Seasonal climate conditions and land use associated with trends in summer Monarch abundance

By Sarah P. Saunders and Elise F. Zipkin

Migration is a captivating phenomenon that continues to fascinate both biologists and the public. In many species—ranging from the smallest insects to the largest mammals—thousands or even millions of individuals move across large portions of the Earth’s surface over extensive periods of the annual cycle. Continuing declines in the abundance of migratory species globally suggest that broad-scale conservation may be increasingly necessary to prevent the loss of migratory populations. However, because few studies have simultaneously quantified the relative contributions of breeding, non-breeding, and migratory processes to population dynamics over large areas, there is little empirical information to test theoretical predictions about how seasonal interactions manifest at the population level. Understanding range-wide fluctuations in the dynamics of migratory populations requires investigations of how events and conditions during different stages integrate across the annual cycle.

The annual cycle of monarch butterflies encompasses three stages outside of migratory periods. Unlike most migratory species, which have distinct winter non-breeding and summer breeding phases, monarchs also have a spatially distinct spring breeding phase. Consequently, the link between population dynamics from winter to spring to summer may be more complex than in other species.

Data indicate that the size of the overwintering monarch population decreased over the past 19 years, a trend that has steepened in the last decade. Several hypotheses have been proposed to explain the long-term decline. Mortality during autumn migration due to low nectar availability, disease, and parasitism may be contributing to population dynamics. During the winter, monarchs are forest specialists and roost exclusively in Oyamel fir forests, a species with an extremely limited range. Illegal logging of Oyamel forests in overwintering sites has led to decreased habitat availability in Mexico. During the spring and summer, monarchs lay their eggs on milkweed hostplants. Reduction in abundance of common milkweed, one of their primary hostplants, due to increased herbicide (glyphosate) use in Midwestern agricultural fields is hypothesized to be a major driver of decreased monarch recruitment. However, the extent to which milkweed loss is contributing to population dynamics, and whether the summer population is declining, is controversial.

We modeled monarch citizen science data collected at sites across Illinois to assess the impacts of climate and land use variables during the winter, spring, and summer stages of their annual cycle on summer abundances. We included variables on local summer climate, crop cover, and county-level glyphosate application. We also incorporated cross-seasonal effects, including annual abundance of wintering monarchs in Mexico and climate conditions during spring migration and breeding in Texas.

Sites where monarchs were surveyed in Illinois and northwest Indiana from 1994 – 2013.​

Our results provide the first empirical evidence of a negative association between county-level glyphosate application and local abundance of adult monarch butterflies in Illinois, particularly in areas with concentrated agriculture. However, this association was only evident during the initial years of the adoption of herbicide-resistant crops (1994 - 2003). Our study also found cross-seasonal associations of both spring climate in Texas and overwintering monarch abundances in Mexico with summer abundances in the Midwest, adding to the growing consensus that seasonal carry-over effects can impact long-term population dynamics.

Expected counts of monarch butterflies at sites with 10% (yellow line) and 50% (blue line) crop cover within 10 km relative to the county-level percentage of corn and soybeans sprayed with glyphosate in Illinois and northwest Indiana from (a) 1994 – 2003 and (b) 2004 – 2013.​

Our results also provide evidence of a decline in abundance of adult monarch butterflies on their summer breeding grounds in Illinois from the first decade in our analysis (1994-2003) to the second (2004-2013), consistent with patterns suggested less strongly by previous research.

Given the limited spatial extent of the survey data relative to the breeding range, we cannot directly assess the extent to which land use, climate, and carry-over effects are contributing to the overall dynamics – and any potential declines – of monarchs across eastern North America. Additional research incorporating data from a larger portion of the summer breeding range and dynamics during autumn may provide useful information on broad-scale monarch population dynamics.

Our study underscores the importance of examining species-specific climate and habitat associations across the annual cycle to capture fine-scale fluctuations in responses to varying environmental conditions. Responses of population dynamics to climate and land use at several points in time are rarely assessed, but may be spatially and temporally complex, especially for migratory species that travel through multiple geographic regions and diverse ecosystems.

Comments