• density-dependence;
  • key factor;
  • k-values ;
  • λ-sensitivity

1.  Key factor analysis is widely used as the first step in analysing census data to identify factors responsible for population change, but is generally considered to be flawed. The conceptual problems can be overcome by assessing the effects of variation in the life-history parameters on population growth rate, λ. We refer to this as λ-contribution analysis. The difference from key factor analysis is that now each life history parameter is weighted by the sensitivity of λ to that parameter. The rationale for this modification is that population growth rate is the best available measure of population change.

2.  The advantages of the new method are: that it correctly assesses the effects of life history parameters on population growth rate; that birth rates are included in the analysis in a natural way without making arbitrary assumptions about birth rate mortalities; that post-reproductive individuals who do not contribute to population growth rate are zero-weighted; and that the analysis can be applied to populations with overlapping generations.

3.  It is proposed that λ-contribution analysis should replace conventional key-factor analysis as the first step in a wider analysis of population change and density dependence. λ-contribution analysis also links census studies of natural populations with the use of life-table response experiments.