Restoration of Sandhill Ecosystems

Tabitha Petri Mentor: Dr. Cynthia Bennington Restoration of sandhill ecosystems: The importance of soil fungi on the growth of a native perennial, Liatris tenuifolia Abstract Possessing a highly diverse understory, sandhill ecosystems are found on dry ridges of the southeastern United States. These systems are highly fragmented due to lumbering, agriculture, and urban development; such practices have significantly impacted soil traits including nutrient levels, pH and microbial communities. The Volusia Sandhill Ecosystem (VSE) was established as a small-­‐scale restoration project in a degraded 2-­‐hectare site on the perimeter of Stetson University’s campus. There we tested how the introduction of soil microbes from an established sandhill ecosystem affected the growth of shortleaf gayfeather (Liatris tenuifolia). We collected soil from the rhizosphere of individual plants in Heart Island Conservation Area (HI), DeLeon Springs, FL, an established sandhill. Half of the soil from Heart Island was treated with fungicide (Daconil®). Two hundred seedlings were divided evenly into four treatment groups: a) untreated HI soil, b) HI soil treated with fungicide, c) VSE soil inoculated with untreated HI soil, and d) VSE soil inoculated with HI soil treated with fungicide. We measured the length of the longest leaf and the total number of leaves over thirteen weeks. The addition of native (HI) soil to VSE soil had no significant effect on number or size of leaves (p>0.15). However, plants grown in native soil treated with fungicide had fewer and smaller leaves than those grown in untreated native soil (p<0.0001). Yet, the inoculation of degraded soil with this native soil is ineffective at improving seedling growth, as other soil characteristics (e.g. pH and nutrient levels) may have a greater effect. Introduction Restoration ecology is a subset of ecology that focuses on the restoration of native areas that have been damaged by unsustainable human activity across the globe. Restoration ecologists work to repair a balance with nature, such that humans function as a constructive part of the ecosystem (SER, 2004). However, restoring these areas is not as easy as removing the non-­‐native species presently in the area and replacing them with natives. First, the “original” ecosystem that should be restored must be determined; the large scale of geologic time makes this decision highly subjective to the inclinations of those leading and funding the restoration. In North America, the primary time period that managers restore their sites to is the early 1800s, just prior to European-­‐American settlement. This time period allows for close climate conditions to the present, while still evoking a more “natural” feeling by glimpsing what the area may have looked like without human interference (Allison, 2004). However, not all sites can be repaired to their original conditions. Novel restoration sites are places so different from their original system that trying to regain the initial conditions is unrealistic. This is often the case with abandoned open pit mines, capped landfills, dredges spoiled islands and abandoned urban lots (“Reference”). Such sites often have highly degraded soil, a lack of wind or sun protection for plants, and potentially a long history of non-­‐native species competition, all of which shape the ways this site will react to given management efforts. Thus when performing restoration in urban sites, it is important to ensure that restoration is feasible and if so to which period of time. Furthermore, the environment, soil composition, surroundings, and organisms that natives will have to compete with most likely differ from those that the natives in the “original” ecosystem previously confronted. All of these factors alter the food web, hydrology, plant-­‐soil interactions, and other functions of an established ecosystem. Ecologists restoring an area must find ways to aid in or completely substitute these functions during the restoration process, particularly during the early stages of restoration when the environment will be most different from its established state. For instance, Rowe et al. (2007) found that phosphorus and mycorrhizal fungi both helped natives in a restoration area to outcompete the invasive species presently dominating the area. Water distribution also affects all stages of plant life; therefore, moisture gradients and the impact that changes in elevation may have on the native species must be gauged for each individual site (Peet & Allard, 1994). Desired native seeds often require direct management efforts in restoration sites to ensure germination and establishment. Many seeds need scarification to germinate but introducing fire or waiting for animals to digest the seeds are not realistic circumstances in a restoration situation; as such, artificial alternatives must be implemented to ensure timely germination (Perez-­‐Garcia & Gonzalez-­‐Benito, 2005). Additionally, the seed bank in a degraded site, especially one with with altered ground cover, may not be helpful for restoring the desired end ecosystem; Buonopane et al (2013) found that seed banks provided only minor contributions to post-­‐disturbance understories, especially in sites were leaf liter was cleared away (Buonopane et al, 2013). For these reasons, managers may alternatively choose to transplant seedlings or adult plants rather than sow seeds during the first years of restoration. Therefore, special accommodations may need to be taken in the early years of projects to ensure that natives can grow in the proper above-­‐ and below-­‐ground conditions. While easily overlooked in restoration efforts, proper soil communities are crucial for the long-­‐term survival and growth of native seedlings. Soil contains a vast assortment of life such as fungi, bacteria, protozoans, nematodes, earthworms, and arthropods. Microbial species help with nutrient recycling, plant community structure, regulation of plant productivity, and decomposition of organic matter (Herzberger et al, “Bouncing”, 2014). Estimates indicate that around 50,000 fungal species form mycorrhizal associations with over 250,000 plants; these fungi can provide up to 80% of a plants nitrogen and phosphorus requirements (Van der Heijden et al, 2015). Both beneficial and parasitic fungi exist and these relationships vary by the combination of plant-­‐soil species; even a single species of fungi have been shown to affect different plant species in different ways (Bever, 1994). Soil fungi can develop symbiotic relationships with plants wherein they increase the surface area of the roots by attaching to the rhizosphere, allowing for a more efficient uptake of nutrients and water. However, in return for their help fungi take sugars from the plant, which means that if a plant already has adequate access to resources without assistance the relationship becomes parasitic. Thus, experiments that transfer soil from one place to another must ensure that the target plant species and abiotic soil conditions are similar enough to those in the original environment to achieve the same plant-­‐soil relationships at a new site. Additionally, soil biota are sensitive to changes in location and atmospheric composition; if soil is transported it should travel the least distance possible (Nie et al, 2013). When restoring soil, there will be tradeoffs occurring in any decision and this process may involve some trial and error, but it is worth the investment to determine the best combination of plant and soil communities for each ecosystem. Previous restoration sites have shown varying success with the altering of the microbial community depending on the age of the plants and the ecosystem type. While fungal associations may lead to intraspecific competition between plant species in later life stages because of limited resources, these fungi tend to be highly beneficial in the early life stages for providing necessary water and nutrients while root systems are still maturing. In the early stages of community development, positive feedback loops allow a few species to maximize their uptake of nutrients, accelerating their growth (Herzberger et al, “Plant”, 2014). Ecosystem characteristics also interact with the soil microbes and in turn affect plant growth. Conrad and Segraves (2013) found that mycorrhizal colonization decreased in a scrub community as the time since fire increased. This suggests that in restoration sites of fire-­‐adapted communities a lack of fire may cause a shortage of soil fungi. This decrease has also been explained as plants relying more heavily on fungal associations during early developmental stages until their root system is mature enough to collect adequate resources on its own (Friese and Cipollini, 2010). Thus, soil microbes develop specialized relationships with plants that can help accelerate growth in disrupted environments in conjunction with other belowground factors. Sandhills are dry, fire-­‐adapted areas established on ancient sand dunes across the southeastern United States. Soils in these ecosystems tend to have low levels of pH and cation exchange capacity, as well as low levels of organic matter and nutrients (Anderson & Menges, 1997). Sandhills rely on fire to create layers of ash that continually replenish nutrient levels and limit above ground competition. Below ground symbiotic relationships are particularly important in these low resource environments because roots alone can be insufficient for gathering all the necessary water and nutrient requirements. Liatris tenuifolia is a sandhill perennial of the Asteraceae family. Its common name, Shortleaf Gayfeather, refers to the vibrant pinkish purple color of the individual flowers. This species is widespread in southeastern United States upland ecosystems. L. tenuifolia produces large (1 to 2 meters) flowering stalks in the fall and thin leaves that start forming shortly after germination. L. tenuifolia grows from a spherical corm that is several centimeters below ground (Anderson & Menges, 1997). Yet, the expanse of roots that form relationships with the soil microbes are at shallower levels allowing for easier access and manipulation. Additionally, Anderson and Menges (1997) found that Liatris tenuifolia formed some level of mycorrhizal associations under both burned and unburned conditions. These factors make Liatris tenuifolia a prime test subject for exploring the importance of fungal associations in sandhills. Our experiment asked two questions relative to our established sandhill ecosystem and our early-­‐stage restoration sandhill ecosystem: 1) If soil from an established sandhill ecosystem contains beneficial fungi, then Liatris tenuifolia seedlings grown in the absence of those fungi will be smaller than those grown with an intact fungal community 2) If the growth of Liatris tenuifolia in a restoration sandhill site is limited by insufficient fungal communities, than inoculation by soil from an established sandhill should increase growth. Methods Study Sites-­‐-­‐ My experiment was carried out in the Volusia Sandhill Ecosystem Teaching Landscape on the Stetson University campus in DeLand, FL. This project aimed to take a 2-­‐hectare lot, consisting of mainly invasive weeds and turf grass, and recover the sandhill ecosystem that once existed here. Unfortunately, the present urban location will prevent the site from being able to support large animals and function as a complete sandhill ecosystem; however, it has a lot of potential as an educational site to help members of the community understand how sandhill ecosystems function and why these systems are important. The Volusia Sandhill Ecosystem was established in the summer of 2011 with the planting of 80 upland trees, primarily longleaf pines. Since then multiple students, service groups, and other community partners have helped add understory plant species. As of May 2014, the site included a pollinator garden, wiregrass area, and native seedbeds. All of this work has restored about one-­‐eighth of the property to date. Future goals include adding interactive activities and a native species kiosk throughout the garden to guide and education guests (“Sandhill”, 2015). Research and experiments are still needed to find ways to aid natives in outcompeting invasive species, like Bahia grass (Paspalum notatum). To gauge the success of our restoration efforts, we located a sandhill ecosystem that was degraded in similar ways to ours but further along in its restoration process. Heart Island Conservation Area in DeLeon Springs, FL is the established sandhill restoration area we used as a comparison for our site. The St. Johns River Water Management District obtained this 14,246-­‐acre area in 1994 and began by replanting longleaf pine trees, as the previously forest had been largely lumbered. This conservation area is called an ‘island’ because it is a fragment of the 35,380-­‐acre Lake George Wildlife Management Area (“Heart”, 2016). Experimental Methods-­‐-­‐ We collected soil from both Heart Island Conservation Area (HI) and the Volusia Sandhill Ecosystem Teaching Landscape (VSE). Samples taken at Heart Island were from around the roots of L. tenuifolia by digging up a total of 4L of soil from 10 to 15cm below ground. Half of the soil from HI was treated with a concentration of 200 ug of fungicide (Daconil®) per 1 gram of soil (Aziz et al, 1991). Daconil has the active ingredient chorothalonil, which has been shown to limit fungal activity without altering nutrient levels in the soil (Chen et al, 2001). The treated HI soil sat undisturbed for one week, so the fungicide had time to kill the fungi and to limit the likelihood of high fungicide levels directly harming the seedlings. Soil samples from VSE were collected from the same depth throughout the experimental plot area of our sandhill. The diagram below illustrates my four treatment groups, note that dots represent fungal presence. The two pots on the left represent the first hypothesis about the established sandhill. The leftmost pot represents Heart Island soil with fungi (untreated soil). The second pot represents Heart Island soil treated with fungicide; thus why there are no dots present. The two pots on the left represent the second hypothesis concerning the Volusia Sandhill Ecosystem. All VSE soil was untreated because there is no management scenario where adding fungicide to an entire landscape would be a viable option; thus, the point of inoculation is to determine if native fungi can outcompete or work around the present degraded fungal community in VSE soil. The pot on the left shows VSE soil being inoculated with HI soil containing fungi; the rightmost pot shows VSE soil being inoculated with HI soil treated with fungicide. Figure 1: Dots indicate fungal presence. The leftmost pot represents HI soil untreated. The next pot represents HI soil that was treated with fungicide. The third pot represents VSE soil that was inoculated with HI untreated soil. The rightmost pot represents VSE soil that was inoculated with HI soil treated with fungicide. My experiment divided two hundred seedlings evenly into four treatment groups: a) untreated HI soil, b) HI soil treated with fungicide, c) VS soil inoculated with untreated HI soil, and d) VS soil inoculated with HI soil treated with fungicide. Each treatment group consisted of 10 pots holding five seedlings each in; each pot was filled to one-­‐inch below the top with 1200mL of soil and the two VSE treatment groups were inoculated by the direct addition of 108mL of HI soil. The inoculation amount was base on Middleton and Bever (2012) using 9% by volume to inoculate their treatments. We germinated the L. tenuifolia on filter paper in petri dishes under 24-­‐hour fluorescent light at room temperature prior to experimentation. L. tenuifolia has had a low germination rate in our sandhill in the past, so using seedlings ensured that we would have a large enough sample size for reliable statistical analysis. The seedlings were grown in potting soil in a greenhouse for three weeks before being transplanted to further ensure viability. During the experiment, the seedlings were potted and grown out in the Volusia Sandhill Ecosystem to ensure realistic photoperiod and temperature. Each of the 40 pots was randomized within a 10-­‐by-­‐4 grid to ensure that the curvature of the sandhill or patchy canopy coverage did not confound the results. The pots were placed 0.25meters apart in the ground such that the level of soil in the pot was even with the level of soil in the sandhill. We placed one seedling in the middle of the pot and the other four evenly around the outside; seedlings were identified by marking the topmost seedling with a stake as seedling 1, then following a clockwise pattern ending with the innermost seedling as seedling 5. Since the greenhouse the seedlings were germinated in contained sprinklers, all seedlings were watered four times a week throughout the first six weeks of the experiment to limit transplant shock. Each week up to six weeks after transplant, we measured the length of the longest leaf and the number of leaves of each seedling. We measured these factors once more at week thirteen to gain a better understanding of the trends of each treatment. Averages were then taken for the length of the longest leaf and average number of leaves for each treatment. These averages were analyzed using ANOVA in JMP (SAS Institute, Cary, NC) to determine if there was an interaction between fungal presence and growth. Results Differences among treatments in the growth of Liatris tenuifolia seedlings were initially small and insignificant, but they increased over the 13 weeks of the experiment (Figures 2 & 4). At the end of the experiment, thirteen weeks after transplanting into the sandhill, there was a significant difference between Heart Island soil untreated (with fungi) and Heart Island soil treated with fungicide for the length of the longest leaf (p<0.0001, Figure 3i) and the average number of leaves (p<0.0001, Figure 5i). However, when seedlings were grown in Volusia Sandhill Ecosystem soil that had been inoculated with Heart Island soil with or without fungi present, there was no effect on seedling growth. At week 13, neither the length of the longest leaf (p=0.65, Figure 3ii) nor the average number of leaves (p=0.15, Figure 5ii) differed between the Volusia Sandhill Ecosystem treatments. Initial measurements start one week after transplanting to ensure that any transplant shock did not confound our calculations. Figure 2: Average length of the longest leaf on Liatris tenuifolia (n=200, 50 per treatment). There were no differences in length of the longest leaf among any treatment group prior to week 6 (p>0.05 in all cases). There were no differences in survival at any time (p>0.05 in all cases). Figure 3: Average length of the longest leaf in HI and VSE treatments, respectively. i) Seedlings grown in HI soil treated with fungicide had significantly shorter leaves than those grown in untreated HI soil (p<0.0001). ii) There was no significant difference in the length of the longest leaves between VSE treatments (p=0.65). Figure 4: Average number of leaves on Liatris tenuifolia (n=200, 50 per treatment). There were no differences in the average number of leaves among any treatment group prior to week 6 (p>0.05 in all cases). There were no differences in survival at any time (p>0.05 in all cases). 3i 3ii Figure 5: Average number of leaves in HI and VSE treatments, respectively. i) Seedlings grown in HI soil treated with fungicide had significantly less leaves than those grown in untreated HI soil (p<0.0001). ii) There was no significant difference in the number of leaves between VSE treatments (p=0.15). Discussion My results suggest that fungi play an important role in promoting seedling growth in established sandhill soil. In our established sandhill, Heart Island Conservation Area, the introduction of fungicide had a large negative effect on the overall growth of Liatris tenuifolia. One explanation for this, is that our initial rationale was correct-­‐ native fungal presence is necessary for increasing seedling growth. A second explanation could be that the fungicide directly harmed the seedlings in that treatment rather than only killing the fungi. However, we do not think this was the case for a number of reasons. First of all, this fungicide and concentration was used in a previous study by Aziz et al (1991) and did not harm their plants. Secondly, the difference between the Heart Island treatments increased over time for both measurements. If the fungicide was harming the plants, we would expect the effect to be greatest when the plants were smallest and before the active ingredient, chlorothalinol, was leaked and/or degraded since the half-­‐life is 10-­‐60 days (Syngenta Inc). Also, we dug up some specimens from each treatment and there were no visible impairments to the roots or corm in this 5i 5ii treatment. Lastly, there was no significant difference in survival between the treatments at any time and survival rates for all treatments were above 78% at week 13. Thus, our results support our initial hypothesis that in an established sandhill seedlings grow bigger in the presence of native fungi. The test of my second hypothesis revealed that the inoculation of degraded soil with native soil was ineffective at improving seedling growth. One explanation for this was that there was not enough HI soil used to inoculate the VSE treatments. This is unlikely, because the inoculation level of 9% by mass of the soil was used by Middleton and Bever (2012) and was effective. However, their study did have a longer running time, so it is possible that the native fungi did not have long enough to reproduce and establish themselves. Another reason for this result could be that the large proportion of non-­‐native fungi present in the VSE soil inhibited the growth of our native plant species. Vogelsang and Bever (2009) found that native plant species had less growth when planted in soil conditioned by a mix of non-­‐native plant species. This result supports findings of previous studies that large areas dominated by non-­‐native plant species are in some way resistant to reinvasion by native plant species (Vogelsang and Bever, 2009). The lack of difference between the VSE treatments could then by explained as the soil fungi from VSE dominating the soil or outcompeting the additional fungi from HI. A third explanation could involve other soil limitations of the Volusia Sandhill Ecosystem. As an early stage restoration site, the soil is degraded in many ways and some of these other characteristics may be more limiting than the presence of soil fungi. The VSE has high levels of nutrients and pH compared to a more established ecosystem like Heart Island (Figure 6). Previous research in the VSE site showed that pH had a significant effect on the growth of Carphephorus and Pityopsis, two other sandhill perennials, so it is very possible that it impacts the growth of Liatris as well (Rivera, 2014). Figure 6: pH and nutrient concentrations for Heart Island Conservation Area (HI) and the Volusia Sandhill Ecosystem (VSE), levels determined by University of Florida soil laboratory. Additionally, Longleaf pine (Pinus palustris) seedlings, a dominant sandhill species, are known to increase root density as competition for water and nutrients increases (Brockway & Outcalt, 1998). Yet, in a site like VSE, where there are elevated levels of water and nutrients, survival pressures may not be strong enough to maximize below ground resources. As noted in the introduction, Anderson and Menges (1997) showed that Liatris tenuifolia invests in root hairs and mycorrhizal associations regardless of the whether soil in a healthy uplands ecosystem was burned or unburned. This would support the idea that L. tenuifolia should form fungal associations at a wide range of soil nutrient abundances. However, since the VSE soil is degraded such that its nutrient levels are significantly higher than even healthy soil post-­‐burning, the seedlings may have been able to support themselves with root hairs alone. Also, some studies suggested that soil communities play an important role in the spatial distribution of plants; however Kalisz and Stone (1984) did not find enough evidence of differences in soil composition for that to be determining the spatial patterns of longleaf pines in Ocala National Forest. This study did not try to describe the influence of the soil community as insignificant, but rather discussed the idea that an interaction between microbes and other below ground factors, such as nutrient levels, more strongly determines community distribution (Kalisz & Stone, 1984). This same overshadowing by additional soil factors could be attributed to our lack of reaction to the inoculation of native fungi. Thus, while our inoculating degraded soil did not improve seedling growth, our research does demonstrate that in a healthy, sandhill ecosystem the presence of native soil fungi does significantly improve growth. Therefore, as our restoration efforts proceed, we must continue to be aware of biotic and abiotic soil parameters to ensure that healthy, appropriate soil biota are establishing themselves. Literature Cited Allison, S. K. (2004). What do we mean when we talk about ecological restoration? Ecological Restoration, 22(4), 281-­‐286. Anderson, R. C. & Menges, E. S. (1997). Effects of fire on sandhill herb: nutrients, mycorrhizae, and biomass allocation. American Journal of Botany, 84(8), 938-­‐ 948. Aziz, T., Habte, M., & Yuen, J.E. (1991). Inhibition of mycorrhizal symbiosis in Leucaena leucocoephala by chlorothalonil. Plant and Soil, 131(1) 47-­‐52. Bever, J. D. (1994). Feedback between plants and their soil communities in an old field community. Ecology, 75(7), 1965. Brockway, D.G. & Outcalt, K.W. (1997). Gap-­‐phase regeneration in longleaf pine wiregrass ecosystems. Forest Ecology and Management, 106(1998), 125-­‐139. Buonopane, M., Snider, G., Kerns, B.K., Doescher, P.S. (2013). Complex restoration challenges: weeds, seeds, and roads in a forested Wildland Urban Interface. Forest Ecology and Management, 295(2013) 87-­‐96. Chen, S.K., Edwards, C.A., and Subler, S. (2001). Effects of fungicides benomyl, captan and chlorothanil on soil microbial activity and nitrogen dynamics in laboratory incubations. Soil Biology and Biochemistry, 33(2001) 1971-­‐1980. Conrad, A. O., & Segraves, K. A. (2013). Mycorrhizal colonization of Palafoxia feayi -­‐4 (Asteraceae) in a pyrogenic ecosystem. Mycorrhiza, 23(3), 243-­‐249. Friese, C. & Cipollini, D. (2010). Arbuscular mycorrhizal fungi protect a native plant from allelopathic effects of an invader. Journal of Chemical Ecology, 36(4), 351-­‐360. “Heart Island Conservation Area.” St. Johns River Water Management District. Bureau of Land Management, 2016. Web. 22 Feb. 2016. Herzberger, A. J., Duncan, D. S., & Jackson, R. D. (2014). Bouncing Back: Plant-­‐ Associated Soil Microbes Respond Rapidly to Prairie Establishment. Plos ONE, 9(12), 1-­‐14. Herzberger, A. J., Meiners, S. J., Towey, B., Butts, P. A., & Armstrong, D. L. (2014). Plant-­‐mircobe interactions change along a tallgrass prairie restoration chronosequence. Restoration Ecology, 1-­‐8. Kalisz, P. J., & Stone, E. L. (1984). The longleaf pine islands of the Ocala national forest, Florida: a soil study. Ecology, 65(6), 1743-­‐1754. Middleton, E.L., and Bever, J.D. (2012). Inoculation with a native soil community advances succession in a grassland restoration. Restoration Ecology, 20(2), 218-­‐226. Nie, M., Pendall, E., Bell, C., Gasch, C. K., Raut, S., Tamang, S., & Thrall, P. (2013). Positive climate feedbacks of soil microbial communities in a semi-­‐arid grassland. Ecology Letters, 16(2), 234-­‐241. Peet, R. K., & Allard, D. J. (1994). Longleaf pine vegetation of the southern Atlantic and eastern gulf coast regions: a preliminary classification. Tall Timbers Fire Ecology, 45-­‐81. Perez-­‐Garcia, F., & Gonzalez-­‐Benito, M. (2005). Seed germination of five Helianthemum species: Effect of temperature and presowing treatments. Journal Of Arid Environments, 65(4), 688-­‐693. “Reference Sites in Ecological Restoration.” University of Washington, n.d. Web. 23 Apr. 2015. Rivera, D. 2014. Acidic Soils Benefit Seedling Growth in Two Perennial Plant Species Native to the Florida Sandhill Ecosystem. Research Report, Stetson University Department of Biology. Rowe, H. I., Brown, C. S., & Claassen, V. P. (2007). Comparisons of Mycorrhizal Responsiveness with Field Soil and Commercial Inoculum for Six Native Montane Species and Bromus tectorum. Restoration Ecology, 15(1), 44-­‐52. "Sandhill Ecosystems." Gillespie Museum. Stetson University, 2015. Web. 26 Mar. 2015. SER (2004). The SER Primer on Ecological Restoration, Verson 2. Society for Ecological Restoration Science and Policy Working Group, 2015. Web. 26 Mar. 2015. Van der Heijden, M.G.A., Martin, F.M., Selosse, M., and Sanders, I.R. (2015). Mycorrhizal ecology and evolution: the past, the present, and the future. New Phytologist, 205(4), 1406-­‐1423. Vogelsang, K.M., and Bever, J.D. (2009). Mycorrhizal densities decline in association with nonnative plants and contribute to plant invasion. Ecology, 90(2), 399-­‐ 407.