the effects of dredging on Sandwich Tern populations:

    • How does an increase of turbidity affect the 'findability' of food by terns? A point of attention is the possible influence of wind waves (Taylor, 1983), as high turbidity levels often occur simultaneously with high wave events.
    • More measurement data on the relation between breeding success and the amount of food provided are desirable. This will enable a better founded estimate of the relation between these factors.
    • Do temporal variations in the amount of food that is provided to tern chicks have a significant influence on the breeding success?
    • Is the relation between water transparency and prey capture probability, found by Baptist & Leopold (2009), generally applicable?
    • Black headed gulls are known rob fish caught by terns (kleptoparasitism). If more quantitative information is available about the impact of kleptoparasitism on the amount of food provided to tern chicks, this effect can be incorporated in the model. A point of attention is the correlation between kleptoparasitism and turbidity, because both factors depend on weather conditions.
    • Do terns migrate to other colonies if food conditions are limiting at their regular breeding colony? If so, do they return to their former breeding grounds when the disturbance has disappeared?
    • The values of several parameters are estimated or based upon research on other tern species. More research on the values of these parameters for Sandwich terns is desirable. This would result in better estimates of the values and their probability density functions (pdf's). When insight into these pdf's is gained, it is recommended to consider whether or not these uncertainty margins can have a significant influence on the final results of the probabilistic modeling. In that case these uncertainty margins have to be taken into account in the probabilistic model.
    • Within the foraging area, possibly relevant spatial differences exist in turbidity and the presence of prey fish. A relation might exist between spatial differences in turbidity and the distribution of patches of fish over the foraging area. Research into this subject is recommended, as it may be relevant to incorporate the effects of the local differences in the modeling of the impact of dredging activities on tern populations.
    • Other visual hunting birds could be affected by sand mining in a similar ways as Sandwich terns. The vulnerability of a species will depend on several factors such as their diet restrictions and the location and size of their foraging grounds.

    Monitoring recommendations

    For the prediction of effects of specific dredging projects on tern populations, monitoring the following parameters in advance is recommended:

    • The number of fishes that can be caught by terns per day. If terns are able to catch a large number of fishes easily, the amount of available time for foraging is not limiting. On the basis of observations of prey capture rates, it would be possible to estimate a probability density function for the prey capture rate at optimal conditions. If this information is available, using conservative assumptions on this variable is not necessary anymore. Using measurements, instead of these conservative assumptions, leads to a more realistic and probably less pessimistic estimate of the impact on terns (see van Kruchten & van der Hammen (2011), Chapter 4, results Run2a and 5a).
    • The background turbidity at the foraging location (SPM- and chlorophyll-a concentrations). The background turbidity has a large influence on the expected impact from dredging activities (see Impact of dredging scenario 1 for different background turbidity conditions, where the average SPM concentration is 23.0 mg/l (C1); 16.3 mg/l (C2); and 9.3 mg/l (C4).
    • In addition, it is relevant to know whether large spatial differences exist in turbidity levels within the foraging area (see van Kruchten & van der Hammen (2011), section 4.2).

    If the probability of occurrence of significant effects, predicted in advance, is not negligible, monitoring capture rates and turbidity during the dredging process is recommended. By substituting the uncertain input variables in the model with the measured values, the prediction of the ecological impact can be adjusted. If the probability on significant effects increases to an undesirable value, adaptation of the dredging process will be desirable. Preferably the additional SPM-concentration, caused by the dredging process, is hindcasted during the dredging process. This makes it possible to distinguish the effects of changes in the background turbidity (reference situation) from the effects that are caused by the dredging activity.
    Regarding the adaptation of the dredging process, attention should be paid to the effectiveness of the possible measures. As tern populations are expected to be more affected by long term, far field effects of the dredging activities than by short term effects close to the dredging site, a certain time lag will exist between the moment of taking measures and the effect of these measures in the foraging area. It is recommended to do research on this time lag, as information on this time lag is highly relevant for the development of an effective, adaptive monitoring strategy. In case of the tern population: measures that are taken on the basis of today's monitoring results should have a positive effect in the foraging area before the breeding period ends.
    Information on the development of the tern population is also valuable. If information on the population size and composition is available, it can be used to validate the population dynamical model. In addition, in the assessment of the predicted impact, it may be relevant to know whether the population shows an autonomous increase or decrease.

    References

    • Ainley, D.G. (1977) Feeding methods in seabirds: a comparison of polar and tropical nesting communities in the eastern Pacific Ocean. In: Llano, G. A. (ed.) Adaptations within Antarctic ecosystems. Smithsonian Inst., Washington, D.C., p. 669-685
    • Baptist & Leopold, (2010). Prey capture success of Sandwich Terns Sterna sandvicensis varies non-linearly with water transparency. Internatioonal Journal of Avian Science, Ibis (2010), 152, 815–825.
    • Baptist, M. J., & Leopold, M. F. (2010). Prey capture success of sandwich terns sterna sandvicensis varies non-linearly with water transparency. Ibis, 152(4), 815-825.Baptist, M.J. & Leopold, M.F. (2007). De relatie tussen zichtdiepte en vangstsucces van de Grote Sterns van De Petten, Texel. IMARES, (Rapport / IMARES C097/07) - p. 45 (in Dutch)
    • Baptist, H. (2007). Lesliematrix bij kolonievogels, Toegepast voor Visdiefjes in de Westerschelde (Leslie Matrix in colony birds, Applied for Common Terns in the Westerschelde), Ecologisch Adviesbureau Henk Baptist, oktober 2007 (in Dutch)
    • Brenninkmeijer, A., Stienen, E.W.M., Klaassen, M. and Kersten, M. (2002a). Feeding ecology of wintering terns in Guinea-Bissau, Ibis (2002), 144, p 602-613
    • Brenninkmeijer, A., Doeglas, G. & De Fouw, J. (2002b) Foerageergedrag van sterns in de westelijke Westerschelde in 2002 (Foraging behavior of terns in the Western Scheldt in 2002), Altenburg en Wymenga ecologisch onderzoek, A&W-rapport 346 (in Dutch)
    • Dunn, E.K. (1972a). Effect of age on the fishing ability of Sandwich Terns Sterna sandvicensis. Ibis 114: 360-366.
    • Fasola, M. & G. Bogliani (1990). Foraging ranges of an assemblage of Mediterranean seabirds. Colonial Waterbirds 13: 72-74.
    • Haney, J.C. & A.E. Stone (1988). Seabird foraging tactics and water clarity: are plunge divers really in the clear? Mar. Eco. Prog. Ser. 49: 1-9.
    • Raaijmakers, M.H.J.E., P.W.M. van Beers & J.M.P.M. Habraken (1993). Telemetrische bepaling van de foerageerplaatsen en foerageerroutes van de op Griend broedende Grote Sterns Sterna sandvicensis (Telemetric determination of foraging sites and foraging routes on Sterna sandvicensis). Internal report. Instituut voor Bos- en Natuuronderzoek, Arnhem. (in Dutch)
    • Stienen & Brenninkmeijer (1998). Effects of changing food availability on population dynamics of the Sandwich Tern. BEON-rapport 98-3.
    • Stienen, E.W.M., A. Brenninkmeijer & C.E. Geschiere (1993). De relatie tussen de foerageermogelijkheden van de grote stern Sterna sandvicensis en het doorzicht van het water, alsmede de invloed van kleptoparasitisme door de kokmeeuw Larus ridibundus op de visaanvoer en het broedsucces van de grote stern: voorlopige resultaten (The relationship between foraging of the tern Sterna sandvicensis and the transparency of the water, and the impact of kleptoparasitism by black-headed gull Larus ridibundus on the feeding and the breeding success of the sandwich tern: preliminary results). IBN-DLO, DG-512.(in Dutch)
    • Stienen, E.W.M. & A. Brenninkmeijer (1994). Voedselecologie van de grote sterns (Sterna sandvicensis): onderzoek ter ondersteuning van een populatie-dynamisch model (Food ecology of terns (Sterna sandvicensis): research supporting a population dynamic model). IBN-DLO, IBN-rapport 120. 94 pp. (in Dutch)
    • Stienen, E.W.M. & A. Brenninkmeijer (1997). Fluctuaties in de lokale voedselbeschikbaarheid in relatie tot de populatiedynamiek van de Grote Stern Sterna sandvicensis: resultaten 1995-1996 (Fluctuations in local food availability in relation to the population dynamics of the Sandwich Tern Sterna sandvicensis: results 1995-1996). BEON project IBN 95 H 24. BEON rapport nr 97-1, 40 pp. (in Dutch).
    • Stienen, E.W.M., Van Beers, P.W.M, Brenninkmeijer, A., Habraken, J.M.P.M., Raaijmakers, M.H.J.E. & Van Tienen, P.G.M (2000). Reflections of a specialist: patterns in food provisioning and foraging conditions in Sandwich Terns Sterna sandvicensis, Ardea 88: 33-49
    • Stienen E.W.M. (2006). Living with gulls : trading off food and predation in the Sandwich Tern Sterna sandvicensis. Alterra scientific contributions 15, Proefschrift Rijksuniversiteit Groningen, ISBN 90-367-2480-5.
    • Stienen, E.W.M., W. Courtens & M. van de Walle (2007). Population dynamics of gulls and terns. In: Vincx, M., E. Kuijken & F. Volckaert (Eds.). Higher trophic levels in the southern North Sea (TROPHOS). Final report EV/25 Belgian Science Policy: Brussel. 89 pp.
    • Suijlen, J.M. & R.N.M. Duin (2001). Variability of near-surface total suspended matter concentrations in the Dutch coastal zone of the North Sea; Climatological study on the suspended matter concentrations in the North Sea. Rijkswaterstaat Report RIKZ/OS/2001.150X
    • Taylor, I.R. (1983). Effect of wind on the foraging behaviour of Common and Sandwich Terns. Omis Scandinavica 14: 90-96. doi:10.2307/3676011.
    • Thorpe, J.E. 1978. Rhythmic activity of fishes. Academic Press, London.
    • Van Boven, R.M. & Schobben, J.H.M. (1993). Risico-analyse voor een indicator-soort van het zeemilieu: De populatiedynamica van de grote stern in Nederland, (Risc analysis for an indicator species of the marine environment: The population dynamics of the sandwich tern in Netherlands), Rijkswaterstaat - Dienst Getijdewateren, Rapport DGW-93.006, maart 1993 (in Dutch).
    • Van Kruchten, Y. & van der Hammen, T. (2011). Case Study Sanwich Terns - a probabilistic analysis of the ecological effects of dreding IJmuiden : IMARES, (Report C055/11) - p. 52..
    • Veen, J. (1977). Functional and Causal Aspects of Nest Distribution in Colonies of the Sandwich Tern (Sterna Sandvicencis). Behaviour Supplement. 20:1-193

    Back to Top

    PDF