Stochastic environments shape life-history traits and can promote selection for risk-spreading strategies, such as bet-hedging. Although the strategy has often been hypothesised to exist for various species, empirical tests providing firm evidence have been rare, mainly due to the challenge in tracking fitness across generations. Here, we take a ‘proof of principle’ approach to explore whether the reproductive strategy of multiple-batch spawning constitutes a bet-hedging. We used Atlantic cod (Gadus morhua) as the study species and parameterised an eco-evolutionary model, using empirical data on size-related reproductive and survival traits. To evaluate the fitness benefits of multiple-batch spawning (within a single breeding period), the mechanistic model separately simulated multiple-batch and single-batch spawning populations under temporally varying environments. We followed the arithmetic and geometric mean fitness associated with both strategies and quantified the mean changes in fitness under several environmental stochasticity levels. We found that, by spreading the environmental risk among batches, multiple-batch spawning increases fitness under fluctuating environmental conditions. The multiple-batch spawning trait is, thus, advantageous and acts as a bet-hedging strategy when the environment is exceptionally unpredictable. Our research identifies an analytically flexible, stochastic, life-history modelling approach to explore the fitness consequences of a risk-spreading strategy and elucidates the importance of evolutionary applications to life-history diversity. 

Data for this study was simulated with the eco-evolutionary model parameterised for Atlantic cod (.R code is provided). The process is described in the method sections.

There are two types of data sets:

1) A simulation code to drive the eco-evolutionary dynamics of multiple-batch and single-batch spawning cod populations. The code is an extended version of the code developed by https://doi.org/10.1111/j.1752-4571.2011.00215.x 

2) Last-year transcripts of populations that have been preadapted to each scenario of environmental pressure. These transcripts serve as import files or initial populations that are needed to run the simulation code.