S. Luke Flory » Ecologist, Indiana University

How do plant invasions affect native species and ecosystems?

Invasions of non-native species occur in many habitats but we have suprisingly little experimental evidence about how they are affecting native species. Research is needed to determine how invasions change the composition of native communities and alter ecosystem processes. Most reports on the impacts of invasions are based on observational (i.e. non-experimental) studies which cannot determine if invaders are driving observed differences or if invasions are a consequence of environmental conditions.

Along with several collaborators, I am evaluating the impacts of Microstegium vimineum (Japanese stiltgrass), a shade tolerant annual grass that is rapidly invading forests in the eastern U.S., on native systems. In 2005, I established 32 - 5.25 x 5.25 m plots on a property of the IU Research and Teaching Preserve and then experimentally added Microstegium to half of the plots.

Within this experimental setup we are determining the effects of invasion on:

1. Native trees at the seedling and sapling stages
2. Native herbaceous community composition, abundance, and diversity
3. Arthropod abundance and diversity (with Carolina Simao and Jennifer Rudgers at Rice University)
4. Plant resource availability including light, water, and nutrients (with graduate student Brett Mattingly, Indiana University)
5. Decomposition and nutrient cycling (with Brett Mattingly, Rich Phillips, Sarah Hoffman)
6. Survival of two tick species (with Dave Civitello and Keith Clay)

Thus far we have found that the invader reduces the survival and alters the growth of some, but not all tree species and dramatically decreases the biomass and diversity of herbaceous communities. Microstegium appears to indirectly impact tree regeneration by increasing vole activity. We have found clear evidence that invasion decreases the diversity and abundance of arthropods and alters species composition. We also found significant effects of invasion on the survival of two tick species that are important disease vectors (Civitello, Flory, and Clay 2008).

These results suggest that invasions will significantly alter forest communities in the eastern U.S. by changing tree species composition and herbaceous community structure. Importantly, these changes in native plant communities also have strong effects on higher trophic levels.

Does post-introduction evolution of invasive species explain their success?

A leading hypothesis to explain species invasions suggests that some species have rapidly evolved following their introduction. The Evolution of Increased Competitive Ability (EICA) hypothesis posits that because species leave their enemies behind when they are introduced and no longer need to defend themselves, they can rapidly evolve greater competitive traits such as high growth rates and reproduction. In addition, multiple introductions of species may result in high rates of hybridization.

I am testing the EICA hypothesis in collaboration with Furong Long, a forestry researcher from Yunnan Province China. We are comparing the performance of 10 populations of invasive Microstegium collected throughout the eastern U.S. with 10 native populations collected in China. The photo at right is of Microstegium growing in its native habitat. We are using greenhouse and common garden experiments conducted across a wide variety of habitat conditions to determine if invasive populations grow more rapidly than native populations and if populations differ in phenotypic plasticity. In addition, we are experimentally evaluating differences in plasticity using a shade treatment in the greenhouse.

I am also collaborating with Theresa Culley to use microsatellite markers for genetic analysis of the native and invasive populations.

What determines the invasibility of a community?

A primary need in understanding the dynamics of species invasions is identifying the traits of susceptible communities. Throughout the eastern U.S., agriculture, urban sprawl, and industrial development have resulted in highly fragmented forests that are dissected by roadways. Roads are known to facilitate the spread of introduced plants by providing dispersal corridors and suitable habitat. Independent of roads, successional age is also a key factor in determining community susceptibility to invasions, with younger habitats often more vulnerable to invasions than mature communities.

We quantified the influence of roads and forest successional age on non-native shrub invasions in southern Indiana and found that invasions were more dense near roads than in interior forest and in young and mid-successional forests than in mature forests (Flory and Clay 2006).

A three-year experimental study using seeds and seedlings of shrub invaders and native species subsequently showed that invader growth was limited at interior and mature sites, but that those habitats did not reduce germination or survival of invaders (Flory and Clay, in press). Interestingly, native species received over 80% more herbivore damage than natives suggesting that release from natural enemies may be contributing to non-native shrub invasiveness - further experimental work is needed to evaluate this possibility.

Overall, it is clear that roads are providing corridors for invasions in eastern forests and that release from enemies may be contributing to invasions. Furthermore, forest disturbances that decrease the amount of forest far from roads will increase forest susceptibility to invasion.

Oviposition preference, host plant response, and cicada offspring performance

During the periodical cicada emergence of 2004, Brett Mattingly and I conducted a field experiment to examine cicada oviposition preference and ciciada offspring performance and the response of host plants to cicada oviposition. We established a plantation with three native and three non-native host plants and experimentally excluded cicadas from half of the plants. Our goal was to answer the following specific questions:

1. Are native plant species more susceptible to cicada oviposition than exotic species?
2. Does cicada oviposition decrease the performance of host plants?
3. Do cicadas oviposit based on species identity or plant architectural traits?
4. Does cicada hatching success depend on host plant species identity, egg nest density, or structural attributes of the plant?

Our experimental results showed that oviposition damage caused by cicadas did not generally affect the growth or reproduction of host plants (Flory and Mattingly 2008). Cicadas preferentially oviposited on some species and chose oviposition sites based on certain architectural features of the plants but offspring performance was largely independent of oviposition site.

Management of Microstegiumm vimineum invasions and recovery of resident plant communities

Restoration of invaded habitats requires both the removal of invasions and reestablishment of native species. However, the effectiveness of methods used to remove plant invaders varies widely and the response of native species may also vary among removal methods.

I tested the effectiveness of three commonly used methods for treating Microstegium invasions and evaluated the response of native communities over three growing growing seasons. I removed invasions with hand-weeding, a post-emergent herbicide, and a post-emergent plus pre-emergent herbicide treatment. 

Our results showed that all methods were highly effective in removing invasions during the growing season. However, unlike the herbicide treatments, hand-weeded plots had more than 75% cover of Microstegium the following spring (Flory 2009). The response of native species varied widely across removal methods and plant functional groups. For example, tree regeneration was increased by 123% in post-emergent treated plots but the number of trees was no greater under the other removal methods compared to untreated plots (Flory and Clay 2009).

These results indicate that invasions can be managed using multiple techniques but that the removal method will determine native community composition following invader eradication. Given these results natural areas managers should select invasive plant control methods based on both their effectiveness and the response of native species. In addition, removal and restoration experiments should consider the efficacy of removal techniques and their effects on native plant community recovery.