Independency in the volumes of eggs of the Heerman's gull (Larus heermanni), on base to their clutching sequence

Abstract

In most birds egg volume has been recorded directly dependant on clutch sequence. Furthermore, the egg size has been also considered as dependent on the size of its predecessor, thus, they are constantly analyzed with non-parametric statistics, given the independency required by the parametric one. We studied the Heermann's gull (Larus heermanni) colony at El Callo Islet (south of San José Island, B.C.S.) during its reproductive season (1996). Twenty-seven nests were followed and identified with plastic flags. The eggs were marked (according to their appearance) and measured (length and width), in order to calculate their volume. The analysis of the volume of 66 eggs showed normality and homocedasticity (in all cases α = 0.05). The volume comparison by clutch sequence evidenced that the first and second eggs were similar, but both differed from third eggs ( x^- ± SE: 57.59 ± 1.52, 56.62 ± 1.75 and 52.66 ± 2.48 cm³, respectively). Furthermore, three simple linear correlations were performed (first-second, first-third and second-third), in order to evidenciate the supposed size-dependence on previous eggs. No relationship could be established in either comparisons (r = 0.143, 0.08 and 0.291; d.f.= 23, 14 and 14, P = 0.502, 0.784 and 0.313, respectively). This means that the egg volume depends on the clutch sequence, but not necessarily on the volume of previous eggs, indicating the need of testing such dependency and, consequently, the type of statistical analysis to be used.