How well do food distributions predict spatial distributions of shorebirds with different degrees of self-organization?


  • Eelke O. Folmer,

    Corresponding author
    1. Animal Ecology Group, Centre for Ecological and Evolutionary Studies (CEES), University of Groningen, Haren, The Netherlands
      Correspondence author. E-mail:
    Search for more papers by this author
  • Han Olff,

    1. Conservation and Community Ecology Group, Centre for Ecological and Evolutionary Studies (CEES), University of Groningen, Haren, The Netherlands
    Search for more papers by this author
  • Theunis Piersma

    1. Animal Ecology Group, Centre for Ecological and Evolutionary Studies (CEES), University of Groningen, Haren, The Netherlands
    2. Department of Marine Ecology, Royal Netherlands Institute for Sea Research (NIOZ), Texel, The Netherlands
    Search for more papers by this author

Correspondence author. E-mail:


1. Habitat selection models usually assume that the spatial distributions of animals depend positively on the distributions of resources and negatively on interference. However, the presence of conspecifics at a given location also signals safety and the availability of resources. This may induce followers to select contiguous patches and causes animals to cluster. Resource availability, interference and attraction therefore jointly lead to self-organized patterns in foraging animals.

2. We analyse the distribution of foraging shorebirds at landscape level on the basis of a resource-based model to establish, albeit indirectly, the importance of conspecific attraction and interference.

3. At 23 intertidal sites with a mean area of 170 ha spread out over the Dutch Wadden Sea, the spatial distribution of six abundant shorebird species was determined. The location of individuals and groups was mapped using a simple method based on projective geometry, enabling fast mapping of low-tide foraging shorebird distributions. We analysed the suitability of these 23 sites in terms of food availability and travel distances to high tide roosts.

4. We introduce an interference sensitivity scale which maps interference as a function of inter-individual distance. We thus obtain interference-insensitive species, which are only sensitive to interference at short inter-individual distances (and may thus pack densely) and interference-sensitive species which interfere over greater inter-individual distances (and thus form sparse flocks).

5. We found that interference-insensitive species like red knot (Calidris canutus) and dunlins (Calidris alpina) are more clustered than predicted by the spatial distribution of their food resources. This suggests that these species follow each other when selecting foraging patches. In contrast, curlew (Numenius arquata) and grey plover (Pluvialis squatarola), known to be sensitive to interference, form sparse flocks. Hence, resource-based models have better predictive power for interference-sensitive species than for interference-insensitive species.

6. It follows from our analysis that monitoring programmes, habitat selection models and statistical analyses should also consider the mechanisms of self-organization.


In the theoretical resource-based literature, animal–habitat relationships are derived from fitness or intake maximization (Fretwell & Lucas 1969; Kacelnik, Krebs & Bernstein 1992). Intake rates are assumed to depend on resource density and interference (Beddington 1975; Ruxton, Gurney & De Roos 1992; Moody & Houston 1995). Under the assumption that animals behave ideally and freely and maximize intake rates, aggregative response functions may be derived (Sutherland 1983; Moody & Houston 1995; Van der Meer & Ens 1997). This approach bases predictions of the spatial distribution of foraging animals on straightforward mechanistic principles.

The empirical resource-based literature takes a phenomenological approach and investigates statistically the relationships between habitat characteristics and animal densities (Bryant 1979; Piersma et al. 1993; Yates et al. 1993; Zwarts, Wanink & Ens 1996; Guisan & Zimmermann 2000; Manly et al. 2002; Granadeiro et al. 2007). These studies find mixed results and heterogeneous relationships, amongst others because animal densities may depend on habit characteristics in nonlinear ways. Specifically, ecological factors may impose upper or lower limits on response variables so that the impacts within and outside the limits substantially differ (Thomson et al. 1996; Cade & Noon 2003). Put differently, ecological factors may operate as constraints on, rather than as exact determinants of behaviour. Moreover, multiple limiting factors may interact.

Both the theoretical and the empirical literature are pre-occupied with the negative impacts of co-occurrence of conspecifics while possible benefits are frequently ignored. The presence of many eyes and ears in a group increases the chance that predators (Pulliam 1973; Beauchamp 1998; Krause & Ruxton 2002; Whitfield 2003) or resources (Valone & Templeton 2002; Danchin et al. 2004) are detected. Additionally, animals may have developed social behaviour in response to past selection pressures (Byers 1997). Behaviour of individuals thus depends on behaviour of group-members (Sirot 2006). Therefore, not only negative aspects of interdependent relationships among individuals must be considered when studying habitat selection but also the positive aspects (Melles et al. 2009). It follows that the gregarious nature of animals may be a source of heterogeneity in the relationship between foraging animals and habitat characteristics.

Conspecific attraction is not necessarily beneficial, but may also lead to the selection of suboptimal foraging patches (Giraldeau, Valone & Templeton 2002). Specifically, if predecessors choose suboptimal foraging patches and followers copy their behaviour and select contiguous or nearby patches, a collective mistake results. Hence, conspecific attraction may lead to a mismatch between the spatial distribution of foraging animals and the spatial distribution of their food. For solitary species the risk of mismatch is smaller because patch selection will be based on expected intake rate only rather than on a combination of expected intake rate and conspecific attraction.

Recently, the concept of self-organization has been introduced to understand collective animal behaviour of groups without permanent leaders (Camazine et al. 2001; Krause & Ruxton 2002; Sumpter 2006). Central to this line of work is the notion that group formation results from repeated interactions among neighbours. These types of models view animals as interacting particles that make movement decisions in response to the locations and movements of their neighbours (Reynolds 1987; Couzin et al. 2002). This framework has been useful to understand and predict properties of groups with many individuals, such as insect swarms, fish schools and bird flocks (Sumpter 2006).

Flocks of shorebirds, particularly of dunlin (Calidris alpina) and red knot (Calidris canutus), may consist of many thousands of individuals displaying synchronized movements in flight, often in response to predation (Piersma et al. 1993; Van de Kam et al. 2004; Van den Hout, Spaans & Piersma 2008). These flocking patterns are maintained during foraging (Goss-Custard 1970). Despite their ubiquity, flocking patterns tend to be ignored in most studies of low-tide shorebird spatial distributions, as it is generally assumed that animal–habitat relationships only result from individual choices in response to resources and interference (Nehls & Tiedemann 1993; Piersma et al. 1993; Van Gils & Piersma 2004; Vahl et al. 2005; Van Gils et al. 2006; Spruzen, Richardson & Woehler 2008).

We hypothesize that shorebirds choose foraging patches based on exogenous factors (e.g. food availability, danger and travel costs) and, at varying degrees, in response to the presence of conspecifics (Fig. 1). Handling time and prey type determine the distance between conspecifics. For instance, oystercatchers (Haematopus ostralegus) foraging on bivalves, require long handling times making it possible for competitors to steal prey (kleptoparasitism) (Ens, Esselink & Zwarts 1990; Stillman et al. 2002). Hence, oystercatchers are sensitive to interference and therefore maintain relatively large inter-individual distances (Moody et al. 1997). In contrast, for species with short handling times (e.g. red knot) the cost of interference is small and animals may easily form dense flocks (Van Gils & Piersma 2004). Hence, they are interference-insensitive, i.e. there is a small impact of interference on spacing behaviour.

Figure 1.

 Clustering of foraging shorebirds as a function of food availability, attraction and interference. The darker the patch colour, the higher the food availability. 1: weak interference and no attraction: clustering at the patch with the highest food density; 2: weak interference and strong attraction: strong clustering at the patch with the highest food density; 3: strong interference and no attraction: weak-to-moderate clustering at the patch with the highest food density and weak clustering at patches with low food densities; 4: strong interference and attraction: moderate clustering at the patch with the the highest food density and weak clustering at the patch with the next highest food density.

The objective of this study is to test the hypothesis that resource availability and distance to high tide roost are more important determinants of the spatial distributions of interference-sensitive species than of interference-insensitive species. This will be reflected in a larger residual variance of a regression of bird density on these predictors for the latter than for the former. The reason is that in the case of interference-insensitive species systematic predictors (i.e. the joint impact of conspecific attraction and interference-insensitivity) are missing. The hypothesis will be tested, at landscape scale, for six common shorebird species in the Dutch Wadden Sea.

The Wadden Sea is an area par excellence to study resource availability–animal density relationships. First, many shorebird species in the Wadden Sea are abundant (Zwarts & Wanink 1993; Van de Kam et al. 2004). By focussing on abundant species, the role of accidental relationships (i.e. relationships that may occur by chance) is reduced. Secondly, there is detailed information available on food availability in the Wadden Sea because of an ongoing benthos monitoring programme (Piersma et al. 1993, 1995; Van Gils et al. 2007; Kraan et al. 2009a). Thirdly, there is large variation in food density and in the level of flocking between shorebird species (Goss-Custard 1970). Finally, the Wadden Sea is an open and well-known landscape such that the risk of not identifying possible confounding site characteristics affecting dispersion is small. Moreover, even if they are overlooked, they may not affect the analysis when they are constant between species.


The study area

The Dutch Wadden Sea is shallow and contains large soft-sediment flats that emerge twice a day during low tide. The mudflats alternate with permanent channels (Fig. 2). The flats are characterized by smooth gradients both in terms of physical properties, such as sediment grain size distributions, and biological properties, such as density of macro-zoobenthic species (Kraan et al. 2009a). As a result of the semi-diurnal tides, the mudflats are accessible to shorebirds approximately twice per day. High tide roosts of non-breeding shorebirds are found on the mainland and on all islands (Koffijberg 2003; Van de Kam et al. 2004).

Figure 2.

 Map of the Dutch Wadden Sea. Benthos sampling stations and the sites for which shorebird distributions were mapped are indicated.

Benthos sampling

As part of a long-term benthic research programme (Piersma et al. 1993; Kraan et al. 2009a, 2009b), we determined the density of macrozoobenthos in the Dutch Wadden Sea between July and September 2004. Benthos sampling was performed over 250 m grids (Fig. 2). The sampling stations were visited by foot during low tide and by boat during high tide (by boat to maximally utilize the number of working hours while in the field).

When sampling by foot, one sample was collected at each station. Each sample consisted of sediment taken down to a depth of 20–25 cm with a core with area of 1/56 m2. The top (0–4 cm) layer of the sample was separated from the bottom layer. The top and bottom layers were sieved separately over 1-mm mesh. As polychaetes are able to move from the bottom to the top part layer, their vertical location in the layer was not recorded. At the same locations, mudsnails (Hydrobia ulvae) were also sampled but with a smaller core (1/267 m2) to a depth of 4 cm. Mudsnail samples were sieved over a 0·5-mm mesh. When sampling by boat, at each station two samples were taken, down to a depth of 20–25 cm, each with a core with area of 1/115 m2. We took two samples to obtain similar precision of benthos density estimates as in the samples collected by foot. The two samples were sieved jointly. As a result of practical limitations, for these samples the top layers were not separated from the bottom layer.

In the field, the numbers of adult and juvenile individuals of each macrobenthos species were counted. All molluscs and shore crabs (Carcinus maenas) that were retained in the sieve were frozen at −20 °C for later analysis in the laboratory. In the laboratory the lengths of all individual specimens were measured to the nearest 0·1 mm. For bivalves, the flesh was separated from the shell and dried at 55–60 °C. After determination of the dry mass (to the nearest 0·1 mg), the flesh was incinerated at 550 °C for 2 h. The weights of the ashes were measured to the nearest 0·1 mg. In this way species and length-specific values for ash-free dry mass (AFDM) were obtained. Further details about prey sampling and analysis can be found in Piersma et al. (1993, 1995, 2003) and Kraan et al. (2009a) studies.

For the specimens counted in the field and not brought to the lab (polychaetes and isopods), we obtained estimates of energy values from the literature (Appendix S1). Note that the values thus obtained are approximations. This, however, is not a problem in the present analysis, as its objective is to analyse the significance of food as predictor of patch choice and flocking variance rather than precisely estimating and comparing regression coefficients of food variables. Particularly, some inaccuracy in the regression coefficients does not affect the predictive power of the estimated models for the entertained objectives of the paper. Moreover, we also considered higher than conventional (5%) significance levels of the coefficients.

Bird mapping

For the 23 sites that were sampled for benthos, shorebird distributions were mapped (Fig. 2). The maps were drawn between 3 days before and 3 days after benthos sampling. Several studies (e.g. Piersma et al. 1993; Van Gils et al. 2003), show that depletion and death of benthic species affecting their densities occur over longer time periods than 6 days.

Benthos sampling and shorebird mapping took place around the centres of the mudflats where submersion times are shortest. This ensures accessibility of the mudflats for most of the time throughout the tide. Observation points were chosen centrally on the mudflats (>1 km away from gullies). Bird distributions were mapped in between 2 h before until 2 h after low tide. The area of exposed mudflat changes little in this time span so that the spatial distribution of the birds is not affected by tidal movement. Furthermore, disturbance because of the presence of the observer is minimal under these conditions, because the extent of available mudflat is at its largest.

The observer (EOF) arrived at the observation points by foot well before mapping started, so that disturbed birds would have sufficient time to return to the areas through which the observer had arrived. Observations were started in opposite direction from which the observer had arrived. Only Curlew (Numenius arquata) seemed disturbed and was never recorded within 200 m from the observer. Positions of individuals and flocks were determined with the aid of GPS, compass and rifle scope with a ranging reticle (mill dots). GPS was used to determine the position of the observer; the compass to determine the observation direction. The rifle scope mounted on the telescope enabled the observer to measure the distance between each individual bird or flock edges and the horizon in terms of mill dots. Based on principles of projective geometry this distance was used to calculate the true distance from the observer (Heinemann 1981). The procedure was regularly calibrated using objects with known locations and true distances. All individual birds and flocks in a 360° circle around the observer were plotted on maps with 100 m grids. Individual birds were plotted as points and flocks as polygons in which the numbers of individuals were registered. In some cases flocks rather than individuals were considered because when the number of birds covering a small area was large, it was not feasible to plot all individual positions.

In early morning and late afternoon, visibility could be poor because of reflecting light making it impossible to make a full 360° map. In those cases observations in the direction of poor visibility were cancelled. During a single low-tide period, depending on the average bird density, either one or two censuses were performed on different sites on the same mudflat. A typical census would relate to a circular area with radius between 650 and 800 m. This census area, denoted ‘site’, is the spatial unit of analysis below.

Data preparation

Regarding benthos availability for short-billed birds potentially feeding on small bivalves (i.e. red knot and dunlin), we only considered bivalves in the top layer (Van Gils et al. 2009). As it was not possible to separate the top and bottom layers for samples collected by boat, we obtained estimates for the top layer benthos in this case by using the proportion of top layer benthos found in samples collected by foot. The proportions of benthos in the top and bottom layers may differ between species, size classes and between the eastern and western Wadden Sea (Van Gils et al. 2009). We therefore used species-, size- and location [western (a–c) and eastern (d–l) Wadden Sea (see Fig. 2)] specific proportions. For example, if the proportion of top layer ingestible Baltic tellin (Macoma balthica, <16 mm) in the samples collected by foot in the eastern Wadden Sea turned out to be 75%, the amount of ingestible top layer Macoma in samples collected by boat in the eastern Wadden Sea was obtained by multiplying the total amount of ingestible Macoma by 0·75.

All benthos samples and bird maps were organized in a GIS. Digital point maps of bird distributions were constructed by digitizing the scanned and geo-referenced field maps. Flocks were represented as polygons in which the numbers of birds were registered. The points inside the polygons were distributed evenly (by hand) over the polygon area. Single birds were plotted as individual points.

The points thus obtained were aggregated in 50 × 50 m grids (i.e. cells) that fully covered the censused sites. The number of birds inside a gridcell was transformed to density and related to its centroid. Only cells with more than 50% of the area inside the site were included in the data set. The resulting lattice formed the basis for calculating occupancy and degree of packing.

For the landscape-level analyses, the bird and benthos data sets were aggregated to site level resulting in 23 data points for all species, except red knot where the number of data points is 16. The reason is that the population of red knots in the Wadden Sea is highly variable in August because of turnover of two distinct populations. By the beginning of September members of the canutus subspecies have departed while the other subspecies, islandica, has arrived (Zwarts, Blomert & Wanink 1992; Piersma et al. 1993; Nebel et al. 2000; Kraan et al. 2009b). For red knot we only considered observations after 1 September.

Bird density was calculated by dividing the total number of individuals by the area of the site corrected for the disturbance effect of the observer. Depending on species-specific sensitivity, we subtracted the area around the observer calculated by π × r2 where r is the distance over which the animals are disturbed. We used the following distances: dunlin and red knot: 150 m, oystercatcher, grey plover (Pluvialis squatarola) and bar-tailed godwit (Limosa lapponica): 200 m, curlew: 300 m (Spaans, Bruinzeel & Smit 1996). The site level density of each benthos species was calculated by averaging the benthos densities of the sampling stations that were inside the site but outside the disturbed area. A map with locations of high tide roosts was used to calculate the distance between the centroid of each site and the nearest high tide roost (Koffijberg 2003).

Packing of individuals relates to inter-individual distance; it indicates the local density of animals. It was obtained by dividing the density by the proportion of occupied cells (birds ha−1).

Statistical analysis

The landscape-level model

We investigated the landscape–level relationship between the spatial distribution of foraging shorebirds and its predictors with a linear model. The dependent variable is density as defined above. To normalize the data, bird densities were log-transformed (Gelman & Hill 2006). We added the value of 1 to avoid taking logarithms of zero.

We used benthos availability and travel distance to high tide roosts as predictors. For each benthos species and sampling station, AFDM values were obtained by summing the AFDM of the benthos items that were ingestible and accessible (i.e. for red knot and dunlin: only small bivalves from the top layer; for long-billed shorebirds: benthos from both layers). (Appendix S1 shows a synopsis of the literature on the summer diet for the six abundant shorebird species.) Profitability and digestibility may differ widely between prey species, even after adjustment for caloric values (Zwarts & Blomert 1992). Therefore, densities of the prey species were entered as separate variables and not combined to give an overall measure of food availability. To avoid spurious correlations, we only used the benthos species that were known to be regular prey and reasonably abundant in the Wadden Sea.

Benthos items are assumed to have a positive or zero impact on the dependent variable. Negative impacts are ecologically implausible, because a shorebird may ignore, but will not be deterred, by benthos. As large travel distances from high tide roosts to foraging sites imply extra time and energy costs, they negatively affect density (Dias et al. 2006; Van Gils et al. 2006; Rogers, Piersma & Hassell 2006). Therefore, we hypothesize a negative impact of this variable. Because there is no a priori reason to expect interactions among predictors, and because the number of predictors for some species is large, only linear and additive combinations of the predictors were considered.

For each bird species, we applied the following modelling procedure. First, we estimated the initial (full) model based on food availability and distance to high tide roosts. Next, we reduced the initial model applying a stepwise, backward procedure in that predictors with ecologically implausible coefficients were deleted, i.e. a positive coefficient for distance to nearest high tide roost and negative coefficients for benthos items. In case of several incorrect coefficients, the former was deleted first. Food predictors with negative coefficients were deleted one by one in order of increasing P-values. The model thus obtained is labelled ‘ecological model’. It is plausible on the basis of ecological considerations and permissive in that higher than conventional P-values are accepted.

The ecological model was further reduced on the basis of statistical criteria to find the model with the best predictive power. We therefore selected the final model, from all possible ecological models, on the basis of minimization of the corrected Akaike’s information criterion (Burnham & Anderson 1998). Data were analyzed with the statistical package R 2.9.0 (R Development Core Team 2009).

Standard deviation of the residuals (σ) and packing

The variance of the residuals of the regression of bird density on food availability and distance to high tide roosts is expected to be higher for gregarious species than for solitary species because of stronger predictive power of benthos for the latter than for the former. This will be reflected by the standard deviation of the residuals (σ). Particularly, for solitary species we expect σ to be smaller than for gregarious species. The relationship between gregariousness and σ was tested by regressing the latter on packing.


Number of birds and density

We counted approximately 26 000 birds at 23 sites on 12 mudflat areas covering a total of 3943 ha (Fig. 2). Dunlin was the most abundant species (12 884 individuals) followed by red knot (5654 on 16 sites), oystercatcher (5365), curlew (887), bar-tailed godwit (604) and grey plover (245).

Figure 3 shows densities by species and site, showing large variations. The summary in Fig. 4a shows that red knot and dunlin had the highest densities, followed by oystercatcher, bar-tailed godwit and curlew, with grey plover having the lowest density. Especially, red knot and dunlin showed high variability while the other four species occurred at relatively constant densities between sites (Fig. 3). When red knot or dunlins were observed at a particular site, there typically were many of them forming dense flocks. In contrast, grey plover and oystercatcher occurred at all mudflats in relatively constant numbers.

Figure 3.

 Species-specific variation in bird density among sites. Sites ordered from west to east. Each letter corresponds to a unique mudflat. Sites are labelled by a combination of letter and number. Observe the different scales on the y-axis. Panels are ordered from top to bottom in order of increasing variance.

Figure 4.

 (a) Mean density and (b) mean packing of six shorebird species. Length of error bars correspond to standard errors. Packing is calculated by dividing the site density by the proportion of occupied cells. Estimates of packing are based on the sites where more than 5% of the 50 × 50 m cells were occupied.

Packing patterns

The packing of individuals varied strongly between species (Fig. 4b). Red knot and dunlin, when encountered, occurred in high local densities. Bar-tailed godwit, grey plover, oystercatcher and curlew showed a much lower degree of packing.

Regression of density on food availability and distance to high tide roosts

Table 1 shows the full, ecological and final model estimates including regression coefficients, standard errors as well as several goodness-of-fit measures. The ecological models show that for each bird species there are benthos items with ecologically plausible coefficients. Distance to high tide roost has a correct negative sign for oystercatchers, curlew and grey plovers only.

Table 1.   Regression models of shorebird density for six species on landscape scale in the Wadden Sea Thumbnail image of

Compared with the ecological models, the final models generally show a substantial reduction of predictors. For oystercatcher, however, the full and ecological models are the same. Distance to high tide roost has dropped out for every species in the final model. Moreover, we find that for dunlin only Nereis diversicolor is a significant predictor; for bar-tailed godwit Arenicola marina, Nephtys hombergii and Scoloplos armiger; for oystercatcher Nereis diversicolor and Cerastoderma edule; for grey plover Arenicola marina and for red knot Mya arenaria.

For curlew there are no significant food predictors in the final model. This may be related to the wide variety of prey that curlews select (Appendix S1). Furthermore, individual specialization on specific prey species probably takes place (Leeman et al. 2001; Bolnick et al. 2003). Curlews have a preference for large Carcinus above other prey species in summer (Goss-Custard, Jones & Newbery 1977; Petersen & Exo 1999; Ens et al. 1990). Densities of these large shore crabs were probably not adequately determined by our sampling method. We therefore also tested whether mudflat elevation [obtained from the National Institute for Coastal and Marine Management (Het Rijksinstituut voor Kust en Zee (RIKZ)), The Netherlands, data collected between 1997 and 2002], silt content (Zwarts et al. 2004) and distance from high tide roost impacted on curlew density. The final curlew model based on silt content (coefficient = 0·0215, P = 0·003) and mudflat elevation (−0·003, P = 0·052) and distance from high tide roost (coefficient = −0·056, P = 0·112) indicates a preference for muddy and low sites that are near the high tide roosts (R= 0·39 and σ = 0·27). Note that the standard deviation is slightly smaller than for the food model.

As discussed above, the missing of systematic predictors is reflected in the R2 of the final models. The R2 vary from more than 0·50 for oystercatcher and bar-tailed godwit to <0·30 for the gregarious dunlin and red knot. It follows that for gregarious shorebird species important systematic predictors are missing, i.e. the joint impact of conspecific attraction and interference-insensitivity.

Regression of the residual standard deviation (σ) on packing

Figure 5 shows that the residual standard deviation (σ) is positively related to packing (slope coefficient 0·032 ± 0·009; F = 14·31; d.f. = 4; P = 0·019; R= 0·78). As hypothesized, for the solitary species curlew, oystercatcher, bar-tailed godwit and in particular grey plover, we find relatively small σ values but for the gregarious red knot and dunlin large σ values.

Figure 5.

 Regression of standard deviation of residuals (σ) on packing. σ was obtained from the regression models in Table 1. In the regression model shorebird density was ln-transformed so that σ also is on ln-scale. Packing is defined as local bird density. The horizontal bars correspond to the standard errors of packing. Regression equation: σ = 0·198 (0·097) + 0·032 (0·009) × packing; P = 0·019; R=0·78. The numbers between parentheses correspond to standard errors.


The main finding of this study is that the predictive power of a resource-based model for the instantaneous spatial distribution of foraging shorebirds deteriorates with the tendency to flock, because the presence of conspecifics may be taken as an indication of the absence of predators or the availability of food (Krause & Ruxton 2002). Positive feedback in the form of conspecific attraction in combination with insensitivity to interference limits the predictability of the spatial distribution of foraging shorebirds by food availability and distance to high tide roost (E.O. Folmer, H. Olff & T. Piersma, unpublished data). The mere presence of food availability and absence of conspecific attraction in a bird density model implies a correctly specified model for solitary species, but a misspecified model for gregarious species leading to an increase in residual variance.

We hypothesized handling time to be a decisive factor with respect to the tendency to group or not. The shorter the handling time, the less sensitive a species would be to e.g. kleptoparasitism, and thus the shorter inter-individual distances that need to be maintained and the greater the tendency to flock. Shorebirds such as dunlin and red knot that forage on small prey requiring short handling times do not suffer much from interference of nearby conspecifics allowing them to benefit from nearby conspecifics (Nehls & Tiedemann 1993; Van Gils & Piersma 2004). Such species therefore occupy a relatively small proportion of the suitable habitat which makes it hard to predict their instantaneous distributions. This may also explain why for such species suitable areas often are not occupied for some time, as observed for red knot by e.g. Piersma et al. (1993) and Van Gils & Piersma (2004). Such absences are less common for interference-sensitive species (oystercatchers, grey plovers and curlews) that maintain large minimal distances from conspecifics (Vines 1980; Moody et al. 1997). For interference-sensitive species a large proportion of suitable habitat becomes occupied which strengthens the predictive power of a resource-based model. For future research, we suggest that attention should also be paid to other factors that may influence inter-individual distances. For example, if small shorebirds were more vulnerable to predation than larger ones, shorebird size would correlate with packing, i.e. small species would flock more densely than large species.

Many studies indicate that complex micro-level relationships may become simple at aggregated levels (Levin 1992). Here we show the opposite: complex patterns at landscape level arise because of small-scale interactions, i.e. flocking behaviour. Hence, the key to prediction and understanding of landscape-level patterns of shorebirds also lies in the elucidation of their social behaviour.

This study provides some insight into the question whether space or food availability limits population size. It follows from the above that species that are interference-insensitive are merely limited by total food resource stocks, whereas for interference-sensitive species both resource availability and the extent of foraging habitat are important.

Traditionally, miss-matches between the spatial distributions of resources and animals have been explained by perceptual constraints (Abrahams 1986), despotism (Fretwell 1972) and inequality amongst competitors (Parker & Sutherland 1986). The joint impact of conspecific attraction and interference adds an additional explanation: for social species actions based on perceptions (right or wrong) are amplified by the collective behaviour of the members of a group, as suggested by Abrahams (1986).

Adequate management of natural reserves depends on the quality of the information about behaviour and distributions of its animal populations. Monitoring programmes should be designed such that all the behavioural and distributional determinants are addressed. We have shown that, based on the distribution of animals, habitat suitability is more difficult to determine for scarce and gregarious species than for abundant, solitary species. Monitoring programmes should therefore also take into account to the ‘gregarious nature’ of the species. For solitary and abundant species random sampling is adequate. Gregarious species need to be followed over longer periods of time or at larger spatial scales, as shown by Piersma et al. (1993) and Colwell et al. (2003) who find that the processes driving instantaneous spatial heterogeneity between sites also underlie heterogeneity along the time axis on a day to day basis on mudflats and on roosts.

The finding of this study that social species tend to occupy a relatively small proportion of the available and suitable habitat contrasts with findings by Piersma et al. (1993, 1995) and Van Gils et al. (2006). They conclude that over a whole season (or even years; Piersma et al. 1993) the cumulative distribution of interference-insensitive shorebird species (red knots), taking the distance to high tide roosts into account, matches relevant food distributions. In these studies the spatial and temporal scales, method of data collection and the statistical models and criteria used differ from the ones employed in the present study. Comparison of results is hence not straightforward. Further research on the statistical relationships of gregarious species with their exogenous predictors observed over long-time periods and large spatial scale is needed to reconcile these findings.


A. Dekinga and C. Kraan organized and carried out much of the benthic sampling in the western portion of the study area. B. Spaans and M. Brugge provided valuable support to the benthic sampling in the eastern part of the Wadden Sea. We thank K. van de Star, T. van der Vis and H. de Vries, crew of the Royal NIOZ-research vessel RV Navicula, for their help. Many volunteers contributed to the collection of the benthos samples. C. Kraan did all the laboratory work on the samples from the western part of the Waddden Sea; M. Lange and E. Marks most of the laboratory work on the eastern Wadden Sea samples. We thank C.J. Camphuysen for suggesting the field method of measuring distances. Jan A. van Gils and an anonymous reviewer provided valuable comments on earlier versions of the manuscript.