Sex‐biasing gene drives for island mouse eradications: Modelling dispersal and mating behaviour for improved risk assessment preparedness

Invasive rodents remain a major source of environmental damage worldwide. Sex‐biasing gene drives may provide an alternative to toxicant‐based eradication strategies. Risk assessments for gene drives will rely partially on models, which must account for intraspecific variation in the ecological and life‐history characteristics of target populations. Dispersal patterns and mating system characteristics in particular are expected to impact drive spread and suppression outcomes, though empirical data on natural populations are lacking.

We developed a stochastic metapopulation model implemented using numerical simulations that simulates the introduction and spread of male‐biasing X‐Shredder or female‐biasing Y‐Shredder gene drives to suppress an island population of house mice ( Mus musculus ). Our model includes parameters for mating system complexity, sperm competition, dispersal frequency and distance, as well as several features related to release strategy and drive efficiency (shredding efficiency for both drives, homologous site cutting efficiency for Y‐Shredder drives).

We found that both drive strategies are capable of complete population eradication within ~15–35 years depending on parameterization. Drive efficiency had the strongest impact on eradication probability and timing across both drive types. Increasing dispersal frequency usually improved the probability of successful eradication and reduced the time required to achieve eradication, though we identified several exceptions to this trend. Polyandrous mating limited the success and speed of X‐Shredder drives, while the effect of polygynous mating varied for Y‐Shredder drives, improving eradication probability across certain parameter combinations at lower dispersal frequencies when cutting efficiency exceeded shredding efficiency and shredding efficiency remained above a minimal threshold.

Across parameter space and drive strategies, ‘chasing’ events, in which wildtype individuals repeatedly recolonize eradicated areas, were common and often prevented successful eradication at low dispersal frequency. Dispersal behaviour exhibited a strong influence on the spatial characteristics of recolonization and chasing events.

Synthesis and applications . Though drive efficiency was highly influential to model outcomes, such parameters will likely be known prior to any experimental gene drive introductions. We conclude that understanding the ecological and life‐history characteristics specific to target populations will be critical to accurately predict gene drive outcomes and inform risk assessments.