“…Angert et al, 2009;Usinowicz et al, 2012;Letten et al, 2018;Hallett et al, 2019). Analysis of IGRs can be supplemented when necessary with other persistence metrics (Jeltsch et al, 2019); a challenge for future research will be to understand when and how the positive relationship between IGR 1 and T ext breaks down. In principle, any persistence metric, including T ext , can be partitioned into contributions from different mechanisms, as Simulation results for an asymmetric lottery model with lognormal temporal variation in per capita fecundity, logB j t ð Þ ∼ Normal μ j ,σ 2 ,μ 1 ¼ 0:45,μ 2 ¼ 0:5: (a) IGR 1 (the smaller of the two IGRs) with δ ¼ 0:5.…”