Early life stage performance of native and non-native congeners of Polygonum, Celastrus, and Parthenocissus: Assessing methods of screening new plant introductions for invasive potential

I compared early life stage performance of invasive species ( Polygonum perfoliatum, Celastrus orbiculatus) with their native congeners (P. sagittatum, C. scandens) and a noninvasive introduction (Parthenocissus tricuspidata) with an abundant native congener (P. quinquefolia) to assess screening methods for new introductions. Experiments measured seed production/dispersal, seed predation, seed longevity, seedling emergence, herbivory, growth, and survival in three habitats, and growth, plasticity, and competition in the greenhouse. I compared predictions generated from generalizations of invasive species characters with actual outcomes to determine critical differences between invasive and non-invasive species. I used field data to estimate establishment probabilities in three habitats. Polygonum perfoliatum had greater establishment probabilities than P. sagittatum in successional field and edge habitats, but neither species established in the forest interior. Polygonum perfoliatum appears to be invasive because of (1) greater tolerance to low soil moisture, (2) greater dispersal potential because of (a) attraction of avian frugivores and (b) wide seed production period, and (3) greater seed longevity. Celastrus orbiculatus had greater establishment probabilities than C. scandens in each habitat. Celastrus orbiculatus appears to be invasive because of (1) greater tolerance to shade, (2) greater germination rates, and (3) greater seedling survival in all habitats.; Invasive species from Polygonum and Celastrus had greater transition probabilities at nearly every stage, but inclusion of pre-emergence transitions had little effect on relative establishment probabilities in most cases. Neither invasive species appeared to be significantly more competitive or plastic than their native congeners. My studies could not determine why Parthenocissus tricuspidata was not invasive. Parthenocissus tricuspidata establishment probabilities were greater than P. quinquefolia in two of three habitats. Relative differences for individual transition probabilities were evenly mixed between Parthenocissus congeners. Pre-emergence transitions were important in determining relative establishment probabilities for Parthenocissus species. The potential roles of performance-based screening methods and alternatives are discussed. I recommend a hierarchical screening procedure consisting of (1) multivariate statistical methods, (2) regional assessment of early life stage performance under various field conditions and/or greenhouse trials to determine tolerances across various abiotic treatments, and (3) monitoring of newly admitted species that are assumed to have minimal invasive potential.