Temporal and spatial variation of waterbirds at Sayula Lagoon, Jalisco, Mexico: a five-year winter season study

Study Extent

Waterbird surveys were conducted monthly over five winters, from October to March 2004-2007 and from October to March 2009-2011, as the migratory species are present in the study area in these months.

Sampling

Thirty (1 ha) permanent plots were randomly stratified using a numbered grid for each zone type. Each plot was located at a minimum distance of 500 m from each other in order to avoid double counts of the same individuals, following Ojasti and Dallmeier (2000). Plots were delimited with red and yellow sticks to allow their rapid location. Observations were made in the eight hours after sunrise. We recorded all the bird species seen and total abundance was also recorded. The block method (Howes and Bakewell, 1989) was used to estimate numbers whenever large flocks (> 300 birds) were present. The seasonal status of species was contrasted with those published in Howell and Webb (2001). The category of species risk was assigned using the Mexican Law (NOM-059-SEMARNAT-2010) and the IUCN Red List of Threatened Species. The on-line taxonomic Check-list of North American birds (AOU, 2015) was used. All plots were classified as one of the five zones according to Colwell and Taft (2000), with modifications as follows: Deep Zone (DEEZ > 100 cm), Aquatic Zone (AQUZ > 20 y < 100 cm), Shallow Zone (SHAZ < 20 cm), Muddy Zone (zone with wet soil and some small waterlogging) (MUDZ) and Sandy Zone (zone with dry soil) (SANZ). This classification was possible because during the observation period (October-March) in each study year, the water level remained the same. We recorded the area or areas in which each observed individual bird it was observed

Method steps

1. Temporal changes were analyzed by comparing seasons and months, and spatial changes were analyzed by comparing the zones in the study area. We also studied attributes of community structure. Richness (S), was estimated through species accumulation curves using EstimateS v, 9.1.0, (Colwell, 2009). Using abundance (numbers of individuals) data and the same software, rarefaction curves were performed to compare richness between zones, months and sampled seasons. In the latter analysis, we compared the richness mean values and their 84% confidence intervals at p=0.05 (MacGregor-Fors and Payton, 2013). Species relative abundance was estimated according to the following categories: abundant (90-100 %), common (65-89 %), moderately common (31-64 %), uncommon (10-30 %), and rare (1-9 %). These percentages were obtained for each species as the number of individuals of a species divided by the total number of individuals considering all species and multiplying by 100 (Pettingill, 1969). Relative frequency was estimated to determine species representativeness over time and the following categories were assigned: very frequent (0.76-1), frequent (0.51-0.75), moderately frequent (0.26-0.50), and sporadic (0- 0.25). This estimate refers only to the number of plots containing a given species divided by the total number of plots, which is not redundant with abundance because it does not refer to individuals (Krebs, 1985). In order to ensure non-redundant data, a fixed time period was assigned (10 minutes) for observation and the distance between plots was longer than 500 m. An abundance data matrix - estimated with the Bray-Curtis’ index using the 4th root data transformation to reduce the contribution of abundant species - was implemented to perform the following analyses using PRIMER 6 (Clarke and Gorley, 2005). A nonmetric multidimensional scaling (NMDS) complemented with hierarchical cluster analysis (Bray-Curtis´ index using average group linkage methods, Clarke and Warwick, 2001) was implemented to compare species composition between zones, months, and seasons. To assess significant differences between groups of samples, a one-way non-parametric similarity analysis was performed (ANOSIM) using 10,000 permutations (Clarke and Gorley, 2005). We also used a one-way similarity percentage method (SIMPER) to identify the most representative species in each zone and to determine the percentage of similarity between zones. Species were selected considering those contributing with 90% of the observed similarity in this study. Functional groups for species in the lagoon were determined according to Escofet et al. (1988) and Terres (1991) on the basis of their foraging strategy (shorebirds, ducks, small grebes, jacanas and large wading birds) and dietary strategy (herbivore, piscivore). Our categories are coarse, but match the generality of the ecological questions we addressed.

Waterbirds at Sayula Lagoon, Jalisco, Mexico

1. Aves
   rank: class
2. Anseriformes
   rank: order
3. Anatidae
   rank: family
4. Ardeidae
   rank: family
5. Charadriiformes
   rank: order
6. Charadriidae
   rank: family
7. Jacanidae
   rank: family
8. Laridae
   rank: family
9. Pelecaniformes
   rank: order
10. Pelecanidae
    rank: family
11. Suliformes
    rank: order
12. Phalacrocoracidae
    rank: family
13. Podecipediformes
    rank: order
14. Podicipedidae
    rank: family
15. Gruiformes
    rank: order
16. Rallidae
    rank: family
17. Recurvirostridae
    rank: family
18. Scolopacidae
    rank: family
19. Threskiornithidae
    rank: family

Sayula Lagoon in the state of Jalisco in west central Mexico