Landscape Feature Effects on the Invasibility of Chicago Region Restored Prairies
Tallgrass prairie is one of the most endangered ecosystems in North America, with its current extent at less than one percent of its historical range. For this reason, there has been a large push within the field of restoration ecology to better understand restored prairies. One of the biggest threats that faces the prairie ecosystem is the invasion of non-native species, which can negatively affect the ecosystem in various ways. The focus of this paper is the relationship between landscape features and the invasibility of restored prairies in the Chicago region. To analyze possible relationships between landscape features and prairie community invasibility, community data were collected from 19 restored prairie sites. Species were then categorized as either native, exotic, naturalized, or invasive. Geographic information system (GIS) software was then used to examine the topology and surrounding land cover of each site. These data were then analyzed in conjunction with the plant community data. No strong relationships were found between prairie topology and non-native species richness. A positive relationship was found between increasing surrounding development and exotic species richness, however this relationship was not significant. I conclude that while landscape features are not a large influence on the invasibility of restored prairies, restoration practitioners should still consider these features when making management and seeding decisions, as they may affect other ecosystem functions.
Tallgrass prairie is among North America’s most diverse ecosystems yet one of its most endangered; it is for these reasons that one of the most prevalent conservation strategies of recent years has been the restoration of these ecosystems (Rowe 2010). Beyond increasing biodiversity and habitat area for a range of plant communities, restoring lost or degraded ecosystems allows for restoration of a number of ecosystem services, including nutrient and water retention and pollution absorption (Benayas et al. 2009).
One of the biggest threats to effective restoration is the invasion of non-native species, which can have a host of negative impacts on ecosystems on both large and small scales (Parker et al. 1999). In grasslands, these effects can range from alteration of nutrient and water cycles to adversely affected pollination patterns (Toledo et al. 2014). Invasive species can drastically alter resource availability within an ecosystem and change competitive relationships and even evolutionary paths over large temporal scales (Jones et al. 1994). Invasions can also have contradictory effects, with increased biomass but lower diversity in mixed-grass prairies (Henderson and Naeth 2005). For these reasons, understanding the relationship between prairie restorations and invasive plant species is crucial for establishing sustainable ecosystems.
Improving prairie restoration projects involves multiple considerations, including the informed selection of plants to be selected for prairie seeding. A study of dominant species yielded data suggesting that the native grass (Andropogon virginicus) being observed in some plots resisted invasion better than the non-native species observed in other plots. (Emery and Gross, 2006). Considering invasion-related traits when selecting species for prairie restorations can lead to sites with increased invasion resistance (Mahaney et al. 2014). Additionally, introducing native plants in restored prairies has been shown to be correlated with decreased non-native plant density as well as increased ecosystem stability. (Middleton et al. 2010). To combat invasive species in prairie ecosystems, it is important to continue utilizing adaptive management techniques like prescribed burns to maintain native plant communities. (Grant et al. 2009).
A relatively lesser-studied factor in prairie restorations is the effect of surrounding landscape features. A review by With (2002) cited six effects of landscape properties on invasion. These include: directly and indirectly influencing seed dispersal, altering species interactions, and creating spatiotemporal fluctuations in resource availability. The same study concluded that habitat fragmentation can both reduce invasive establishment and increase their relative amounts of seed dispersal success. In a study of the invasive shrub Lonicera maackii, Bartuszevige et al. (2006) concluded that landscape structure is more important than community structure in wooded areas, with edge effects and proximity to centers of population being the most influential factors. The researchers emphasized the importance of the seed dispersal method to plant invasions. Community assembly filters that determine how certain communities establish themselves are very important to prairie restoration efforts (Grman et al. 2015) and could include landscape features, but more empirical research in the area is needed.
The likelihood that an ecosystem will experience invasion is influenced by the shape of the area. All else held equal, habitats with complex topology are more susceptible to invasion, however topology and other related features are often difficult to quantify (Cumming 2002). Site topology is connected to a number of features such as amount of edge and the ratio of edge to total site area.
A better understanding of the effects of landscape features on invasion resistance could lead to more effective establishment and management of restored prairies. Biological invasions are likely to increase in both extent and severity as humans continue to introduce new species to places outside their evolutionary contexts; fighting invasions will be important in the future to conserving biodiversity around the world (Hobbs and Humphries 1995).
In the present research, I hypothesize that that the surrounding landscape features will have a quantifiable effect on restored prairie communities. Specifically, a higher ratio of perimeter to area is expected to yield higher proportional numbers of invasive species present, as has been observed in previous studies (Bartuszevige et al. 2006). I do not expect to find any relationship between prairie size and invasive species richness, based on the previous findings (Cully et al. 2003). Moreover, prairie sites surrounded by both highly and moderately developed areas are not expected to experience higher or lower amounts of invasive species, compared to sites surrounded by low development, agriculture, or preserved lands. This is likely due to the lower species richness and lower potential for invasive species to be accidentally introduced in urban/suburban areas.
A total of 19 restored prairies in the Chicagoland region were surveyed (Figure 1). The prairies were restored by the same firm, Pizzo & Associates. Studying sites restored exclusively by a single firm meant that variables related to seed sourcing and planting methods were similar across sites. Factors such as invasive species management, controlled burns, and time since restoration varied across sites.
To quantify non-native species at each site, I vegetatively sampled sites. At each study site, I laid out two 50-meter transects across areas that appeared to be most representative of the prairie as a whole, avoiding site edges. Study plots were selected every five meters along each transect, for a total of 20 plots per site. Each ¼ m2 plot was randomly chosen to be either left or right of the transect line, at a random distance of 2-7 whole meters. I recorded all species present in each plot. Species that I could not identify in the field were either photographed or collected and identified at the Chicago Botanic Garden. All species identified were placed into one of my categories: exotic (not from North America), naturalized (spreading but not invasive), or invasive (invading native communities). An invasive species was also considered naturalized and exotic, and a naturalized species was also considered an exotic. Additionally, I recorded GPS points at each plot, as well as along each 50-meter transect.
I performed all necessary GIS analyses using ArcMap, part of the ArcGIS suite. Data layers including a digital elevation model, an orthophoto, and county boundaries were sourced from the Illinois Geospatial Data Clearinghouse (Illinois Geospatial Data Clearinghouse. 2015). The land cover data layer, from 2011, was sourced from the US Geological Survey’s Land Cover Institute (United States Geological Survey. 2015). Boundaries of restored prairies managed by the Chicago Park District were sourced directly from the City of Chicago. These layers, in addition to the GPS points recorded during my surveys, were all uploaded into a single ArcMap document for analysis.
I used the 2014 orthophoto to create a layer of polygons to represent the study sites. Next, I created three separate layers of buffer zones to surround each prairie. The layers included buffer zones of radii 0.5, 1, and 2 kilometers around each site (Figure 2). Using the 2011 land cover data as the input layer, I generated a zonal statistics table for each of the three buffer zone layers. These tables, along with the polygon layer’s attribute table, were then exported from ArcMap for further analysis.
To analyze prairie shape and its potential relationships with non-native species establishment, I divided each site’s perimeter by its area and used this value as a buffer trait for topology. Low values indicate low relative amounts of edge and a relatively continuous site, while high values indicate larger amounts of edge and a more dissected site.
In order to evaluate the land cover surrounding the prairies being studied, I took the mode land cover value from each prairie for all three buffer zones and grouped them into my categories: preserve (surrounded mostly by preserved lands, which in all cases meant more prairie), agriculture (which includes both cultivated lands and pastures), low-density development (developed open space and low intensity development areas covered in less than fifty percent impermeable surfaces), and high-density development (medium and high density development areas covered in greater than fifty percent impermeable surfaces) (Figure 2). These categories were then assigned numerical values in order to create a gradient of land cover development and run linear regressions: 1 (preserve), 2 (agriculture), 3 (low-density development, and 4 (high-density development).
To determine the strength of relationships between landscape effects and prairie invasabiltiy I used linear regressions and calculated R2 values for exotic, naturalized and invasive species against prairie area, edge, and surrounding land cover. Using the non-native species data and the land cover classifications (1-4), I ran ANOVA tests and post-hoc Tukey tests using software from statistica.moo.com (Vasavada, 2014).
Prairie topology was not generally a good predictor of the species makeup of the ecosystems surveyed (Figure 3). No relationship was found between prairie size and the species composition of the restoration. Similarly, there was no correlation between topology and the abundance of exotic or naturalized species. This pattern is broken by invasive species abundance, however, which is negatively correlated with the amount of edge at a given site.
The land cover surrounding each study site was not, in general, related to the numbers of naturalized or invasive species (Table 2). The exception to this trend was exotic species abundance, which was related to land cover at all three buffer zone sizes (Figure 4).
Iran ANOVA tests using the my land cover classifications, with no significant results generated. A one-way ANOVA tests of land cover within the 0.5 km buffer did not yield any p-values greater than 0.1 for exotic, naturalized, or invasive species levels (Vasavada, 2014).
The results of this research does not support all of the hypotheses I used to direct the course of this study. My first hypothesis, that increased amounts of edge would lead to higher numbers of exotic, naturalized, and invasive species was not supported. In fact, I observed a negative relationship between invasive species richness and the amount of edge at a site. In addition to being the opposite of my hypothesis, this finding seems to contradict much of the conventional wisdom about edge effects (Laurance and Yensen 1991, Murcia 1995). It should be noted that that site 11, which can be seen in Table 1 and Figure 3C as having zero invasive species present, did have at least one invasive species present at the site. However, no invasive species were found within my study plots, and were thus not included in analyses. In addition, all transects were chosen to specifically avoid the edges of restored prairies. Sampling the edges instead of the core of the prairies could reveal different relationships. The observation that there is no relationship between restoration size and exotic, naturalized, or invasive species richness supported my original hypothesis.
My hypothesis that the surrounding land cover would have minimal effect on non-native species richness was partially supported. There were weak relationships between increasing development of the surrounding landscape and the number of naturalized or invasive species present. This trend was broken by exotic species richness, which positively correlated with increasing development at all three buffer zone sizes. This result implies that increasing levels of development could be tied with the invasion of prairies by exotic species. This could be due at least in part to more developed areas having more potential for the introduction of exotic species. However, as indicated by the ANOVA tests run using the land cover classifications, there were no significant differences in the amount of exotic, naturalized, or invasive species between prairies surrounded by different landscape types.
The results of this study could potentially impact the practices of ecological restoration creators and managers. First, the lack of a relationship between prairie edge and prairie size with non-native species richness implies that these landscape factors do not need to be given significant weight when making restoration decisions. It could mean that areas once thought to be too small or fragmented for restorations are now seen as viable candidate sites. Additionally, while land cover could be a factor when making decisions about restoration practices due to the relationship between development and exotic species richness, it should not be heavily weighted. When trying to create prairie restorations with high biodiversity and low numbers of non-native plants, it is probable that the most important factors are seeding and management methods.
It is difficult to say with certainty what the current consensus is regarding landscape effects on restored prairie invasibility. Much of the current literature is focused on the seeding of restorations, with emphasis placed on the ability of seed mix richness to produce higher-diversity prairies (Larson et al. 2011) and the ability of more diverse prairies to block invasive species (Foster et al. 2015, Oakley and Knox 2013). More research in the area is needed.
There are a number of ways in which future research could build upon the results presented here. Future surveys could record relative abundance data in order for more complex diversity and community indices to be created. The creation of a data set that contained management history, such as controlled burns or efforts to reduce invasive species abundance, would also be helpful to future research. In addition, higher resolution land cover data would allow for more nuanced analysis and perhaps the discovery of additional trends. These improved methods, whether utilized separately or in conjunction with one another could very well lead to greater knowledge about restored prairie landscapes. Further, this research could lead to the more effective restoration of an ecosystem that has been largely destroyed due to human activity.
I would like to thank Evelyn Williams for her feedback and guidance throughout the various stages of writing this paper. I would also like to thank Andrea Kramer, Matthew Rossi, Emily Yates, and Rebecca Barak for their assistance with my research. I thank Evelyn Williams, Rebecca Barak, Meghan Kramer, Gabriela Carr, and Alyssa Wellman Houde for their help with prairie surveying. Finally, I thank Northwestern University and the Chicago Botanic Garden for their support. This research was supported by NSF Award #1354426.
Bartuszevige, AnneM. et al. 2006. The relative importance of landscape and community features in the invasion of an exotic shrub in a fragmented landscape. Ecography. 29 (2): 213-222.
Benayas, Jose M. Rey et al. 2009. Enhancement of Biodiversity and Ecosystem Services by Ecological Restoration: A Meta-Analysis. Science. 325 (5944): 1121-1124.
Cully, Anne C. et al. 2003. Invasion of Exotic Plant Species in Tallgrass Prairie Fragments. Conservation Biology. 17 (4): 990-998.
Cumming, G. S. 2002. Habitat Shape, Species Invasions, and Reserve Design: Insights from Simple Models. Conservation Ecology. 6 (1): 3.
Emery, Sarah M. and Gross, Katherine L. 2006. Dominant species identity regulates invasibility of old-field plant communities. Oikos. 115 (3): 549-558.
Foster, Bryan L et al. 2015. Does tallgrass prairie restoration enhance the invasion resistance of post-agricultural lands? Biological Invasions. 17 (12)
Grant, T. A. et al 2009. An Emerging Crisis across Northern Prairie Refuges: Prevalence of Invasive Plants and a Plan for Adaptive Management. Ecological Restoration. 27 (1): 58-65.
Grman, Emily et al. 2015. Dispersal and establishment filters influence the assembly of restored prairie plant communities. Restoration Ecology. 23 (6): 892-899.
Henderson, Darcy C. and Naeth, M. Anne. 2005. Multi-scale impacts of crested wheatgrass invasion in mixed-grass prairie. Biological Invasions. 7 (4): 639-650.
Hobbs, Richard J. and Humphries, Stella E. 1995. An Integrated Approach to the Ecology and Management of Plant Invasions. Conservation Biology. 9 (4): 761-770.
Jones, Clive G. et al. 1994. Organisms as Ecosystem Engineers. Oikos. 69 (3): 373-386.
Larson, Diane L. et al. 2011. Effects of planting method and seed mix richness on the early stages of tallgrass prairie restoration. Biological Conservation. 144 (12): 3127-3139
Laurance, William F. and Yensen, Eric. 1991. Predicting the impacts of edge effects in fragmented habitats. Biological Conservation. 55 (1): 77-92.
Mahaney, Wendy M. et al. 2014. Impacts of prairie grass species restoration on plant community invasibility and soil processes in abandoned agricultural fields. Applied Vegetation Science. 18 (1): 99-109.
Middleton, Elizabeth L. et al. 2010. The Effect of Restoration Methods on the Quality of the Restoration and Resistance to Invasion by Exotics. Restoration Ecology. 18 (2): 181-187.
Murcia, Carolina. 1995. Edge effects in fragmented forests: implications for conservation. Trends in Ecology and Evolution. 10 (2): 58-62.
Oakley, Clinton A. and Knox, John S. 2013. Plant species richness increases resistance to invasion by non-resident plant species during grassland restoration. Applied Vegetation Science. 16 (1): 21-28.
Parker, I. M. et al. 1999. Impact: toward a framework for understanding the ecological effects of invaders. Biological Invasions. 1 (1): 3-19.
Rowe, Helen I. 2010. Tricks of the Trade: Techniques and Opinions from 38 Experts in Tallgrass Prairie Restoration. Restoration Ecology. 18 (S2): 253-262.
Toledo, David et al. 2014. Extent of Kentucky Bluegrass and Its Effect on Native Plant Species Diversity and Ecosystem Services in the Northern Great Plains of the United States. Invasive Plant Science and Management. 7 (4): 543-552.
With, Kimberly A. 2002. The Landscape Ecology of Invasive Spread. Conservation Biology. 16 (5): 1192-1203.