Abstract Decisions about which processes to include in a population model can have substantial impact on estimates of population trends and extinction risk. This is particularly important for species of conservation concern, whose conservation status is decided in part on the basis of these models. In their recent paper, Meehan and Crossley (2023) Insect Conservation and Diversity, 16, 566–573 revisit a time series of overwintering monarch butterfly abundances, which previous assessments had characterised as rapidly declining and at risk of extinction. The authors, in contrast, reported no evidence for declines in the past 10 years and characterised extinction risk as low. A primary reason for the difference between these conclusions was that Meehan and Crossley used a more complex model that included a parameter for density dependence. While negative density dependence may play a role in monarch population regulation, there are a variety of unresolved issues with how and if density dependence should be included in models of monarch populations, including widely known issues with separating observation error from density dependence in noisy time series and taxon‐specific issues with fitting models to data surveyed every fourth generation and pooled at continental scales. These issues make the conclusions of Meehan and Crossley about monarch population viability much less robust than implied in their article. They do not provide convincing evidence that density dependence reduces extinction risk in monarch butterflies. Our commentary supports their general conclusion that population viability projections depend on model assumptions, but the ways in which monarch butterfly populations are regulated remains an open question.
Read full abstract