Factors influencing nest-site selection
Correlations among predictor variables (water depth, vegetation height, offshore distance of the nest and grazing intensity) were first considered to test for multicollinearity. The VIF values were low (<1.3), and models did not show evidence of overdispersion (Φ ≈ 1), and therefore all predictor variables were included in the model without correcting for multicollinearity or overdispersion. The model selection procedure yielded a single best model (ΔAICc <2) with the predictor variables i) water depth, ii) vegetation height, and iii) offshore distance of the nest (Table 1). These variables had a significant influence on nest-site selection: water depth (p =0.005), offshore distance of the nest (p =0.037), and vegetation height (p =0.035) (Table 2).
A total of 30 nests out 63 active nests and an equal number of randomly selected nest-sites, were measured. These nests were distributed in all the three sections of the lake (north 14, middle 9, south 7), thus were a representative of crane’s nesting-sites. The nests were located in the lake’s marshes with varying levels of standing water, the mean water depths being north (35.76 cm), middle (55.76 cm) and south (64.86 cm), and a range of 14.2–123.63 cm. Cranes located their nests in water points above 50 cm deep, vegetation above 60-90 cm, and preferably 100 m offshore (Figure 3).