Habitat variety and behaviour of the Sanderling ( Calidris alba ) in an urban neotropical coastal wetland

Sanderling Calidris alba is a migratory coastal bird species whose populations are affected by climatic and anthropogenic issues. Additionally, its behavioural patterns depend on habitat variety. The aim of this study was to evaluate the variation in the frequency and duration of C. alba behaviour in relation to habitat variety in a Peruvian coastal wetland. Based on types of behaviour recorded in a previous study of C. alba during August and September 2019 at low, mid and high tide periods in three habitats of the Coastal Wetland Poza La Arenilla (HCPA) (La Punta, Callao, Peru), differences and diversity in the frequencies and durations of such behaviour between habitats were assessed according to tide condition. Additionally, a correlation analysis between the duration and frequency of behaviour for each habitat according to tide condition was conducted. General differences in behaviour (frequency and duration) between habitats were found, although there was a high similarity of behaviour between areas particularly when mid-and high tide levels were present. The greatest diversity of behaviour, both in frequency and duration, was observed generally during mid-tide conditions in rocky habitats. In most cases, there was a high correlation between the frequency and duration of behaviour for habitats according to tide condition. It is concluded that mid-tide conditions are associated with greater availability of soft substrates, favouring a wide variety of types of behaviour for C. alba , including behaviour related to foraging and locomotion.


Introduction
The Sanderling Calidris alba (Pallas, 1764) (Fig. 1) is a boreal migratory bird species that prefers habitats in the surf zone and near the breakwater line (Cotillo et al. 2018;Mazzochi et al. 2021).Calidris alba is a small shorebird (between 19 and 21 cm in length) belonging to the Scolopacidae family, characterised by abundant and seasonally fluctuating populations (Scherer and Petry 2012;Podestá and Barona 2021) and facing issues related to population decline, habitat loss and pressures due to sea-level rise as a consequence of climate change (Payne 2010;Galbraith et al. 2014).
The behaviour of shorebirds like C. alba can be directly influenced by tidal cycles, food availability and access to prey, resulting in local movements to access suitable roosting areas during high tide and feeding sites during low and mid-tide periods (Martínez-Curci and Petracci 2016; Quiñonez and Hernandez 2017;Houpt et al. 2020), where they can find a wide variety of marine invertebrates crucial for their diet (Ferrari et al. 2012;Tallei et al. 2021).Additionally, birds may exhibit different feeding or resting behaviour in different habitats because of the influence of physical factors, such as temperature (Myers et al. 1985;Londe et al. 2021).
Regarding food availability, Grond et al. (2015) found that sanderlings needed less time for foraging in habitats with great availability of soft body preys, such as polychaetes, while they needed more time for foraging in habitats with great availability of hard body preys, such as bivalves.Lourenço et al. (2015) found similar food availability patterns associated with foraging time budgets, these being sandy shores habitats with greater diversity of adequate prey for sanderlings.Sanderling behaviour is expected to differ between sandy or rocky habitats and between tide levels, because it means different physical conditions in which birds could feed, rest or walk (Grond et al. 2015;Lourenço et al. 2015).Fonseca et al. (2017) found that the total time spent by shore birds in foraging behaviour decreases when neap tides occur, mainly because of the reduction in the intertidal zones available.However, foraging behaviour of sanderlings appear to increase in intertidal zones (Mazzochi et al. 2021).Podestá et al. (2022) characterised the behaviour of C. alba in the Poza La Arenilla wetland (HCPA), La Punta, Callao, Peru, in great detail.Currently, HCPA is home to 98 bird species and is amongst the three most significant wetlands in Callao (the other two being the Ventanilla Wetlands Conservation Area and the "El Mirador" Lagoon in Ventanilla).One of its standout features of HCPA is its proximity to the urban environment, separated by a promenade for tourism and birdwatching (Podestá and Barona 2021;Podestá et al. 2021).It also serves as a mandatory stopover point for migratory birds during their annual migration, with the presence of the species C. alba, a sandpiper from the Scolopacidae family, being particularly notable during the spring months (Podestá et al. 2017;García-Olaechea et al. 2018).
The objective of this study was to evaluate the variation in the frequency and duration of C. alba behaviour in relation to habitat variety in a HCPA.We aimed to answer the following questions: Do behaviour patterns change with tide level and habitat use?Do frequency and duration of behaviour show correlation between them?This study presents the results of observations of C. alba behaviour in two types of habitats and tides during the months of August and September 2019.

Study area
The Coastal Wetland "Poza La Arenilla" (HCPA) is located in the district of La Punta, Callao, on the coastal strip between 10°04'00"S and 10°04'30"S latitude and 78°10'30"W longitude.Average annual temperature is 18.75 °C and average annual precipitation is 0.6 mm.The average altitude of the wetland is 0.5 m.This wetland was formed because of the construction of two rocky barriers between 1965 and 1967 to protect the southern zone of the La Punta District from abnormal sea surges (Troll 2000).Additionally, due to those strong abnormal sea surges, a 50-m wide inlet that connected these rocky barriers was filled, forming the Isla de Guilligan (DHM 1987) and enclosing this area with calm and semi-stagnant waters, which is nowadays known as the Coastal Wetland "Poza de la Arenilla" (HCPA) (Troll 2000).This study takes into account the zoning of the wetland, based on the use of the area by birds, as proposed by Cotillo et al. (2018).
HCPA has eleven zones or habitats (Fig. 2).Three zones or habitats were assessed for the current study: Cobbles from Boulder Zone (BZ), Sand Shore 1 (SS1) and Sand Shore 2 (SS2).These three areas are the primary zones used by C. alba for activities, such as feeding and resting within the wetland.BZ has been described as an area with relatively shallow waters (Cotillo et al. 2018).Initially, SS1 and SS2 differ in that the former receives water flows from the West inlet, while the latter borders the Water Mirror 2 (W2), which is a lentic water body receiving water from the East inlet and is the longest of the water bodies.Additionally, it is characterised by a lower water flow compared to the Water Mirror 1 (W1), which is much closer to its respective inlet (Cotillo et al. 2018).

Study species
The Sanderling Calidris alba (Pallas, 1764) (Fig. 1) is a boreal migratory bird species whose preferred habitats are the surf zone and near the breakwater line (Cotillo et al. 2018;Mazzochi et al. 2021).C. alba is a small shorebird belonging to the Scolopacidae family, characterised by abundant and seasonally fluctuating populations (Scherer and Petry 2012;Podestá and Barona 2021).Although this bird species is categorised by the IUCN Red List of Threatened Species (BirdLife International 2016) as Least Concern (LC), it is facing issues related to population decline, habitat loss and pressures due to sea-level rise because of climate change (Payne 2010;Galbraith et al. 2014).

Recording of behaviour
The types of behaviour used for analysis were those documented by Podestá et al. (2022) (See Suppl.material 1), who developed an ethogram for the species C. alba through sampling conducted during the period when this species arrives at the wetland as part of its migratory cycle.This evaluation covered 12 days in the months of August and September 2019, a period in which the species arrives at the wetland, with a total of 670 effective minutes of observation (40,200 seconds) and the recording of 58 different types of behaviour belonging to eight behaviour categories (Podestá et al. 2022) (See Suppl.material 1).Observations were made by one observer, who was accompanied by two field assistants.
The sampling was conducted during periods of low, mid and high tide.Observation was carried out by tracking a focal individual.The focal sampling technique was employed, which involves recording the occurrence (frequency) and duration of behaviour of a randomly selected single individual from a group of organisms (Suen and Ary 2014).However, when the focal individual interrupted its behaviour by leaving the area, the observer shifted immediately to another individual from the population, fulfilling the timeframe specified by the observer, without interrupting the sampling.
The recording of the activity of each focal subject was supported by 10-minute videos, which were later analysed for better identification of bird behaviour.Additionally, the observation time for each focal subject was 10 minutes.

Statistical analysis
The Kruskal-Wallis test was applied to verify statistically significant overall differences in behaviour both amongst the three evaluated zones (habitats) and amongst tides (low, mid and high).
Subsequently, the following secondary analyses were conducted: 1) a similarity analysis for both the frequency and duration of behaviour between zones and tides using the Bray-Curtis similarity index, which is useful for assessing quantitative data (Birkl et al. 2018), 2) the calculation of the equivalent number of types of behaviour (frequency and duration) ( 1 D) based on the Shannon-Wiener entropy index (Ligon et al. 2018) to determine which of the three habitats exhibited greater diversity of behaviour in terms of their frequency and duration and 3) a Spearman correlation analysis between the duration and frequency of behaviour for each habitat, based on tide conditions.

Results
The greatest diversity of behaviour, both in frequency and duration, was observed generally during mid-tide conditions in rocky habitats.In most cases, there was a high correlation between the frequency and duration of behaviour for habitats according to tide condition.
The non-parametric Kruskal-Wallis test indicated that there were overall differences in the frequency (P < 0.0001) and duration (P < 0.0001) of behaviour between habitats based on tides.The similarity analysis conducted using the Bray-Curtis index revealed the existence of clusters with similarity greater than 50% for both similarity of behaviour frequency (Fig. 3A) and behaviour duration (Fig. 3B).In the first case, three paired comparisons showed similarities greater than 50% (SS1-mid versus SS2-high: 80%; BZ-mid versus SS1-mid: 54%; BZ-mid versus SS2-high: 52%), i.e. habitats based on tide condition with more than half of shared types of behaviour (Fig. 4A).In the second case, three paired comparisons also showed similarities greater than 50% (SS1-mid versus SS2-high: 68%; SS2-mid versus SS2-high: 52%; SS1-mid versus SS2-mid: 48%) (Fig. 4B).No different types of behaviour were recorded for habitat BZ and SS1 during high tide and for habitat SS2 during low tide.
Regarding the diversity analysis, the equivalent number of types of behaviour based on the Shannon entropy index revealed that, concerning the frequency of behaviour, the highest equivalent number was for the BZ habitat at mid-tide ( 1 D=26.8),while the lowest value was obtained by SS2 at mid-tide ( 1 D=7).Concerning the duration of behaviour, the highest equivalent number was for the BZ habitat at mid-tide ( 1 D=25.3) and the lowest value was for BZ at low tide ( 1 D=3.9) (Table 1).Additionally, Figs 5, 6 display the frequency and duration of the main behaviour and behaviour categories in each habitat according to each tide condition.
In most cases, a high correlation was observed between the frequency and duration of behaviour for habitats according to tidal condition (ρ Spearman > 0.82; P < 0.05), with the only exception being habitat SS2 under mid-tidal condition (ρ Spearman = 0.36; P > 0.05).Table 1.Types of behaviour in each habitat according to each type of tide.Habitats BZ at high tide, SS1 at high tide and SS2 at low tide have been omitted since no behaviour was recorded in these habitats under these tidal conditions.The names of the types of behaviour and their abbreviations are in the Suppl.material 1.

Discussion
Overall differences in the frequency and duration of behaviour between habitats based on tides were observed.Additionally, the greatest diversity of behaviour was observed generally during mid-tide conditions in rocky habitats.This outcome is complementary with what Grond et al. (2015) and Lourenço et al. (2015) found, i.e. that the greater availability of prey occurs in sandy shores.
In HCPA, the greater diversity of behaviour occurred at the boulder zone (BZ) during mid-tide, although the most common behaviour in that zone was resting, not foraging.On the other hand, sanderlings showed foraging and locomotion behaviour as the most common behaviour in sandy shores during low tides, where there is a high invertebrate biomass (J.Podestá, personal communication, 07 July 2023).
No types of behaviour were recorded for habitat BZ during high tide.Podestá et al. (2017) attribute the absence of shorebirds (such as C. alba) during the summer season in several key areas of HCPA to the occurrence of anthropogenic activities.Additionally, high tide levels contribute to several species having difficulties in finding specific areas to use as roosting sites (resting areas for birds) (Senner and Howe 1984;Becerra and Ferrari 2012;Giner and Pérez-Emán 2015;Senner et al. 2017), which is relevant information since the "Resting" behaviour is one of the main types of behaviour exhibited by C. alba in the HCPA.Shores are areas with soft substrates and it is precisely in these areas where the highest activity of the species C. alba was observed, especially during foraging (Fonseca et al. 2017;Mazzochi et al. 2021).Yates et al. (1993) and Scheiffarth et al. (1996) attribute this behaviour to the greater accessibility and availability of prey in the wet substrate, making it easier for these birds to probe into the substrate.However, similar to habitat BZ, no types of behaviour were recorded in SS1 during high tide.Hernández et al. (2012) and Houpt et al. (2020) relate the movement of shorebirds to the accessibility of prey during cyclical tide periods, with the birds moving short distances to other areas with similar characteristics that provide their basic needs (foraging).This could also be due to the preferred depth for each species, which limits their presence to certain specific habitats (Vargas-Fonseca 2014).Levey (1988) and Daniel et al. (2019) consider that ecological interactions, such as inter and intraspecific competition, determine the dispersal of birds amongst different wetlands.This is important for planning conservation actions in different areas, taking into account the movement of C. alba, as well as its presence and absence in specific locations, particularly because various ecological-evolutionary processes are related to the development of behaviour linked to interspecific interactions (Lawrence et al. 2012).
On the other hand, greater similarity was found when considering both the frequencies and duration of behaviour between SS1 at mid-tide and SS2 at high tide (Figs 3, 4).This could be related to the location of these two areas within the HCPA, which are connected to spaces where the water flow constantly varies with the tide movements; for instance, water mirror 1 adjacent to SS1 and water mirror 2 to SS2.This would allow C. alba to engage in nearly the same activities in both places.The similarity in behaviour between SS1 at mid-tide and BZ at mid-tide, as well as between BZ at mid-tide and SS2 at high tide, may be due to their proximity, as BZ and SS1 are adjacent habitats affected by tides coming from the West inlet, while BZ and SS2 are also relatively close to each other (Figs 3, 4).
The greatest diversity in both the frequency and duration of all types of behaviour for habitat BZ was observed at mid-tide.For SS1, this greater diversity occurred at low tide and for SS2, it was at high tide (Table 1).According to previous studies, the sandpiper C. alba is present in HCPA for most of the time in the mid-littoral zone, near the breakwater adjacent to BZ.This area is more frequently used by this species, as well as areas near the shore (Podestá et al. 2017;Cotillo et al. 2018).However, there is also a high diversity of behaviour associated with the shores during both low and high tide also having been recorded.This is related to the species' preference for locations with continuous tidal movement, due to the presence of muddy and sandy-mud areas where it can easily find food (Mazzochi et al. 2021;Podestá et al. 2022).
Regarding the BZ habitat, the categories of resting, locomotion and grooming are the most frequent, although the locomotion category has the longest duration, covering more than 65% of the total time (Figs 5,6).Types of behaviour included in the foraging category, on the other hand, were only observed during mid-tide, which can be explained by the greater availability of food.This is because both low and high tide, with their rocky nature and water levels, do not favour food availability (Cotillo et al. 2018).Carmona et al. (2003) pointed out that physical factors, such as substrate moisture and texture, are associated with prey acquisition due to tidal inundation in two ways: 1) easier substrate penetration and 2) increased invertebrate activity, making them more susceptible to capture by predators, especially shorebirds.These prey items may be more available during periods of low and mid-tide when the substrate texture is wet and soft, benefitting various shorebirds, especially C. alba, which is one of the smallest and has a short beak, requiring a soft substrate for penetration.Additionally, Beerens et al. (2011) found that different bird species with different feeding strategies were influenced by water physical features, such as water depth.
Furthermore, Ramli and Norazlimi (2016) observed differences in shorebird behaviour in relation to tidal cycles, as they influence population abundance and behaviour (greater abundance and more foraging during low tide periods).Similarly, Manrique and Williams (2005) observed differences in relation to pro-longed diving behaviour of birds and tides (more diving activity during low tide and less surface resting time during high tide).It is also important to note that high and low tide cycles are extreme points, resulting in prolonged periods of mid-tide, precisely when most activities carried out by C. alba are favoured.
In SS1 and SS2, the most frequent types of behaviour were those classified under the categories of foraging and locomotion, although during low tide in SS1, foraging decreases in frequency and behaviour related to the grooming category occur more frequently.The high duration of the behaviour 'long range flight' (LRF) in SS1 during low tide is also significant, despite the low frequency of this behaviour (number of occurrences) in that habitat under such tidal conditions.This is consistent with the study by Burger et al. (2018), which showed that the local abundance of C. alba tends to decrease during low tide hours, so the increase in LFR is probably due to individuals searching and moving to other habitats with better conditions for resting or feeding.Bird et al. (2019) reported that, as the tide drops, invertebrates accustomed to intertidal zones that inhabit muddy or sandy substrates are exposed to drastic temperature changes, moving to greater depths of the substrate, thus compensating for these changes to avoid desiccation.These behavioural adaptations may be reflected in the results of the present study, which identified a lower frequency and duration of behaviour belonging to the foraging category by C. alba in SS1 during low tide.
The duration and frequency of behaviour are highly correlated and, therefore, evaluating either of them for C. alba provides very similar information about the behaviour pattern for the studied species.However, the low association observed between the duration and frequency of behaviour for habitat SS2 in midtide conditions occurs because the duration of very infrequent behaviour like 'rest with beak between wings' and LRF, with only a single occurrence, were up to seven times longer than other more frequent types of behaviour, such as 'pecking', for that habitat and tidal condition.This could be due to the fact that habitat SS2 is one of the most diverse habitats in HCPA in terms of sandpipers' community (Cotillo et al. 2018) and, thus, greater competition for space may influence the higher occurrence and longer duration of non-feeding-related behaviour.

Conclusion
C. alba prefers areas with soft and moist substrate, where there is little competition for space with larger species and where food is more readily available.These features mainly occur during mid-tide, when there is greater diversity of behaviour and higher frequency/duration of behaviour related to foraging and locomotion.

Figure 3 .
Figure 3. Cluster plot based on Bray-Curtis similarity A frequency of similarity of behaviour B duration of similarity of behaviour.

Figure 4 .
Figure 4. Map depicting the three combinations (habitat and tide) with the highest Bray Curtis similarity in terms of behaviour frequency (A) and duration (B).Modified from Cotillo et al. (2018).

Figure 5 .
Figure 5.Primary types of behaviour recorded in each habitat under different tide conditions.The top five types of behaviour by frequency (orange bars) and duration (blue bars) are displayed (out of 58 recorded types of behaviour), while the remaining types of behaviour are grouped under the category "other".Behaviour nomenclature is provided in Suppl.material 1.

Figure 6 .
Figure 6.Categories of behaviour recorded in each habitat under different tide conditions A frequency of behaviour categories B duration of behaviour categories.