General patterns of age-by-stage distributions

1 The methods of Cochran & Ellner (1992) make it possible to calculate the within-stage stable age distributions for populations modelled with a stage-projection (Lefkovitch) matrix. 2 Applying these methods, I found a general pattern in the form of these distributions across a broad range of taxa: with increasing stage category, the stable age distributions become lower, more symmetric, and flatter. 3 This pattern is found in most but not all populations, and thus is not simply an artefact of the method. Exceptions tend to be populations with multiple new-born types (e.g. with both vegetative and sexual reproduction). 4 The pattern remains qualitatively the same when the number of stage classes is varied, despite quantitative differences, When the number of stages is large, the skewness and kurtosis of the within-stage age distributions appear to decline exponentially with stage. 5 The cause of the pattern seems to be the accumulation of lag times in individuals' progressing to larger stages. 6 This result can be used to test for stability, when populations can be aged directly; to modify methods for estimating age from stage, when they cannot; and to suggest the need to look for vegetative reproduction in the field, when it does not hold.