Shorebirds exhibit niche partitioning on multiple dimensions at a small staging site on the Northumberland Strait, New Brunswick, Canada

We examined staging ecology of shorebirds at Petit-Cap (PCAP) and Cape Jourimain (CJ) on the Northumberland Strait in New Brunswick, Canada (Fig. 1). Behavioural observations and habitat samples were taken at both sites, but shorebirds were captured only at PCAP.

We used a stratified random sampling design to sample the mud- and sandflats at PCAP and CJ to estimate prey availability. Transects were 300 m long, and ran perpendicular to shore, starting at the high tide line. At CJ, we placed three transects 100 m apart. At PCAP, we placed two transects facing the ocean on the open side of the barrier beach, 400 m apart, and three transects facing the interior flats, 100 m apart. We collected invertebrate and biofilm samples from two random points within each 100 m stratum, for a total of six samples per transect. We replicated transect sampling early (July 17–18, 2019) and late (September 6–12, 2019) in the season using different sets of random points.

We collected invertebrate samples with a 5.5 cm diameter sediment core, and sieved samples through 500 µm mesh to remove sediment. Retained invertebrates were stored in 95% ethanol until they could be sorted by taxon. Sorted invertebrates were dried at 90 °C for at least 24 h (40GC Gravity Convection Oven, Quincy Lab Inc., Chicago, IL) and then weighed. Following weighing, shelled invertebrates (phylum Mollusca) were ashed at 550 °C for 2 h (Isotemp Programmable Muffle Furnace 650-750 Series, Fisher-Scientific, Waltham, MA) to burn off consumable biomass and reweighed. The difference between pre- and post-ashing mass (ash free dry mass) was used to estimate consumable biomass.

We observed foraging behaviour of shorebirds between June 17th and October 9th, 2019 at PCAP and CJ. Behavioural observations took place throughout the day during periods when tidal flats were exposed. We observed foraging behaviour of flocks with more than five individuals and recorded whether the flock was foraging in a puddle. We used focal animal sampling (Altman 1974) to estimate pecking and probing rates of foraging shorebirds by recording the number of times an individual shorebird pecked or probed for a maximum of 1 min. Focal animal samples were taken on each bird in a flock, or until the flock flew away. We also simultaneously performed scan sampling (Altman 1974) on the same shorebird flocks to determine the average number of individuals of each species in the flock throughout the focal observation period. This method was used to account for individuals joining or leaving the flock throughout the sampling time.

After collecting behaviour observations, we took two invertebrate samples from the location at which each flock was seen foraging, following the methods described for transect sampling. Flocks were flushed to take samples, or in the case flocks flushed during observations due to disturbance or unknown reasons, we sampled invertebrates immediately after flushing to ensure the most accurate representation of the prey base being consumed. At each foraging location, we also measured the penetrability of sediments (Gerwing et al. 2020) and collected two sediment samples and four biofilm samples. Sediment samples were used to analyze sediment particle size and water content, and were collected using 3 cm diameter vials pressed approximately 8 cm into the sediment. We recorded the wet weight of a small, homogenized subsample of each sediment sample, then dried at 90 °C for at least 24 h and reweighed. We calculated the proportion water content in each sample as (wet mass − dry mass)/wet mass. We sieved the remaining sediment in each sample through a cascade of 850, 500, 250, 125, and 63 µm mesh sizes, and collected water and suspended sediment that flowed through all mesh sizes in a bucket under the finest sieve, termed the pan fraction. A particle size profile for each sample was calculated by drying and weighing material from each fraction, and calculating the proportional contribution of each fraction. Pan fraction sediment mass was estimated by collecting a 20 mL subsample from the well-mixed water and sediment suspension, drying and weighing that sediment, and scaling up the mass to reflect the total volume in the pan fraction.

Biofilm samples consist of the top 2 mm of sediment, collected using a modified 10 cc syringe following Coulthard and Hamilton (2011). We measured chlorophyll a (Chl a) concentration, a proxy for biofilm standing crop, from each sample. Sediments were freeze dried at −50 °C and 0.05 mBar (FreeZone one-liter benchtop freeze drier, Labconco Corporation, Kansas City, MO) for a minimum of 5 h. We extracted Chl a from a 0.2 g subsample of the sediments with 5 mL of 3:2 90% acetone to dimethyl sulfoxide for 30 min (Shoaf and Lium 1976). We centrifuged samples for 20 min at 2500 rpm (Thermo Fisher Scientific Model No. 225A, Waltham, MA; Hettick MIKRO 120, Beverly, MA) to separate the extracted Chl a solution before measuring Chl a concentration with fluorometric techniques (Welschmeyer 1994) (10 AU Chl a Fluorometer, Turner Designs, San Jose, CA). We converted Chl a concentration from relative fluorescence units (RFU) to µg/L using the calibration curve y = 7.064x – 0.7656, where y = RFU and x = µg/L. We then calculated concentration of Chl a per gram of sediment using the following formula:

To collect blood plasma samples for analysis of stable isotopes of ¹⁵N and ¹³C, we captured Semipalmated Sandpipers, White-rumped Sandpipers, Semipalmated Plovers, Short-billed Dowitchers, and Least Sandpipers at PCAP between August 1st and September 13th, 2019. We set up mist nets in arrays across the intertidal zone and captured birds at night when tides were appropriate. Shorebirds were weighed upon capture, and we collected approximately 140 µL of blood from the brachial vein of those that met species-specific minimum weight requirements (Table 1). Shorebirds arrive at staging sites light and become heavier over time as they forage (Tsipoura and Burger 1999; Quinn and Hamilton 2012). Plasma isotope levels reflect diet in approximately the previous week (Hobson and Clark 1993). Therefore, restricting sampling to birds with moderate levels of body fat compared with lean mass estimates (Anderson et al. 2019) coupled with the rapid turnover of plasma helps to ensure plasma samples reflect diets of shorebirds during their stopover in the region. Following bleeding, we fitted shorebirds with aluminum USGS bands on the upper left leg and field-readable alphanumeric flags on the upper right leg and took morphometric measurements for ongoing shorebird banding initiatives, and then released the birds. Blood samples were stored on ice. After returning from the catching site, we centrifuged blood samples for 1 min at 6300g (mySPIN12 Mini Centrifuge, Thermo Scientific) to separate plasma and red blood cells (RBCs). Separated plasma and RBC samples were stored at −20 °C until further analyses. All methods of catching and sampling shorebirds were reviewed and approved by the Mount Allison University Animal Care Committee prior to commencement of the research and adhere to guidelines provided by the Canadian Council for Animal Care.

We collected 139 blood plasma samples from shorebirds of various species at PCAP for analysis of stable isotopes of ¹³C and ¹⁵N. Briefly, δ¹³C values reflect prey origin (e.g., terrestrial prey are more enriched than marine organisms), and δ¹⁵N values reflect trophic level, with higher values found in organisms from higher trophic levels (Kelly 2000). While we did not also collect samples from birds at CJ, we know from previous work in the region that shorebirds make frequent, short movements between these and other sites in the Northumberland Strait (Doiron 2021; Linhart et al. 2023). This, coupled with the fact that the prey base does not differ significantly between PCAP and CJ (Linhart et al. 2022), suggests that plasma isotopes of shorebirds captured at PCAP likely reflect diets across a range of local foraging habitats.

We dried plasma samples at 70 °C for at least 24 h. The target weight for analysis of both δ¹³C and δ¹⁵N from blood plasma was 0.800 mg. We subsampled dried plasma samples using a Mettler-Toledo MX5 microbalance with readability to 0.001 mg, and subsamples were analyzed for δ¹³C and δ¹⁵N signatures at the Environmental Analytics and Stable Isotope Laboratory at Mount Allison University in Sackville, New Brunswick, Canada, using an Elementar PyroCube Elemental Analyzer (Elementar Analysensysteme GmbH, Hanau, Germany) and an Isoprime Precision Isotope Ratio Mass Spectrometer (Elementar UK Ltd, Cheadle, UK). Delta values of isotope signatures are a relative isotope ratio of the sample to international standards based on the following formula:where a = the heavier isotope, X = the element of interest (nitrogen or carbon), and R = the ratio of heavy to light isotope.

Analyses were performed using R (version 4.0.3) with RStudio interface. Nonparametric methods were used as data failed to meet parametric assumptions. Dissimilarity matrices were created using the Bray–Curtis method unless otherwise specified. Data were plotted using the “ggplot2” package (Wickham 2016). Test statistics were compared with an alpha of 0.05. We adjusted the alpha level with the Holm's sequential Bonferroni procedure (Holm 1979; Abdi 2010) for multiple comparisons within datasets.

We fourth root transformed the abundance of shorebird species in each observed flock to improve consideration of rare species. We used a permutational multivariate analysis of variance (PERMANOVA) in the R package “vegan” (Oksanen et al. 2020) to compare species composition between flocks foraging in and out of tide pools (permutations = 999). We then performed a similarity percentage (SIMPER) analysis to compare the contributions of each species to the dissimilarity in species composition. Using untransformed abundance data, we calculated the mean and standard error of the abundance of each species in and out of tide pools.

We examined spatial variation in prey availability on the Northumberland Strait using data collected from transects at PCAP and CJ. We grouped prey into the following taxonomic groups: amphipods, bivalves, decapods, gastropods, polychaetes, and meiofauna. Shelled prey >12 mm long were excluded as they are too large to be consumed by most shorebirds in our study (Kober and Bairlein 2006). We performed a nested PERMANOVA (R package BiodiversityR; Kindt and Coe 2005) with transect nested in site to examine spatial variation in prey availability (permutations = 9999). We used a Mantel test (R package vegan; Oksanen et al. 2020) to test for a relationship between flock species composition (based on average number of individuals of each species in the flock during scans of that flock) and the available prey community where the flock was observed foraging (permutations = 99 999; correlation method = Spearman's coefficient). This was performed with a Bray–Curtis dissimilarity matrix of shorebird species abundances in each flock, and a Euclidean distance matrix of biomasses of prey taxa collected from the location each flock was foraging.

We analyzed foraging strategies of seven shorebird species; however, Greater Yellowlegs and Lesser Yellowlegs were pooled together as “Yellowlegs”. Pecking and probing rates were calculated in actions per minute. We used a permutational multivariate analysis of covariance (PERMANCOVA) to examine differences among species in foraging rates while controlling for habitat covariates (permutations = 9999). The full model tested pecking and probing rates against species and nine covariates: availability of six prey taxa, Chl a concentration, and two components extracted from a principal component analysis of sediment characteristics. Species composition of the flock was also considered as a covariate, but was nonsignificant (Table S1), so it is not included in the main model to improve sample size (one flock lacked composition data). The first component primarily reflected sediment water content and the second sediment particle size. We removed nonsignificant covariates and reran the model. We compared the multivariate dispersions of foraging strategies between species using the permutest function (vegan package; permutations = 9999; Oksanen et al. 2020). Data were square root transformed prior to plotting to improve visualization of results. We performed 15 pairwise PERMANCOVAs using the covariates from the reduced model to compare foraging strategies between each species (permutations = 9999), with critical alpha levels corrected for multiple comparisons.

Isotopic values of δ¹³C and δ¹⁵N were measured from Semipalmated Sandpiper, White-rumped Sandpiper, Least Sandpiper, Semipalmated Plover, and Short-billed Dowitcher plasma (Table 1). Too few Least Sandpipers were sampled to be included in analyses, and one Semipalmated Sandpiper sample was excluded as the height of its N₂ mass spectrometry peak fell below quality assurance limits. We visually inspected plasma isotope data to ensure individual birds’ isotope signatures did not vary with bird mass, which might have suggested influence of prey from elsewhere in the range (Catry et al. 2022). We used a PERMANOVA to test δ¹³C and δ¹⁵N against species (permutations = 9999) and a permutest (permutations = 9999; Oksanen et al. 2020) to compare multivariate dispersions of δ¹³C and δ¹⁵N among species. We plotted the isotopic diet niches using ggplot2, following the methods of stable isotope Bayesian ellipses in R (R package SIBER; Jackson et al. 2011).

Species composition of flocks of foraging shorebirds varied between areas with and without available tide pools (PERMANOVA, F_1,19 = 4.76, p = 0.02). Semipalmated Sandpipers and Semipalmated Plovers were most abundant outside of tide pools, while Short-billed Dowitchers were most abundant in tide pools (Table 2). The remaining species had similar abundances in both habitats (Table 2). SIMPER analysis showed that the contribution to dissimilarity in species composition between flocks observed in and out of tide pools was similar among species except that White-rumped Sandpipers and Least Sandpipers, which were less abundant overall, contribute slightly less than other species (Table 2). Non-metric multidimensional scaling (NMDS) ordination separates flocks with a high abundance of Short-billed Dowitchers and Yellowlegs (both long-legged birds) from those with a high abundance of the smaller Semipalmated Plovers and Calidrid sandpipers (Semipalmated, Least, and White-rumped Sandpipers) (Fig. S2).

There was considerable spatial variation in the availability of invertebrate prey within coastal sites on the Northumberland Strait. Prey availability, measured as the biomass of six different prey taxa, did not vary significantly between PCAP and CJ (nested permutational MANOVA, F_1,83 = 2.43, p = 0.087); however, prey availability did vary among transects nested in site (F_6,83 = 2.13, p < 0.001), suggesting there is significant within-site variation. Flock species composition was correlated with the assemblage of available invertebrate prey (Mantel test, r = 0.22, p = 0.047).

Between July 24 and October 9, 2019, we observed foraging rates of 625 individual shorebirds. Foraging strategies, measured by pecking and probing rates, varied with species when controlled for significant (p < 0.05) covariates (Table 3; Fig. 2). The foraging strategies of Short-billed Dowitchers and Semipalmated Plovers differed from all other species, and the foraging strategy of Yellowlegs differed from all but White-rumped Sandpipers (Table 4; Fig. 2; Table S2). Semipalmated Sandpipers, White-rumped Sandpipers, and Least Sandpipers did not use significantly different foraging strategies (Table 4; Fig. 2). SIMPER analysis showed that these differences in foraging strategy were often largely driven by variation in pecking rates (Table S3). Semipalmated Plovers mostly pecked, but had a low mean pecking rate (11 actions per min), while Short-billed Dowitchers mostly probed with the highest mean foraging rate of all species (71 actions per min) (Fig. 2; Table S3). The Calidrid sandpipers used a combination pecking and probing strategy, but all had higher mean pecking rates than probing rates (Fig. 2; Table S3). Yellowlegs also used a combination of pecking and probing, but their mean pecking and probing rates were similar and relatively low (Fig. 2; Table S3). Dispersion of pecking and probing rates varied significantly among species (Permutest, F_6,619 = 5.94, p = 0.001), and pairwise comparisons revealed that this was driven by Semipalmated Plovers and Yellowlegs, which had foraging niche breadths significantly different from other species (Table 5). Although not a significant covariate in the model of foraging rates among species (Table 3), sediment water content was positively correlated with probing rate (Pearson's correlation; r₍₆₂₃₎ = 0.33, p < 0.001) and negatively correlated with pecking rate (Pearson's correlation; r₍₆₂₃₎ = −0.09, p = 0.03).

Blood plasma signatures for δ¹⁵N and δ¹³C, reflecting dietary differences, varied significantly among shorebird species (PERMANOVA, F_3,80 = 13.65, p = 0.001). However, these differences were driven primarily by differences in dispersion among species (Permutest, F_3,80 = 7.37, p = 0.001). Thus, while there was substantial overlap in dietary niche for all species (Fig. 3), diet niche breadth differed among species (Table 6). Semipalmated Plovers and Short-billed Dowitchers had the broadest niches, followed by Semipalmated Sandpipers and finally White-rumped Sandpipers with the narrowest diet niche (Fig. 3).

Foraging microhabitat is an important dimension for segregation among coexisting species (Burger et al. 1977; Pöysä 1983; Davis and Smith 2001; Kent and Sherry 2020) as habitat influences the type and availability of prey (VanDusen et al. 2012) as well as the efficacy of foraging behaviours (Finn et al. 2008). The most prominent foraging microhabitats on the tidal flats at PCAP and CJ are tide pools, which created clear spatial boundaries between species that are and are not tall enough to forage in standing water (Burger et al. 1977). Although tide pools in the Northumberland Strait often attracted multispecies flocks of shorebirds, this size division of available habitat was evident. Short-billed Dowitchers and Yellowlegs could access prey in deeper water due to their greater bill and tarsal lengths (Baker 1979) and were more abundant in tide pools than the smaller species. Yellowlegs took advantage of their broad range of exploitable water depths by foraging both in and out of tide pools, whereas Short-billed Dowitchers almost exclusively foraged in tide pools. This is likely due to softer substrates in tide pools (VanDusen et al. 2012), which make prey more accessible for tactile foragers such as the Short-billed Dowitcher (Mouritsen and Jensen 1992; Finn et al. 2008). Tide pools may be especially attractive for tactile foragers at coastal sites as sediments are often sandy and dense (Davis 2019). Standing water also prolongs surface activity of invertebrates (VanDusen et al. 2012), which may explain why all shorebirds in this study used tide pools to a certain extent. Semipalmated Sandpipers and Semipalmated Plovers were too small to forage in deep water (Novcic 2016) and were clearly more abundant outside of tide pools. However, even they were attracted to the edges of tide pools where the water was shallow, perhaps because of increased invertebrate surface activity (VanDunsen et al. 2012). Least and White-rumped Sandpipers were not abundant enough for analysis of their microhabitat preference; however, given their size, it is likely that they foraged more frequently outside of tide pools like Semipalmated Sandpipers and Semipalmated Plovers.

Spatial segregation between wet and dry microhabitats has been observed at other migratory stopover areas (e.g., Davis and Smith 2001) but is less prominent at sites with limited habitat diversity (Novcic 2016). Therefore, heterogeneity of structural habitat, such as the presence and absence of tide pools, facilitates spatial partitioning of different shorebird species into microhabitat niches. We found evidence of significant spatial heterogeneity of prey availability within coastal sites, and there is a correlation between species composition of foraging flocks and the available prey community where they foraged. As such, if shorebird species target different prey items, prey heterogeneity may contribute to spatial segregation at coastal sites. However, given that some species that overlap in foraging microhabitat had significantly different diet niches (e.g., Semipalmated Sandpipers and Semipalmated Plovers), there is little evidence to suggest that shorebirds target microhabitats within a site based on prey availability or that spatial segregation is driving variation in diet niches. Morphology and foraging behaviour are more likely facilitators of diet variation, as has been suggested for similar species at different stopover sites (Davis and Smith 2001).

Foraging behaviour is a widely studied niche dimension in birds (e.g., Baker and Baker 1973; Davis and Smith 2001; Novcic 2014, 2016; Choi et al. 2017). However, unless prey are distributed in a way that restricts their availability to particular foraging behaviours, it may not reflect prey taxon-level resource partitioning, as large behavioural differences often translate to only small differences in diet (Kent and Sherry 2020). The main difference between pecking and probing foraging behaviours is the depth of consumed prey (Dit Durell 2000), which largely depends on bill morphology, as birds with longer bills can access a wider range of probing depths. Therefore, shorebird species foraging at different sediment depths target spatially segregated prey items, though the prey taxa in their diets may overlap unless these taxa are segregated by depth. For example, gastropod species typically dwell on or near the sediment surface (Huxham et al. 1995; Chandrashkara and Frid 1998), while other invertebrate taxa, including bivalves and polychaete worms, can be found at a range of sediment depths (Henriksen et al. 1983; Zwarts and Wanink 1989; Touhami et al. 2018). Within the taxonomic levels to which we classified burrowing invertebrates for diet niche models in this study, individuals can be found at a wide range of depths, sometimes spanning 0 to over 10 cm (Zwarts and Wanink 1989; Davey 1994). Larger prey also tend to burrow deeper in muddy or sandy intertidal sites (Zwarts and Wanink 1989; Coulthard and Hamilton 2011; Touhami et al. 2018). Thus, long-billed shorebird species, which in our study were larger than the short-billed plovers and small Calidrids (and therefore could consume larger prey items), may have had access to prey items that their smaller competitors could not have consumed. Variation in foraging behaviour therefore reveals a potential for niche partitioning via foraging depth that could not be detected by diet examination alone. As such, it is important to pair behavioural observations with diet analyses when attempting to explain resource partitioning.

We found significant differences in the foraging behaviours and diets of shorebird species during staging in the Northumberland Strait. Short-billed Dowitchers had very high probing rates and low pecking rates, resulting in little behavioural overlap with other species in this study. Long bills, such as that of the Short-billed Dowitcher, are adapted for probing deep into the sediment (Barbosa and Moreno 1999), and thus, pecking behaviours are rarely observed (Baker and Baker 1973; Novcic 2016). This specialized foraging behaviour has been observed in Dowitchers across their migratory range (Baker 1979; Davis and Smith 2001; Novcic 2016), and likely is necessary to reach prey buried in the sediment under standing water. At our site, this behaviour would have provided these birds with access to prey that most other species could not reach. Analyses of plasma isotopes suggested that Short-billed Dowitchers consumed a broad diet, and their lower range of δ¹⁵N values suggests they consumed prey from lower trophic levels than the other species. The diet niche of Short-billed Dowitchers had some overlap with other species; however, their behavioural specialization and limited foraging microhabitat likely result in a realized niche that has little competitive overlap with the other shorebird species on the Northumberland Strait.

Semipalmated Plovers also had unique foraging behaviour. They pecked more than probed, and overall, their foraging rates were low compared with other species. Semipalmated Plovers are visual foragers that spend more time moving and looking for prey items than making foraging attempts (Nol 1986; Ouellette 2021), which results in lower foraging rates than species that peck or probe steadily while walking (Baker and Baker 1973). MacKellar (2018) found that Semipalmated Plovers at similar coastal sites in northeastern New Brunswick had lower foraging rates than Semipalmated Sandpipers feeding in the same area. Given their short bills and pecking behaviour, the dietary niche of Semipalmated Plovers is likely limited to invertebrates dwelling on or near the sediment surface (Dit Durell 2000). Rose et al. (2016) found that Semipalmated Plovers foraged opportunistically and consumed a broad range of invertebrate prey items across various foraging sites on the nonbreeding grounds. This dietary opportunism was reflected in a broad isotopic niche and likely balances their limited range of accessible prey depth. Semipalmated Plovers likely experienced substantive competitive overlap only with the Calidrid sandpipers in the Northumberland Strait due to their similarity in foraging microhabitat. Further, Semipalmated Plovers at Petit-Cap concentrate feeding during daylight hours, while sandpipers continue to forage nocturnally (Ouellette 2021), suggesting an additional level of temporal segregation that would reduce resource competition among these species. As such, Semipalmated Plovers’ broad diets and narrow range of foraging behaviours combined with diurnal foraging trends may allow them to coexist with sandpipers to the extent observed in the Northumberland Strait.

Unlike Short-billed Dowitchers and Semipalmated Plovers, Yellowlegs pecked and probed at similar rates while foraging. This resulted in foraging behaviour that differed from all but the White-rumped Sandpiper. Yellowlegs are tall shorebirds and often forage in standing water (Danufsky and Colwell 2003). Yellowlegs in the Northumberland Strait foraged both in and out of tide pools. Sediment coarseness affected pecking and probing rates of foraging shorebirds, with coarser sediment inhibiting foraging efficiency, as has been seen in other studies (Danufsky and Colwell 2003). These coarser sediments tended to be drier, while wetter and more penetrable sediments supported more probing behaviours. Therefore, Yellowlegs may adjust their foraging behaviour for the sediment conditions they encounter in different microhabitats, resulting in similar mean pecking and probing rates. Yellowlegs’ broad habitat use likely resulted in competitive overlap with both Short-billed Dowitchers and the smaller sandpipers and plovers. We did not estimate diet of Yellowlegs; however, Andrei et al. (2009) found that Yellowlegs consumed broad diets at stopover sites in the southern United States. Niche complementarity predicts that Yellowlegs should exhibit niche specialization on some dimension to compensate for competitive overlap (Schoener 1974). However, it is possible that Yellowlegs forego foraging specialization and instead use a generalist staging strategy to exploit the full range of available resources at coastal sites.

Semipalmated Sandpipers, Least Sandpipers, and White-rumped Sandpipers did not have significantly different foraging behaviour. These species are congeners and are morphologically similar (Thomas et al. 2004), which likely plays a part in their similar behaviour (Barbosa and Moreno 1999; Dit Durell 2000; Norazlimi and Ramli 2015). All three sandpiper species used both pecking and probing behaviours but had higher pecking rates than probing rates. Resource competition is likely high among the sandpiper species; however, differences in diet breadth may have compensated for spatial and behavioural overlap. The diet of White-rumped Sandpipers was considerably narrower than that of Semipalmated Sandpipers. On their wintering grounds, White-rumped Sandpipers consume high proportions of the most abundant prey items (de los Angeles Hernandez and Bala 2007), which suggests they modify their diet to maximize foraging efficiency at different sites across their migratory range. The diet niche of Semipalmated Sandpipers in our study was also narrow compared with niches of Short-billed Dowitchers and Semipalmated Plovers; however, Semipalmated Sandpipers are considered generalist foragers (Gerwing et al. 2016) and have been found to opportunistically target the most available prey when more favourable prey items are unabundant (MacDonald et al. 2012). The heterogeneity of habitat and prey assemblages in the Northumberland Strait likely allows Semipalmated and White-rumped Sandpipers to opportunistically target the specific prey taxa that they can consume most efficiently, thus limiting dietary overlap with Dowitchers and Plovers.