Final Report: Predicting the Distribution and Dominance of Exotic Species Across Landscapes of Southern AppalachiaEPA Grant Number: R828897
Title: Predicting the Distribution and Dominance of Exotic Species Across Landscapes of Southern Appalachia
Investigators: Huston, Michael A. , Johnston, J. W. , Pounds, Larry R.
Institution: Interdisciplinary Solutions for Environmental Sustainability Inc.
EPA Project Officer: Hiscock, Michael
Project Period: September 1, 2001 through August 31, 2004 (Extended to January 31, 2006)
Project Amount: $448,205
RFA: Exploratory Research to Anticipate Future Environmental Issues (2000) RFA Text | Recipients Lists
Research Category: Ecological Indicators/Assessment/Restoration , Water , Ecosystems
The objectives of this research project were to: (1) test ecological theory of invasions based on environmental properties; and (2) develop a method for predicting spatial patterns of exotic species invasion and dominance across landscapes, as well as the time period over which a particular exotic species is likely to be a problem in a given area. The approach taken here differs from the traditional approach to understanding plant invasions, which focused on the properties of species, specifically those properties that made them more “weedy” or “invasive.” Our approach is based on the dynamic equilibrium model of species diversity (Huston, 1979, 1994) and assumes that both native and exotic species respond similarly to environmental conditions, particularly those related to mortality-causing disturbances and resources necessary for growth. This model predicts that the most favorable conditions for plant establishment and survival (and consequently, high species diversity) should occur where physical conditions allow survival of most species but competitive interactions are weak and competitive exclusion occurs slowly, if at all. These conditions are predicted to occur where resource availability (and thus plant growth and productivity) is low and disturbances are infrequent but also where plant resource availability is high and mortality-causing disturbances are frequent (Huston, 1979, 1994, 2004). One corollary is that the diversity of native and exotic species should be correlated positively. This prediction is the opposite of classical competition theory, which predicts that the correlation should be negative, because communities with a high diversity of native species should be more resistant to invasion than communities with low diversity. In addition, the model predicts that plant dominance should be highest, and thus the potential impact of invasive species greatest, under productive conditions where competitive exclusion occurs more rapidly. Our model of invasions predicts the probability and intensity (impact) of plant invasions, as well as the probable duration of the survival of a population of exotic species, based on the rate of competitive exclusion. This project tested these predictions using field surveys in four regions with different environmental conditions and histories of human impacts.
The traditional approach to understanding invasions has focused on properties of the invasive species and of the communities that are invaded. A well-established concept is that communities with higher species diversity should be more resistant to invaders. Most recently published field data, however, contradict this theory, finding instead that areas with high native plant diversity also have high exotic plant diversity. An alternative approach to understanding patterns of invasions focuses on properties of the environment related to the survival and competitive interactions of organisms. This approach assumes that native and exotic species respond similarly to environmental conditions, and thus predicts that they should have similar patterns of abundance and diversity. Population responses and species diversity are predicted in relation to the interaction of productivity and population growth rates with the frequency and intensity of mortality-causing disturbances. The approach distinguishes between the probability of establishment and the probability of dominance, predicting that establishment should be highest under unproductive and undisturbed conditions and also disturbed productive conditions. The probability of dominance by exotic species, and thus of potential negative impacts, is highest under productive conditions, which also is where native species can achieve high dominance. The theory predicts that a change in disturbance regime can have opposite effects in environments with contrasting levels of productivity. Manipulation of productivity and disturbance provides opportunities for resource managers to influence the interactions among species, offering the potential to reduce or eliminate some types of invasive species.
Consistent patterns were found across all four study regions—Oak Ridge Department of Energy Reservation, Tennessee; Great Smoky Mountains National Park, Tennessee and North Carolina; Big South Fork National River and Recreation Area, Tennessee; and Edwards Plateau, Hays County, Texas.
At each site the total number of exotic species was approximately 10 percent of the number of native species, but most of the exotic species were very rare. We found a positive correlation between the number of herbaceous exotic species and the number of herbaceous native species in all areas. Thus, sites with few native species had few exotic species, and sites with the most exotic species generally had many native species. In each area the highest ratios of exotic species to native species (that is, the slope of the line) were found in areas with the lowest canopy cover from woody vegetation. In the forested regions in Tennessee, the areas with least forest canopy cover tended to be areas that had been disturbed by human activities, such as clearing or mowing, or mortality of tree species. In the savannas of Texas, the areas with the highest woody canopy cover were areas that had been invaded by Ashe Juniper.
Sites with the highest dominance (percent cover) of exotic species tended to be sites with the highest cover or height (and presumably biomass) of herbaceous plants. Many sites with high plant cover, however, had few or no exotic species.
Oak Ridge Reservation, Tennessee
Our 240 sample plots (2.4 ha total area) encountered approximately 480 out of the total of 1100 plant species recorded on the Oak Ridge Reservation. This was 44 percent of the total recorded flora (47% of the native flora and 22% of the exotic flora) on 0.02 percent of the total land area. Of the 480 species, 37 species were nonnative, out of the total of 167 nonnative species recorded on the reservation. Most of the 37 exotic species occurred at very low frequencies, with only three species (Lonicera japonica, Microstegium vimineum, and Ligustrum sinense) present in more than 10 percent of the 240 10 x 10 m plots. The correlation between exotic and native species was positive, with the highest slopes among sites with the least forest canopy cover.
The number of exotic herbaceous species (and of native species) was correlated negatively with the density of the forest canopy (as estimated by total basal area or by a spherical densiometer).
Past land use had a strong impact on the present ratio of exotic to native species. Forest areas with the highest basal area (as a consequence of less past or present disturbance, greater time since disturbance, or greater productivity) had the lowest ratio of exotic to native species. One of these areas was subsequently determined to be primary forest. Some of the areas that were cultivated fields prior to the Manhattan Project had very low living tree basal area at the time of sampling because of recent high mortality of pines that had established on the abandoned fields or were in plantations. These areas were infested heavily with Lonicera japonica, which was widely distributed over the landscape at low densities, but responded to increased light availability caused by any type of disturbance. Given the intensive agricultural use of the landscape prior to the Manhattan Project, however, it is clear that nearly all agricultural weeds have been greatly reduced or eliminated. The formerly agricultural landscape is now covered by forest with a relatively low incidence of exotic species.
Great Smoky Mountains National Park, Tennessee and North Carolina
Unlike our systematic sampling of the entire landscape on the Oak Ridge Reservation, in the Great Smoky Mountains we focused our sampling on areas where we expected to find exotic species. All of our transects began in campsite or trail areas, with different recreational uses (horse or hiking only) and at different elevations. Two of the lower elevation sites were associated abandoned settlements.
In a total of 52 10 x 10 m plots over all sites, we encountered 411 native species and 44 exotic species, which is 32 percent of the 1267 native species identified from the park, but only 13 percent of the 343 exotic species recorded for the park. Of the 44 exotic species encountered, only six were present in more than 10 percent of the plots. These were Microstegium vimineum, Ageratina altissima (Eupatorium rugosum), Lonicera japonica, Rumex crispus, Anthoxanthum odoratum, and Dioscorea oppositifolia.
Similar patterns were found at each elevation. The ratio of exotic to native herbaceous species richness was highest among sites with the least canopy cover, and the number of exotic species declined with increasing canopy density or stand basal area.
Big South Fork National River and Recreation Area, Tennessee
We sampled 32 plots on the three transects across Bandy Creek and 29 on the three transects across Unnamed Branch (plus two additional plots at the mouth of each creek). 374 taxa were identified to species, 48 to genus or family, and 54 observations could not be identified. Of the 374 taxa, only 12 were exotic, which is one-third of the ratio found in the other two regions in Tennessee. In addition to the low species richness of exotic species, they occurred infrequently on the landscape. Out of the 65 plots sampled, the most frequent exotic species was Microstegium vimineum, which was found in five plots. Microstegium was followed in frequency by Stellaria media (four plots), Prunella vulgaris (three plots), Rosa multiflora (two plots), and Chrysanthemum leucanthemum (two plots). None of the exotic species exceeded 2 percent cover in the 1 m x 1 m plots where they occurred.
The rarity of exotic species in these two watersheds precludes any statistical analysis for the patterns found at the previous two sites. The highest numbers of exotic species (two or three) were found in the plots at the confluence of each stream with the river into which it flowed. These plots also had the highest richness of native species (76 species/0.01 ha) and were areas with low canopy cover that were subjected frequently to major flooding. There was a significant difference in total species richness between the watersheds, with Unnamed Branch having nearly twice as many species per 10 x 10 m plot as Bandy Creek ( 55.6 vs. 31 sp/0.01ha). Eight state or federally listed species were identified in the study plots.
In addition to the plots along transects, we also surveyed valley bottoms, roads, and trails in the two watersheds for the present of exotic species, occurrences of which were mapped using GPS. By focusing on these disturbed areas with potential propagule inputs, we encountered many more exotic species than in our transect plots. These surveys found 21 exotic species of trees, shrubs, and herbaceous plants along the 15 miles of Bandy Creek roads, trails, and valleys, and five exotic species along the 10 miles of roads, trails, and valleys in Unnamed Branch.
Edwards Plateau, Texas
Sampling during the no-cost extension was prevented by a severe drought, so no new data were collected from the plots that were established at the Pollard Preserve. In the previous sampling at the Freeman Ranch, 176 taxa, of which 133 were identified to species, were encountered in sampling at the 100 transect locations over the course of three seasons. Of these 133 taxa, only 4 were identified as exotic. Two of these were conspicuous in some areas, whereas two were very infrequent. Medicago minima was found in 35 percent of the plots and could have high cover. Bothriochloa ischaemum (“King Ranch Bluestem”) had lower frequency but could also have high cover where it occurred. The other two species Cynodon dactylon (“Bermuda Grass”) and Lamium spp. were very infrequent.
In spite of the low number of exotic species, there was nonetheless a positive correlation between exotic and native species richness. Total species richness had a unimodal relationship with total plant cover, with highest species richness at intermediate levels of cover for both fall and spring samples. Similar to the observations at Oak Ridge and the Great Smoky Mountains, both the percentage of exotic species cover and the ratio of exotic to native species richness decreased with increasing woody basal area (a correlate of canopy density, which was not measured). In most cases this increase in basal area was the consequence of invasion by Ashe Juniper (Juniperus asheii), a native species that has expanded with fire suppression and converted substantial areas of oak savanna to dense juniper thickets. A final pattern is that the maximum percent cover of exotic species increased with increasing total plant cover, with maximum exotic cover being more than twice as high in the fall growing season as in the spring growing season.
Experimental Research on Microstegium viminium “Japanese Stilt Grass”
Microstegium, an annual grass, has been identified by numerous invasive plant groups as one of the most serious invasive plant species in the Eastern United States, primarily because it can achieve high densities in mesic environments. Dr. Patrice Cole completed her Ph.D. on Microstegium in 2003 at the University of Tennessee, Knoxville, Tennessee, with some of her initial work supported by this grant. Her work was initiated by the observation that the distribution of Microstegium on the Oak Ridge Reservation was limited to a small proportion of the total landscape. Microstegium virtually was absent from most of the oak-hickory forest covering the hills and was found almost exclusively in moist valley bottoms, as well as open disturbed areas, regardless of moisture conditions. It occupies shaded moist areas, but Dr. Cole noticed that is was excluded from some of the areas with deepest shade, such as under stands of Pawpaw (Asimina triloba) and Sugar Maple (Acer saccharum). She conducted field experiments that demonstrated that the primary limiting factor for Microstegium occurrence under Pawpaw was light, and that water, soil allelochemical effects, and seed supply were relatively unimportant. In greenhouse experiments, she grew Microstegium under five levels of light (75, 25, 20, 5, and 2% of full sunlight) crossed with four moisture levels (40, 30, 20, and 10% volumetric water content), with five replicates of each treatment combination. At the end of the growing season above- and belowground biomass per plant were measured as the response variables. Results were consistent with the interpretation that Microstegium survival and growth in the shade were more dependent on high water availability than growth in the sun.
The basic predictions of the Dynamic Equilibrium Model of Plant Invasibility were supported by the field surveys. Native and exotic plant species responded similarly to environmental conditions, to the degree that native and exotic richness were correlated positively, in direct contradiction of the prediction of classical invasion theory that they should be correlated negatively because high diversity communities should be most resistant to invasion. All of the data suggest that high and low diversity communities are invasible equally, because the ratio of exotic to native species stays the same regardless of native species richness. The ubiquity of this positive relationship ratio suggests that species richness has no effect on the plant establishment phase of invasion.
Native and exotic species richness responded similarly to variation in plant productivity (as estimated by plant height and the percentage of the ground surface covered by plants). In contrast to establishment, which occurred across the entire productivity gradient (depending on disturbance frequency; Huston 1979, 1994, 2004), dominance of exotic species tended to be highest where total plant cover was the highest (presumably the most productive sites) as predicted by the model. Dominance by both native and exotic species occurred primarily under the most productive conditions, with adequate supplies of soil nutrients, water, and light. Frequent, severe disturbance, as demonstrated by the flood-scoured riparian zone of the Cumberland Plateau rivers, prevented dominance of both native and exotic plant species and allowed the coexistence of many species under high productivity conditions.
The ratio of exotic species richness to native species richness (slope of the regression line) was highest under disturbed conditions with high levels of light availability and lowest where woody plant canopies reduced light availability in the understory. This pattern primarily is a consequence of the fact that most exotic species are fast-growing, shade intolerant species. Those exotic species that are shade tolerant, such as Microstegium and Garlic Mustard, are less likely to follow the patterns found in this study and can achieve dominance even under shaded conditions.
Historical land use disturbances such as agriculture and logging leave a legacy of effects related to reduced forest cover and changes in species composition, but the majority of the post-disturbance exotic flora does not necessarily persist through time. At Oak Ridge, the areas of forest with the least historical disturbance (including virgin forest) had few or no exotic species. In contrast, areas that had been cultivated and invaded by pines after agricultural abandonment experienced a second wave of disturbance at the time of this study, when the pines had reached maturity and succumbed to pine bark beetle mortality exacerbated by overcrowding and a strong drought.
Implications of this research for management include: (1) locations with high probability of invasions are highly predictable, so detection and control efforts can be prioritized; (2) high species richness does not deter invasions; (3) manipulation of disturbance and/or productivity can alter survival and dominance of invasive (and native) species; (4) undisturbed natural communities are rarely, if ever, invaded, so it is important to minimize disturbance to prevent invasions; (5) succession to larger plant sizes and species with denser canopies eliminates most invaders, so it is important to minimize disturbance to eliminate invaders; and (6) under low productivity conditions, most invaders have little effect on native species.
Huston MA. A general hypothesis of species diversity. American Naturalist 1979;113:81-101.
Huston MA. Biological diversity: the coexistence of species on changing landscapes. Cambridge: Cambridge University Press, 1994.
Huston MA. Management strategies for plant invasions: manipulating productivity, disturbance, and competition. Diversity and Distributions 2004;10(3):167-178.
Lonsdale WM. Global patterns of plant invasions and the concept of invasibility. Ecology 1999;80:1522-1536.
Stohlgren TJ, Binkley DA, Chong GW, Kalkhan MA, Schell LD, Bull KA, Otsuki Y, Newman G, Bashkin M, Son Y. Exotic plant species invade hot spots of native plant diversity. Ecological Monographs 1998;69:25-46.