Common approaches for estimating invasive species risk, such as niche-based or correlation models, focus on average climate conditions and asymptotic dynamics. While these models are useful for understanding indefinite persistence, they may underestimate risk of invasive species impact on timescales relevant to management. One major issue for newly introduced populations is that they are far from an equilibrium stage distribution (e.g., the proportion of juveniles and adults). An unstable stage structure can cause transient dynamics, such as periods of fast population growth. This growth will not persist indefinitely, but nonetheless, can lead to economically or ecologically significant impacts. While such dynamics can arise simply from an unstable stage structure, autocorrelated environmental variation may also contribute by affecting life history parameters such as age at reproduction.
Giant hogweed (Heracleum mantegazzianum) is a recognized invasive plant species in the UK, Europe and the United States. Giant hogweed has been of great concern in Europe because of its human health and economic impact. The sap causes photodermatitis which can range from mild reactions to burned skin with deep blisters. The long generation time of this semelparous plant has perhaps obscured its continued spread in North America. While in some regions it can produce seeds quickly, in others it can take as long as 5 years to reach maturity. We use previously published life history data, and 40 years of climate data in North America, to develop a stage-structured population model of giant hogweed, where age at first reproduction is determined by growing degree days.
Results/Conclusions
We find that unstable stage structure associated with recent introduction, and autocorrelated temporal variation in growing degree days that affects age at first reproduction, can produce transients with large growth rates. The predictions of stage-structured models agree with traditional climate-matching techniques in some regions. However, the population models suggest significant risk in some areas on the periphery of suitable habitat. These regions have a larger degree of autocorrelation in growing degree days between years, and can have multi-year periods of fast growth, even though on average, conditions are barely suitable. We conclude that the use of average climate conditions, and the assumption of asymptotic population dynamics, could underestimate invasive species ecological or economic impact in a large variety of organisms