A Weedy Business: Weed Composition on Prairie Edges

Weed management is a large part of landscaping and farming projects around the world. As excellent competitors with high dispersal rates, weed species have the ability to outgrow and out-compete native species. With this in mind, characterizing weed populations in prairies should be an integral aspect of prairie conservation projects. Knowing what types of weeds can grow under what sorts of conditions can help optimize such projects by reducing the interspecific competition caused by weeds.

Lososova et al. (2004) study the effects of a number of environmental factors on the composition of weed communities in Central Europe and aim to rank them in order of the significance of their influence on weed community patterning. The four main factors they test are altitude, season, year, and crop type, each of which variably affect several micro-factors such as temperature and soil pH. Based on statistical tests they conducted on releve data they collect from several databases, their results indicate that, out of the four tested factors, altitude had the highest influence on weed species richness and abundance and season, the second highest. These results helped me and my own team build hypotheses as we went into our project.

While our experimental site is very different from the large spatial range covered by the Lososova study, there were plenty of weed-affecting variables discovered that t translate onto our study. We examine the differences in weed abundance and diversity on either edge of a 220 metre long prairie strip. One edge of the strip runs along an active railway track and the other runs along a patch of lawn. Since the edge along the railway is sprayed in regular intervals to prevent growth onto the tracks, we thought there might be a difference in the weed communities on either side. Lososova et al. (2004) lays out an array of possible variables that might affect the weed abundance and richness on either edge of the prairie strip we are conducting our project in. In their study, altitude was the most significant factor affecting weed composition. They explain that soil pH varies greatly with altitude in their experimental region, suggesting that the pH might be a crucial variable to take into account when battling weeds. While altitude doesn’t vary greatly within our small experimental area, the regular spraying on one side may cause differences in soil pH which may, in turn, cause differences in weed communities along either edge.

                  Fig 1. Sampling on the lawn edge of our prairie strip. The railroad is in the background.

While the results from Lososova et al. (2004) build a solid foundation to kick off our weed project, they also generate some interesting contradictions when compared to the field of prairie conservation research. While previous prairie research has led us to believe that human influences are the largest cause of disruptions to the ecosystem, Lososova and her team suggest that natural geographical and climatological variability are more important for weed species composition than human-based factors such as crop type and other agricultural management. I wonder whether their results will extend to our study in the prairie strip? Furthermore, given that seasonality, including variables like temperature and precipitation, plays the second largest role in weed composition in the study, would the results of our experiment be different if we extended it into a long term project across various seasons?

References

Lososova, Z.,  M. Chytrý, Š. Cimalová, Z. Kropáč, Z. Otýpková, P. Pyšek, and L. Tichý. 2004.    Weed Vegetation of Arable Land in Central Europe: Gradients of Diversity and Species        Composition. Journal of Vegetation Science 15(3): 415-422.          http://www.jstor.org/stable/pdfplus/3236483.pdf?acceptTC=true&jpdConfirm=true

A sense of community: how plants determine which insects live together

Anthropocentric influences of the last century has left our natural world in a declining state, with decreased natural habitats and a resulting decrease in biodiversity. Generally, the scientific community more heavily focuses on the diversity of vertebrates and plants, with these species abundance and diversity declines more publicized and accessible by the general public. However, insect biodiversity may be declining at an even quicker pace than that of vertebrates and plants, and the lack of public awareness of this requires a more effective method to quantify the risk this organismal group is under. Because arthropods can be categorized into a wide range of trophic levels, or roles within in the food chain, numerous factors determine arthropod community structure, and defining which variables influence species composition the most is essential for conservational purposes.

           

Figure 1. Before this study, the species making up an insect community were though to be determined mainly by environmental conditions (such as soil pH, moisture, etc.) and by vegetation structure (map A.) However, this study brought to light that really only plant species composition has a direct impact on the insect community. Map B demonstrates that this is because the environment influences plant species presence/absence, which in turn, determines vegetation structure. Picture from Shaffers 2008.

            Andrè P.Shaffers’ (et al.) ‘Arthropod Assemblages are Best Predicted by Plant Species Composition’ sought to compare local factors to define what best predicted arthropod species presence, also known as ‘assemblage.’ Specifically, the factors he studies were plant species composition (what species form the plant community,) landscape composition, vegetation structure (the way plants grow vertically,) and environmental conditions. Two years of sampling focused on 47 sites ranging from open grasslands, to hay meadows, to dense tall-herb grasslands in the Netherlands, concentrating on seven arthropod groups (spiders, grasshoppers, ground beetles, weevils, plant-hoppers, hoverflies, and bees.)

Surprisingly, environmental conditions and vegetation structure played a weak role in predicting arthropod species occurrence, while plant species composition was a highly influential factor. Before this study, species presence was thought to be attributable to these weaker factors, but Shaffers shows instead that it is the species of the plant community that end up having a direct impact on arthropod assemblage. Intuitively, this makes sense as specific plant species, products of environmental conditions, are relied upon by herbivorous arthropod species for food, and thus determine the herbivorous insect community. As higher trophic level species, such as predator and parasitic arthropods, prey upon the lower trophic levels that are the herbivorous species, the herbivorous species composition will define the rest of the arthropod community. In sum, arthropod community is determined by which plants can successfully grow in the environment to support herbivorous insect species, which in turn, influence which predator species will be present in the plant community as well. By focusing first on plant species composition, certain arthropod communities can be better targeted for conservation, management, or research.

Figure 2. Variations in ‘Vegetation Structure’ can be seen in these pictures by the differing heights and densities of plant growth. (Picture from Shaffers 2008)

            Though Shaffers’ study did not look into the effects of urbanization upon arthropod communities, his findings end up having implications for our own study concerning the differences in ground beetle species between restored prairie sites and flowerbeds across the Grinnell College campus. Plant species composition will undoubtedly be very different between these two types of sample sites, with the resorted prairie being a ‘naturally developing ecosystem’ while the flowerbeds, aesthetically designed by landscapers, will not result in the same complex plant community structure seen in prairie systems. Theoretically, ground beetle assemblages will differ between the sites as a result of the differing plant community compositions, which may have future implications concerning the importance of flowerbed arrangements towards maintaining arthropod diversity and abundance. If we are concerned about the decline in arthropod biodiversity, even small-planted plots in urbanized areas can affect insect communities and their conservation.

Reference:   Schaffers AP, et al. 2008. Arthropod assemblages are best predicted by plant species composition. Ecology 89:782-794

Getting Around on a College Campus: Sapling Dispersal and Spatial Patterning

A picture of my team’s study site. Flags mark the outlying corners of the study site. The site sits between two academic buildings at Grinnell College and is planted with shade trees, conifers, and shade dwelling ground covers. Emerging saplings are left in the site by the landscapers and so pose an excellent opportunity to observe tree dispersal on campus.

An individual’s dispersal ability greatly affects its chances of success. Animals move when habitat proves unsuitable.  Plants do not have this luxury and must establish in a single location. Thus it follows that plants would be more sensitive to microhabitats, or variation on a smaller spatial scale, and that this would affect their spatial arrangement in a site. Such spatial arrangements would also be sensitive to dispersal barriers, requiring plant communities to develop another way to disperse offspring into suitable microhabitats or tolerate a higher density of individuals within the currently occupied space. The resulting separation between populations stems from a larger issue of isolation. This has particular consequences for restoration efforts within the urban setting because of the extreme fragmentation of natural areas and the barriers created by manmade structures.

Recruitment from established source populations often aids the succession of restoration and reclamation projects.  However, this process requires source communities to be spatially close to the project to work, a frequent limitation in urban settings where intervening buildings and pavements hamper dispersion. Robinson and Handel (2000) experimented with artificial habitat islands in an attempt to attract dispersal agents, namely frugivore birds, to the region. By attracting dispersal agents, established populations could overcome dispersion barriers and influence the spatial patterning of the site, e.g. succession from bare ground to woodland. With a total of 24 plots, they planted varying sized, native trees and shrubs in an old, barren landfill in New Jersey, and kept track of the population over several years. While they examined several aspects, they explicitly analyze the effect of plant size, finding that larger plants had an initial advantage in plant recruitment, but that this advantage diminished over time. While these results suggest smaller starting plants would provide the best benefit for the cost of restoration, they additionally note recruitment in seriously isolated urban settings may never rise to the species diversity levels associated with historic communities.

While my team’s project does not look at a patch undergoing primary succession, we are looking at the effect of microhabitat on dispersion, particularly of saplings. Understanding source population influences and animal dispersal agents will underscore the processes of secondary succession occurring in the site and will inform the interpretation of its spatial patterning. Because we are in an urban setting, we may have some of the issues of isolation. The adult trees in the site may contribute the most to new individuals. However, nearby green spaces may act as additional source populations. We focus on sapling abundance in relation to microhabitat (overhead cover and slope steepness) in our study. If microhabitat works in concert with dispersion to influence spatial patterning of saplings, this could have implications for other human landscape management.

For the full article by Robinson and Handel (2000):

http://www.esajournals.org/doi/full/10.1890/1051-0761%282000%29010%5B0174%3ADSPORD%5D2.0.CO%3B2

Figure 2

 (a) A diagram of the experimental study site of Robinson and Handel (2000) shows the spatial pattern of plots. Each shaded box represents one of the experimentally planted plots, whereas the open and dashed boxes represent the plots that tested for the effects of fencing in attracting birds. Open boxes have fencing but no intentional plantings, whereas dashed boxes have no fencing or intentional plantings. (b) Shows an up close diagram of planting patterns. The tilled strips were used to evaluate the dispersion of species from the planting plots to outlying areas, but are not discussed in this blog.

The Dangers of Being Edgy

Agriculture and settlement in the Midwest almost destroyed the entire tallgrass prairie ecosystem in less than a century; only one percent of the original extent exists today. Intact prairie, though it looked monotonous to early settlers, is a highly diverse system; this diversity is threatened by the intense fragmentation of the prairie ecosystem. Breaking an ecosystem into small tracts is inevitably accompanied by species loss. One important cause for this species loss is competition with invasive species and unwanted weeds; because patches with small surface area have a higher edge to area ratio, small plots of prairie are particularly vulnerable. Edges help disperse invasive plants and are often quite “weedy,” providing the perfect launching site for weeds to invade prairie patches. Understanding these edge effects, therefore, is an essential for conservation efforts. These efforts extend to reconstructing prairies of various size scales. In particular, small patches of planted tallgrass prairie have begun to appear in more urban landscapes, including on Grinnell College’s campus in Grinnell, IA.

Understanding edge effects is crucial in order to make such small-scale, urban reconstructions a success – diverse patches that can, hopefully, be expanded and can at least demonstrate the beauty of the prairie ecosystem. Helen I. Rowe, Joseph Fargione, and Jeffery D. Holland’s “Prairie Restorations can protect Remnant Tallgrass Prairie Plant Communities” (2013) studies edge effects in a new, important way by quantifying differences between types of edges (different types of bordering land). Rowe et al. (2013) wanted to quantify the impact of different edges because reconstructed prairies are often used as a buffer to protect remnant prairies from damaging invasive weeds. In this study, four remnant prairies adjacent to four different types of lands (roads, crops, high diversity reconstructions, and low diversity reconstructions) were studied; the authors quantified edge effects by measuring non-native dominance (how many of the plants in a 0.5x0.5m quadrat were weeds) along a transect that stretched from the edge of the remnant into the interior of the remnant. Rowe et al. (2013) found that nonnative dominance was higher on the edges than in the interior of the prairies adjacent to crops and roadsides; reconstructed prairie served as a good buffer to weeds because there was no significant edge effect on nonnative dominance in remnants next to reconstructions (both high and low diversity). Therefore, small patches of prairie adjacent to non-prairie edges are vulnerable to edge effects but such risk can be diminished, and diversity maintained in remnants, by providing a reconstructed prairie buffer zone. 

Figure 1. Rowe et al (2013) found that remnant prairies near roadsides and cropland demonstrated much larger edge effects (higher levels of nonnative dominance) than prairies bordered by prairie reconstructions (both high and low diversity).

How does reconstructed prairie buffer zones translate to reconstructed urban prairie patches in Grinnell, IA? In an effort to increase native plantings, Grinnell College’s campus is dotted with small reconstructions particularly vulnerable to edge effects. The narrow strips of prairie surrounding the college’s athletic fields are mainly edge; the patches are bordered by either railroad or concrete paths. With so much edge, can these prairies become viable, diverse reconstructions? Rowe et al.'s (2013) findings suggest these narrow prairie strips surrounded by weedy edges would be nonnative dominated and low in native diversity. We hope to investigate this issue, addressing the feasibility of converting small bits of the urban landscape into a healthy tallgrass prairie ecosystem. In doing so, we also plan to quantify the differences between railroad or path edge effects. Rowe et al. (2013) not only highlight some of the issues of our urban prairie patches with differing edges, but their study also informs our methods and analysis. Their use of slopes to compare differing edges (shown in figure 1) could serve as a useful tool for comparing our different edges’ effect on prairie “weediness.” In this way, we hope to use the lessons and methods presented in Rowe et al. (2013) in order to quantify edge effects in small prairie reconstructions to inform urban conservation efforts.

Toni (one of my colleagues) and I sampling a narrow piece of reconstructed prairie near the athletic fields at Grinnell College, IA. For perspective, the picture was taken on the path-edge of the prairie and you can see the railway edge

Literature Cited:

Rowe, H.I., J. Fargione, and J.D. Holland. 2013. Prairie restorations can protect remnant tallgrass prairie plant communities. The American Midland Naturalist 170: 26-38