An Adaptive Management Plan for Restoring a Degraded Oak Barren in West Michigan.

Abstract

Oak barrens are important ecosystems because they provide habitat for rare and threatened species. After European settlement, most barrens were destroyed. Seidman Park in Ada, Michigan is home to a degraded oak barren.  I hypothesize that with proper management, the barrens can be restored to a high quality oak barren. Data was collected on current vegetation composition of Seidman Park and a reference site behind the Lowell Township Hall in Kent County, Michigan. This adaptive management plan includes prescribed burning, invasive species cutting, hand seeding of native grasses and forbs, and monitoring of the plan to assess progress. This plan can be implemented to restore the original rare Oak barrens in West Michigan and create habitat for species such as the Karner blue butterfly (Lycaeides melissa samuelis), the eastern box turtle (Terrapene c. Carolina), and numerous graminoids, legumes, and forbs. 

Introduction

Pre-settlement of the Europeans, Native Americans set fire either accidentally or intentionally to oak barrens. These frequent fires paired with drought, windthrow, and frost kept the areas open and functioning. Once settlement occurred the frequent fires decreased due to the clearing of land for sand mining, agriculture, and residential uses. Making matters worse large legendary wildfires spread across the United States in the 1800s. Franklin Hough and Bernard Fernow were conservationists that argued that if fire suppression was not set in place, wildfires would be detrimental to the timber industry. With the timber industry threatened and conservationists pushing the message that watershed conservation was at risk as well, the government was convinced to set aside national forests. Shortly thereafter, the Forest Service was formed and by 1920 complete and total fire suppression policies were set in place (The Forest History Society 2015).

Fire suppression may have been good for timber production; however, it was detrimental to ecosystems such as oak barrens. Oak barrens are savanna type ecosystems that rely on fire and are typically dominated by a 60% or lower canopy layer with mature overstory oaks. The ground layer is comprised of graminoid species that one would associate with forest and prairie communities. Oak barrens support 26 threatened and endangered animal species and 56 threatened and endangered vegetative species such as the Karner blue butterfly (Lycaeides melissa samuelis), the eastern box turtle (Terrapene c. Carolina), three-awned grass (Aristida), and Gattingers Gerardia (Agalinis gattingeri) (Kost et al. 2007). Fire-dependent communities that are deprived of fire negatively impact rare and endangered species that live in that ecosystem. This project aims to understand what kind of management techniques can restore an oak barrens ecosystem in west Michigan.

        I hypothesize that the oak barrens located in Seidman Park, Ada Michigan, can be restored back to its original composition of mature overstory oaks and a diverse herbaceous layer with the right management techniques. This site shows signs that it was originally a functioning oak barrens ecosystem with the presence of indicator species, sandy soils, and canopy gaps. Multiple species of pines were also planted in the barrens which turned this site into a mixture of mature overstory trees, invasive and noninvasive midstory bushes, native forbs, and invasive grasses. My adaptive management plan will include data from Seidman Park and a reference site in Lowell Michigan and recommend management practices to restore the altered oak barren community into a functioning ecosystem.

Methods

        This study took place at Seidman Park in Ada Michigan, and at the surrounding land behind the Lowell Township building. Seidman Park holds several different ecosystems including what used to be a functioning oak barren. Historically, this area was comprised of black oaks (Quercus velutina), white oaks (Quercus alba), little blue stem (Schizachyrium scoparium), New Jersey tea (Ceanothus americanus), lupine (Lupinus perennis), huckleberry (Vaccinium membranaceum), switch grass (Panicum virgatum), and additional indicator species of an oak barren (Kost et al. 2007). Currently, Seidman Park exhibits oak barren indicator species, Oakville fine sands and open grown, mature oaks. Several invasive woody species, grasses, and forbs have encroached on the area including Black cherry (Prunus serotina), Jack pine (Pinus banksiana), White pine (Pinus strobus), White spruce (Picea glauca), Kentucky bluegrass (Poa pratensis), and spotted knapweed (Centaurea maculosa).

The reference site for this restoration plan is the area behind the Lowell Township Hall near Lowell, Michigan.  This area is comprised of Plainfield sand, a varying canopy cover including areas of high canopy cover and areas with none to create a savanna structure. This area is also composed mostly oak barren indicator species including black oaks, white oaks, three-awned grass, switch grass, and little blue stem (Schizachyrium scoparium), making this the best option of a reference site (Figure 1).

        I established 10 plots at Seidman Park and 10 plots at the barren behind the Township Hall for vegetation sampling. These plots were randomly distributed at each site and measured 10 x 10 m in size.  5 x 5 m subplots were also created and located in the northwest corner of each plot 10 x 10 m plots created. Trees were measured in the 10 x 10 m plot and recorded if they had a diameter at breast height (dbh) greater than 10 cm using a dbh tape. Within each 10 x 10 m plot, the percent vegetation cover of all ground cover in the plot was determined by visual estimation. Canopy cover percentage was determined using a spherical densiometer.

        In each 5 x 5 m plot, saplings were recorded at a dbh ranging from 2.5 cm to 10 cm and seedlings were recorded at a dbh less than 2.5 cm. Forbs and graminoids indicative of oak barrens including switch grass, three-awned grass, huckleberry, lupine, little blue stem, milkweed (Asclepias), and reindeer lichen (Cladonia rangiferina) were identified and recorded as occurring within the plot or not occurring.

Results

        The average number indicator species and ground cover species were higher in the reference site than in the management site. (2.1 indicator species for Seidman Park vs. 3.2 indicator species for Lowell Township. 61.5% ground cover in Seidman Park vs. 82% ground cover in Lowell Township; Table 1). Seidman Park, or the management site, has a higher number of average seedlings, saplings, trees, invasive species, and canopy cover.  Seidman Park has 2.4 trees per plot, 1.9 saplings per plot, 10.4 seedlings per plot, 46.8 % canopy cover and 1.9 invasive species per plot. This is comparable to the reference sites 0.8 trees per plot, 0.6 saplings per plot, 7.4 seedlings per plot, 28.94% canopy cover, and 0.9 invasive species per plot (Table 1). The management site has 5 non-oak species whether they are seedlings, saplings, or trees (Table 2). It also has 5 invasive species present and 6 indicator species (Table 3). The relationship between the number of trees, saplings, indicator species, and invasive species per plot in each location shows a significant difference with a p-value of <0.05 (Figure 2).

        The correlation between trees and indicator species is a negative correlation with an r-squared value of 0.6878 (Figure 3). The positive correlation that was found was between the number of invasive species and trees in general per plot with an r-squared value of 0.6721 (Figure 4).  There was also a positive correlation shown between the non-oak species per plots and the invasive species present that is shown with an r-squared value of 0.4773 (Figure 5). The last negative correlation that showed importance was one between invasive species per plot and indicator species per plot with an r-squared value of 0.4773 (Figure 6).

        The correlation between seedlings and saplings and the number of invasive species present was positive as well. Following the same trend the correlation between seedlings and saplings and indicator species was negative. The negative correlation was present as well between canopy cover and indicator species. However, these correlations had an r-squared value of close to 0 which indicates a trend, however not an overly strong one.

Discussion

        The results of this study propose that the number of invasive species and indicator species present in an area depends directly on the amount of mature overstory trees present, the number of seedlings and saplings able to thrive in the area, the ground cover amount, and the canopy cover amount. Invasive species were higher in the areas that had a lower ground cover, higher canopy cover, more seedlings and saplings, and more mature trees. Vice versa, indicator species thrived in areas that had a high ground cover amount, low canopy cover, few trees, and a limited number of seedlings and saplings. To sum this up the more trees, seedlings, saplings, and canopy cover was present per plot the more invasive species there were per plot and the less indicator species per plot.

The big issue to be focused on however is the number of trees present. The correlation between the number of trees per species and the number of indicator species and invasive species fit the data trendline with an r-squared value close to 1. The correlation between indicator species and invasive species and the number of seedlings, saplings, and canopy cover had an r-squared value of close to 1. This means that while there was a relationship, it did not fit the trendline as well, showing less importance.

Proposed Management

        These results suggest that trees need to be removed, canopy cover needs to be reduced, ground cover needs to be increased, and fewer seedlings and saplings need to be given the chance to survive. Lowell township had only mature overstory white oaks while Seidman park had mature red maples, red oaks, white pine, autumn olive, black oaks, and jack pines (Table 2). Jack pine seeds stored in the seed bank are fire dependent, which will make the key restoration technique, prescribed fire, a challenge (Herring 1999). However, since it will take a few years for the pines in the seed bank to grow to a height that can be managed for prescribed burning should be the first step.

        A high intensity burn should be implemented in spring and then again in the late season for 2 years in a row. This will lower the canopy cover percentage and hopefully rid of some non-oak species (Haney 2008). After this is conducted, monitoring should occur to assess what the composition is. Pines should be cut down and chipped during the winter, leaving the chips where they are. After the mature pines have been removed, Black cherry trees should be targeted as their berries attract birds that in turn carry with them many seeds and various invasive species.

All other species that are not oaks should be removed as well; however even oaks that are not large and indicative of inhabiting the area for a long period should be removed to lower the amount of canopy cover present. After two years when the progress has been assessed, manual seeding should occur. Seeds collected from the local seedbank from the reference site of Lowell Township should be collected and then spread to encourage growth of forbs, graminoids, and other components of the herbaceous layers. Fire should then be reintroduced to the area every other year (Peterson 2014). It is however hard to say exactly how much fire will need to be introduced within a 10-year period. Studies have shown that it is dependent on the area and the condition it is in. Some barrens need 4 fires/decade, and others need up to 6 fires/decade (Haney 2008). This is where the monitoring process takes place.

Monitoring

        While this process seems straightforward it is going to take a lot of work and with that a long monitoring process. All invasive species and tree species that aren’t native to oak barrens are not going to be eradicated immediately, if totally at all. However, to reach these goals a monitoring process of at least 10 years needs to occur. The first 5 years should be spent trying to reduce canopy cover, seedlings, saplings, trees, and invasive species per plot in Seidman Park. The following 5 years should be used to assess the progress and either change routes with more frequent fire, more manual cutting, or possibly other techniques.

Evaluation

        When the monitoring period of at least 10 years has ended there are certain goals and standards that need to be met in order to claim that the management was a success. A canopy cover that is less than 40% or closer to 30%, an invasive species count of 1 or less per plot, an indicator species count of 1 or less, and a tree count of 1 or less would be considered successful.

Conclusion

        The purpose of this management plan is to restore a rare degraded oak barren located in Ada, Kent County Michigan. I hypothesize that the implementation of the proposed management techniques followed by an extensive monitoring process that the restoration of this area to pre-settlement conditions could be successful. The success of this project would mean that there would be more habitats for the 26 threatened/endangered/rare animal species and for the 56 threatened/endangered/rare vegetative species that thrive in oak barren communities. Not only would it create vital habitat but also it would increase the total area that oak barrens span across the Midwest. This plan would give other the opportunity to recreate the same project in other degraded barrens. Lastly, to complete this project successfully would create a means of education. It would allow people to visit a functioning oak barren and learn about their importance to the ecosystem and give them a chance to enjoy the aesthetics.

Figure 1. Study area location maps created with Archmap of the reference site behind Lowell Township Hall and the management site in Seidman Park, both located in Kent County Michigan, 2016.

Figure 2. Average number of species per plot (+/- 1 Standard Deviation) in both Seidman Park and reference site behind Lowell Township in Kent County, Michigan 2016. All relationships are significant with a p-value <0.05.

Figure 3. Linear regression model describing the effect that the number of  trees per plot has on the number of indicator species per plot from data collected at Seidman Park, Ada Michigan 2016. (R²= 0.6878)

Figure 4. Linear regression model describing the effect that the number of  trees per plot has on the number of invasive species per plot from data collected at Seidman Park, Ada Michigan 2016. (R²= 0.6271)

Figure 5. Linear regression model describing the effect that the number of trees that are not oak species per plot has on the number of invasive species per plot from data collected at Seidman Park, Ada Michigan 2016. (R²= 0.4773)

Figure 6. Linear regression model describing the effect that the number of invasive species per plot has on the number of indicator species per plot from data collected at Seidman Park, Ada Michigan 2016. (R²= 0.4773)

Table 1. Averages for data collected at a reference site behind Lowell township hall and a management site in Seidman Park, both located in Kent County, Michigan 2016.

Table 2. Tree, sapling, and seedling species present at Lowell Township, the reference site, and Seidman Park, the management site in Kent County Michigan, 2016.

Table 3. Invasive species and their recorded appearances and also indicator species and their recorded appearances for both Lowell Township, the reference site, and Seidman Park, the management site in Kent County Michigan, 2016.

Literature Cited:

Nielsen S, Kirschbaum, Haney A. 2003. Restoration of midwest oak barrens: structural manipulation or process-only? Conservation Ecology 7(2): 10. [online] Available from: http://www.consecol.org/vol7/iss2/art10/

Haney A, Bowles M, Apfelbaum S, Lain E, Post T. 2008. Gradient analysis of an eastern sand savanna’s woody vegetation, and its long-term responses to restored fire processes. Forest Ecology and Management. [Internet]. [cited 1 Oct 2016]; 256(8). Available from: http://plantconservation.us/Haney2008.pdf

Ralston J, Cook J. 2013. Impact of prescribed fire, timber removal, and the seed bank in understory plant diversity and canopy cover in an oak-pine barrens, central Wisconsin, USA. Ecological restoration. [cited 1 Oct 2016]; 31(4).

Petersen SM. 2014. Effects of biennial fire and clipping on woody and herbaceous ground layer vegetation: implications of restoration and management of oak barren ecosystems. Restoration Ecology. [Cited 1 Oct 2016]; 22(4)

Bried JT, Patterson WA, Gifford NA. 2014. Why pine barrens restoration should favor barrens over pine. Restoration Ecology. [Cited 1 Oct 2016]; 22(4)

Maloney JM. 1997. Oak Savanna Restoration Tips. Student Online Journal. [Internet] [cited 30 Sep 2016]; 2(6). Available from: http://conservancy.umn.edu/bitstream/handle/11299/58861/1/2.6.Maloney.pdf

Brock TD, Brock KM. 2004.  Oak Savanna Restoration: A Case Study. North American Prairie Conference [Internet]. [cited 9 Sep 2016]; 83. Available from: digitalcommons.unl.edu/napcproceedings/83/

Reinhardt JR.  2015. Using Observational, Experimental, and Focal Species Approaches to Inform the Adaptive Management of Oak Savanna Ecosystems in Western Michigan, USA. Retrieved from the University of Minnesota Digital Conservancy. [Internet]. [Cited 1 Oct 2016]; Available from: http://hdl.handle.net/11299/175313

The Forest History Society(US). 2015. U.S. Forest Service Fire Suppression. [Internet] United

States Forest Service History(US); [cited 2016 Oct 7] Available from http://www.foresthistory.org/ASPNET/Policy/Fire/Suppression/Suppression.aspx

Kost MA, Albert DA, Cohen JG, Slaughter BS, Schillo RK, Weber CR, Chapman KA. 2007.

Natural Communities of Michigan: Classification and Description. Michigan Natural Features Inventory, Report No. 2007-21.

Herring D. 1999. Evolving in the Presence of Fire. [Internet] Earth Observatory; [cited 2016 Dec

            12] Available from http://earthobservatory.nasa.gov/Features/BOREASFire/