Plant functional traits reflect different dimensions of species invasiveness

Abstract

Trait-based invasiveness studies typically categorize exotic species as invasive or noninvasive, implicitly assuming species form two homogenous groups. However, species can become invasive in different ways (e.g., high abundance, fast spread), likely relying on different functional traits to do so. As such, binary classification may obscure traits associated with invasiveness. We tested whether (1) the way in which invasiveness is quantified influences its correlation with functional traits and (2) different demography-based metrics are related to different sets of traits. Using a case study of 251 herbs exotic to Victoria, Australia, we quantified species' invasiveness using 10 metrics: four continuous, demography-based dimensions of invasiveness (spread rate, local abundance, geographic and environmental range sizes) and six binary classifications of invasiveness (based on alternative sources and invasion criteria). We examined the correlation between species' invasiveness and a set of four traits known to relate to plant demography and invasion. Then, we examined whether different demographic dimensions of invasiveness were better explained by different sets of traits. We found that the way invasiveness was quantified was important: different traits were linked with different invasiveness metrics, and some traits showed opposite effects across metrics. Species with fast spread were either tall with small seeds (i.e., good colonizers), or had heavy, animal-dispersed seeds. Plants with a large environmental range had greater plasticity for some traits. Locally abundant plants had low SLA and heavy seeds (i.e., strong competitors). Animal dispersal was also key to reach a large geographic range. No traits were consistently related to the six binary classifications. Our results indicate that exotic plants are invasive in different ways and rely on different combinations of traits to be so. Some traits (e.g., seed mass) had complex relationships with invasion: they apparently promote, hampered, or had no influence on different dimensions of invasiveness. Our findings are consistent with the notion that plant species use strategies that may be near optimal under some, but not all, ecological conditions. Compared to binary classifications of invasiveness, the use of invasiveness dimensions advances clearer hypothesis testing in invasion science.Key words: demography metrics; dimensions of invasiveness; environmental and geographic range sizes; exotic plant invasion; invasiveness and invasive species; local abundance; plant functional traits; spread rate.

Addressing context dependence in ecology

“…The X-Y relationship could vary if studies measure X or Y in different ways [57][58][59]. For example, invader occupancy and invader abundance can both indicate invasion level, but factors related to these two metrics can differ [60], meaning studies that use these different indicators of invasion level could reach different conclusions [61]. Composite indicators or classification-based indices may be particularly troublesome as the underlying metrics or criteria may be unclear and metrics may inadvertently be conflated [25,62].…”

“…For example, definitions of species invasiveness vary such that two studies, both examining traits related to species invasiveness, could inadvertently examine different dimensions of invasiveness (e.g., local abundance, spread rate, or habitat breadth) [25]. Studies could consequently reach different conclusions about traits linked to species invasiveness solely because of differences in the underlying Y variable (see Figure IC in Box 1) [61]. Similar problems could arise with indices representingamong manydiversity [63], resilience and stability [64], and ecosystem services [65], each of which can be calculated in numerous ways.…”

“…(B) Invasion success of bullfrogs, Rana catesbeiana, is negatively related to native frog species richness in eastern China, but this relationship disappears when the confounding effect of frog hunting is accounted for in a logistic regression model (confounding factors) [44]. (C) Across 236 alien herbs in southeastern Australia, the relationship between species seed mass and spread rate depends on whether seeds are smooth or hooked (upper panel) (interaction effects), but a different relationship is obtained if abundance, not spread rate, is used as a measure of species invasiveness (lower panel) (methodological differences) [61]. (D) Aboveground biomass of invasive Pinus nigra in New Zealand has a unimodal relationship with plant species richness (methodological differences) [70] (image redrawn from [71]).…”

Dimensions of invasiveness: Links between local abundance, geographic range size, and habitat breadth in Europe’s alien and native floras

“…Functional traits play a role in mediating invasion processes, but efforts to identify characteristics of successful invaders have generally resulted in few or inconsistent associations (38,39). However, distinguishing between different components of invasiveness may provide additional clarity if each is influenced by different traits, or if the same trait has contrasting effects on different dimensions (15,16,21,40,41). For example, many plant traits are associated with general trade-offs between rapid growth (i.e., acquisitive growth strategies) versus stress tolerance and survival (i.e., conservative growth strategies) (42)(43)(44), and one can hypothesize scenarios where these divergent strategies are associated with success in different dimensions of invasiveness (40,41).…”

“…However, distinguishing between different components of invasiveness may provide additional clarity if each is influenced by different traits, or if the same trait has contrasting effects on different dimensions (15,16,21,40,41). For example, many plant traits are associated with general trade-offs between rapid growth (i.e., acquisitive growth strategies) versus stress tolerance and survival (i.e., conservative growth strategies) (42)(43)(44), and one can hypothesize scenarios where these divergent strategies are associated with success in different dimensions of invasiveness (40,41). Another example are specialized adaptations for long-distance dispersal that may promote rapid range expansion, both in extent and into new habitats, but likely do not provide any advantages that would influence local abundances (45,46).…”