Can long‐term diapause/dormancy improve persistence in a positively autocorrelated environment?

Abstract A temporal bet‐hedging strategy spreads demographic risk across multiple years via diapause/dormancy mechanisms. Diapause polymorphisms (i.e. individuals varying in the length of their diapause) are said to lead to greater persistence in white‐noise stochastic environments but to be less advantageous in temporally autocorrelated or ‘reddened’ environments because immediate future conditions are more predictable. Only diapause lasting for up to 2 years has been investigated, despite numerous species possessing longer diapause polymorphisms. We predicted that longer diapause polymorphisms (>2 years) would extend the advantage of a bet‐hedging strategy to times when reddened environments become unpredictable again. Using data from the Sinai Baton Blue butterfly, we modelled the effect of longer‐diapause strategies (>2 years) and varying degrees of temporal autocorrelation of environmental stochasticity. Both 3‐year and 8‐year diapause polymorphisms were more advantageous, characterised by greater stability or slower population decline than a simple one‐year diapause in white‐noise stochastic environments; while the longer 8‐year diapause polymorphism was more advantageous in reddened stochastic environments. This marks the first evidence that longer diapause polymorphisms are an evolutionary response to reddened stochastic environments.