Wetlands provide several ecosystem benefits and are important habitats for waterbirds (Mitsch and Gosselink 2000). Waterbirds are a highly diverse group, exhibiting large differences in their habitat requirements (Fredrickson and Reid 1988). Nevertheless, most of these species select wetlands in relation to the location of their feeding grounds or reproduction areas (Raeside et al. 2007).

Chilean wetlands are complex and diverse systems (Marín et al. 2006; Schlatter and Sielfeld 2006; González et al. 2011). The Carlos Anwandter Nature Sanctuary (Sanctuary), a Ramsar wetland in Southern Chile, is considered a priority site for waterbird conservation (Risk et al. 2010). This wetland hosts about 119 bird species, primarily year-round residents (Schlatter and Sielfeld 2006). Among these species, the Black-necked Swan (Cygnus melancoryphus) is a “flagship species” in the Southern Neotropical area because it is an indicator of the quality of the wetlands (Corti and Schlatter 2002).

During the last decade, the Sanctuary has undergone degradation due to human activity in the area (Universidad Austral de Chile 2004, 2005; Jaramillo et al. 2007). Nevertheless, this Sanctuary is considered a source of Black-necked Swans (swans) for South America (Schlatter et al. 2002). The Black-necked Swan population here has shown strong yearly fluctuations (Schlatter et al. 2002; Lagos et al. 2008). Population size decreases occurred in 1997 and 2004; the former mainly driven by natural perturbations (El Niño Southern Oscillation) (Schlatter et al. 2002), while the later was attributed to changes in the abundance of the aquatic plant Brazilian waterweed (Egeria densa), which is the Black-necked Swan's major food resource in the area (Corti and Schlatter 2002). The biomass of this plant was severely reduced during 2004 (Di Marzio and McInnes 2005) and both its disappearance and the population decrease of Black-necked Swans were attributed to human activities (Universidad de Austral Chile 2004, 2005; Sepúlveda 2011).

Although there is much scientific interest in the consequences of the environmental changes in the Sanctuary, most of the studies have been focused on the Black-necked Swan population inside the protected area, but wetlands adjacent to the Sanctuary have not be considered. Further, there is very little information about foraging and behavioral responses of swans to the environmental changes occurring in this wetland (but see Schlatter et al. 2002). This information is crucial for Black-necked Swan conservation in the area. Based on this information, our goals were to describe: 1) spatiotemporal variation in Black-necked Swan population abundance and distribution; and 2) Black-necked Swan foraging and reproductive ecology.

Methods

Black-necked Swan Surveys

Swan counts were conducted seasonally from May 2006 to January 2010 between 08:00 hr and 19:00 hr using binoculars. Surveys were done by boat and from fixed-point locations on the ground. On each survey, two observers recorded the number of swans, while a third observer recorded the location coordinates of groups of swans using a global positioning system (GPS). The geographical data were mapped using the software R (R Development Core Team 2011). We obtained additional data for swans in this wetland for the period between 2000 and 2005 (Chile's National Forest Corporation 2010a). As they used the same counting method, we included data from 2000 to 2010 in our analyses.

To assess the distribution of the foraging areas in the wetland, we recorded Black-necked Swans on their feeding grounds using a GPS in the same manner as above (Erwin 2005). To estimate Black-necked Swan density, we standardized the abundance of swans by the area of each zone (individual/km²); (48,770 km² inside the Sanctuary and 15,250 km² outside the Sanctuary). We collected samples of the aquatic plants that occurred in these foraging areas using a hook. Plants were identified to the lowest taxonomic level possible using identification guides (Ramírez et al. 1982; Muñoz-Pedreros 2003). We emphasize that the intent of this sampling was simply to document that Black-necked Swans were indeed feeding on plants other than Brazilian waterweed, not to provide a comprehensive analysis of plant cover or Black-necked Swan feeding preferences. Additionally, we determined the size and frequency of foraging groups in the Sanctuary during the autumn (late March to late June) and spring (late September to late December) from 2004 to 2009.

Figure 1.

Map of the study area showing the “wetland complex.” Black dots represent Black-necked Swan counting locations within the study area. The gray shaded portions of the map represent the areas within the Sanctuary (continuous line) and outside the Sanctuary (dotted line).

The distribution of Black-necked Swan reproductive areas was recorded and analyzed using the same methods as the feeding grounds. We characterized reproductive timing, occurrence of Black-necked Swan pairs, and nests and chicks between 2000 and 2010.

Results

Spatial Variation in Black-necked Swan Abundance

Between 2000 and 2004, Black-necked Swans were concentrated in a few sites of relatively high population abundance (> 1,200 individuals) (Fig. 3A). At the same time, most of the areas within the Sanctuary maintained less than 400 individuals of swans (Fig. 3A). From 2005 to 2010, we observed that most areas in the wetland complex maintained a low number of individuals (i.e., < 80); meanwhile, most Black-necked Swans (∼200 individuals) were concentrated in a single area (Fig. 3B). The two sites with the highest Black-necked Swan numbers between 2000 and 2004 supported close to 60% of the population, which was significantly higher than the site containing the next higher abundance of Black-necked Swans (t = 3.41; P < 0.01; Fig. 4A). From 2005 to 2010, the area with the highest number of swans supported on average more than the 42% of the population, which was significantly higher than the site containing the second higher abundance of Black-necked Swans (t = 2.56; P < 0.01; Fig. 4B). Overall, population densities inside the Sanctuary were lower (0.1 individuals/ha) than outside the Sanctuary (0.5 individuals/ha) during 2005–2010 (t = 2.46; P < 0.01).

Figure 2.

Mean number of Black-necked Swans per season (± SE) in the Carlos Anwandter Nature Sanctuary (black circles) and outside the Sanctuary (white circles) during 2000–2010 (Chile's National Forest Corporation 2010b). No data available outside the Sanctuary between 2000 and 2004.

Figure 3.

Histogram of the number of Black-necked Swans in the Carlos Anwandter Nature Sanctuary (inside Sanctuary) during 2000–2004 (A), and in areas inside and outside the Carlos Anwandter Nature Sanctuary during 2005–2010 (B).

Foraging Ecology

We identified five principal feeding grounds for Black-necked Swans in the wetland complex (Fig. 5). Brazilian waterweed was present in some areas inside and outside the Sanctuary but the biomass of other aquatic plants, such as parrot feather water milfoil (Myriophyllum aquaticum) and pondweed (Potamogeton sp.) was greater than that of Brazilian waterweed. Other areas of the wetland showed high cover of the green algae Spirogyra sp. and the floating aquatic Amazon frog bite (Limnobium laevigatum), both co-occurring in San Antonio, whereas floating pennywort (Hydrocotyle ranunculoides) was present in the Cayumapu (Fig. 5). During the summers of 2008 to 2010 (end of December-end of March), Cayumapu showed some cover of Brazilian waterweed, while only two locations (i.e., Pishuinco and Huellelhue) had high percent cover of Brazilian waterweed throughout the entire study period (Fig. 5). However, they were not important as foraging areas for the Black-necked Swans.

Figure 4.

Primary Black-necked Swan concentration sites during 2000–2004 (A) and during 2005–2010 (B) in areas inside and outside the Carlos Anwandter Nature Sanctuary.

Reproductive Ecology

Long-term records of Black-necked Swan reproduction showed that swans started nesting in early June (Fig. 6). Between 2004 and 2006, there were no records of Black-necked Swan reproduction in the wetland. During the last three reproductive seasons (2007–2008, 2008–2009, and 2009–2010), Black-necked Swans showed a pattern of breeding increasingly earlier, evidencing a reproductive phenology similar to that before 2004. The total number of swan nests and chicks recorded between 2000 and 2004 were > 2,000 nests and about 11,000 chicks (Table 1). The nests and chicks of Black-necked Swans were recorded in seven main areas in the Sanctuary (Table 1). During 2004–2006, there were no nests and chicks, whereas from 2006 to 2010 the number of swan nests and chicks increased (Table 1); most of these were found in the Sanctuary.

Figure 5.

Primary Black-necked Swan foraging grounds in areas inside and outside the Carlos Anwandter Nature Sanctuary. Black circles show areas in the Sanctuary, and black stars show the five main feeding locations.

Discussion

Our results showed large spatiotemporal variation of the Black-necked Swan population size and its distribution in the wetland complex of the Cruces River. During the study period, the Black-necked Swan population showed increases in its abundance during the summer-autumn and decreases during the winter-spring. These findings are contrary to those reported previously for Black-necked Swans in this wetland complex (Schlatter et al. 2002; Schlatter 2005). These previous studies showed population increases during the spring-summer and decreases during the autumn-winter, and these changes were related to precipitation levels. The differences between our results and Schlatter et al. (2002) and Schlatter (2005) may be due to differences in the temporal series analyzed by them (1985 to 2000). During this period, there were two population increases (1988–1991 and 1995–1998); these increases were concordant with La Niña phase that can cause desiccation of shallow wetlands and trigger bird migrations from northern to southern Chile (Vilina et al. 2002; Schlatter 2005). Overall, precipitation seasonality influences the hydrology of the wetland, affecting Black-necked Swan population abundance and food availability (Schlatter et al. 2002; Schlatter 2005). In fact, during the winter, high precipitation may also cause a decrease in Black-necked Swan populations in the Sanctuary, whereas the opposite pattern is common during the summer (Schlatter et al. 2002). Black-necked Swan population size was less variable during the spring, and we speculate that the presence of chicks, and consequently their parents, may explain this pattern; this phenomenon occurs from August to March each year and is related to more stable environmental conditions in the wetland (Schlatter 1998).

Figure 6.

Reproductive timing of Black-necked Swans in areas inside and outside the Carlos Anwandter Nature Sanctuary; data collected between 2000 and 2010. Different rectangle colors show distinct reproductive stages: mating pair formation (white rectangles), nesting (light gray rectangles), chicks (dark gray rectangles) and juveniles (black rectangles). Reproductive season 2009–2010 was still developing when this study was finished.

Our spatial analyses showed that a few sites within the wetland host the largest number of Black-necked Swans. This pattern may be explained by the differences in microhabitat conditions within the wetland, such as food availability, shelter or reproductive areas (Brown et al. 1995). Although our study design does not allow us to infer cause-effect relationships, the largest and most stable Black-necked Swan populations were in Cayumapu River, San Antonio, and particularly in Guacamayo during the last 2 years of the study (2009–2010). This suggests that environmental conditions within the wetland have changed and might be responsible for the dynamics of Black-necked Swans.

The reproductive timing of birds may be strongly affected by energetic constraints, suggesting that food availability is the main factor influencing the nesting timing of different populations (Chastel et al. 2003). Black-necked Swans showed no reproduction between 2004 and 2006. We believe this behavior was due to the decrease in their main food resource. After 2006, Black-necked Swans delayed reproductive timing, when Brazilian waterweed was absent from the wetland. This suggests that Black-necked Swans may be using other aquatic plants as food resources (see Ramírez et al. 1982; Corti and Schlatter 2002).

Table 1.

Summary of the number and main locations of Black-necked Swan nests and chicks recorded during the breeding periods between June 2000 and January 2010 in the Carlos Anwandter Nature Sanctuary and outside the Sanctuary in Valdivia, Chile. ND indicates no data available. Data provided by Chile's National Forest Corporation (2010b).

The environmental changes recorded in this wetland have been related to human activities, specifically a pulp mill, which has caused the disappearance of Brazilian waterweed (Jaramillo et al. 2007) and triggered massive Black-necked Swan migration out of the area and mortality due to starvation and high levels of iron and other chemicals in the water (Universidad Austral de Chile 2004, 2005; Mulsow and Grandjean 2006; Sepúlveda 2011). Although the observed patterns showed significant changes in the population dynamics of the Black-necked Swans between 2004 and 2005, we cannot infer cause-effect relationships between Brazilian waterweed disappearance and the decrease in Black-necked Swan numbers. However, the correlation between these two phenomena seems to be strong (Sepúlveda 2011). Overall, the distributional patterns of swans in the Sanctuary and outside this protected area provide useful information about the spatial scale over which Black-necked Swans distribute. This information should be considered in biodiversity conservation and the design of protected areas (see Schlatter et al. 2002).