An Examination of Early Succession at the Indiana Dunes

Examination of Early Succession at the Indiana Dunes

Adrian Raygoza (Senior undergraduate BioS major)

The University of Illinois at Chicago

1200 West Harrison St.

Chicago, Il 60607

Abstract:

This project focused on five different methods of succession. Soil development was investigated by sampling and measuring the percentage of organic matter in the samples collected from the nearest dunes to the furthest from the shore. Diversity was investigated by counting the number of species in five sites. We then calculated the species density in the sites by using a Simpson’s diversity index. Percent ground cover was measured by recording the ground cover every 6 inches for 12-foot successional sequences on the north and south of each dune site. The data was used to run a chi-square statistical test. A comparison of the proportion of grasses and forb by finding the average ratio of the occurrence using the same successional sequence data recorded from the previous exercise. Completed counts of 3 focal grass species and 3 counts of focal tree species were taken and incorporated into chi-square tests for each species to compare the densities of the focal species within each dune. We found that the percentage of organic matter dramatically increased from the first dune until the savannah. The Simpson diversity index showed that the average density of species increased from the first dune steadily with every dune through the savannah. The average percent cover increased steadily from dune 1 through dune 5, then slightly tapered downward in percent coverage in the savannah. A chi-square test was performed on the data and revealed a statistical difference, supporting a change in percent cover in the differing sites. The average ratio of the proportion of grass to forbs showed a gradual decrease from the first dune through the savannah, which supports a decrease with dune age. Counts of focal species of grass showed no difference in each plot. A chi-square test supported this with p-values above 0.05. Two out of three tree species showed similar patterns when a chi-square test was run; with p-values above 0.05. One species, black oak, did, however, show significant statistical change with a p-value of <0.0001. This was also supported by the change in density (#/m^2) from beginning dunes to the end dunes, skewing the graph to the left.

Introduction: 

Indiana Dunes State Park is located along the shores of the southern tip of Lake Michigan and has been a long-studied site for ecological succession (Olson 1958). The concept of ecological succession, most likely, dates as far back as to when man first began cultivating agriculture (Olson 1958). The phenomenon of succession occurs at the Indiana Dunes as the shorelines continue to slowly recede, recovering from the ancient glacier that once existed over the Great Lakes region, and the linear formation of dunes occurs (Olson 1958). Each dune, separately, represents a time frame of succession from the time of its formation (Olson 1958). Succession represents a sequence for the colonization of species in an ecological community following a disturbance that opens up space (Connell and Slayter 1977). At the Indiana Dunes, succession can be seen from the changes in many different ecological dynamics, such as the relative abundance of species, richness in soil, and percentage coverage (Olson 1958; Smith and Houston 1987). Several models exist for measuring processes of succession. Some have argued that succession represents the complexity of more than just one model occurring simultaneously (Chaplin and Walker 2013).

In our research study, we were interested in measuring several different dynamics of ecological successional occurrences. Rather than depending on one model, we investigated five different methods of succession. These included soil development, diversity, percentage of ground cover, the proportion of grasses and forbs, and counting of focal species. We hypothesized that the amount of organic matter would increase in the dunes with age; that diversity would increase with dunes age, as they gradually shifted away from the shoreline to the savannah, with the fifth dune having the highest diversity index; that the percent of ground cover of vegetation would increase with dune age, with the fifth dune having the highest percent coverage; that the proportion of grass to forbs would decrease increase with dune age, with the fifth dune having the lowest proportion of grass forbs; the counts of focal grasses would decrease with dune age, with the fifth dune having the lowest count of focal grass of marram, while the counts of focal trees would increase with dune age, with the fifth dune having highest count of focal trees of black oak.

Materials and Methods:

The site used for this study included 5 divided plots. These plots were the first and second dunes, separately, off the shores of Lake Michigan. The third and fourth dunes were combined into one plot, referred to simply as dune three for conciseness. The fifth dune from the shoreline served as a plot, as well. The final plot was the black oak savannah, a wooded area where succession has flattened into a forested area away from the shoreline. We recorded data on both the north side, facing the lake, and the south side, facing away from the lake, of the dunes and savannah.

Nutrient sampling took place by filling three test tubes, approximately 4.5-5 mL, with samples of soil collected from three specific plots; nine test tubes total were filled. These plots included dune one, dune five, and the black oak savannah. Once the samples were collected, they were immersed in water to help separate the organic materials from sand. The test tubes were left to sit overnight in order for the materials to properly separate. Once separated, the measurements of the separated materials were taken for analysis of the ratio of sand to organic content (Fig. 1).

A secondary task involved randomly selecting a sampling site within the plot, creating a 1 x 1m square quadrat. Within this quadrat, we listed and counted all the stems of grasses and forbs. We repeated this to create a total of four 1 x 1m quadrats within a plot site. This was done for all five plot sites. The data collected from this exercise was used to calculate a Simpson diversity index (Fig. 2) and, separately, to create an average ratio for grass to forbs (Fig. 4).

An analysis was performed using a tape measure to vertically measure out a twelve-foot linear path; keeping the tape measure on the ground as a guide. Using a meter stick as a “point marker” to the ground, an observer measured out every six inches of the linear path. A separate observer recorded whether the meter stick hit bare ground or plant coverage every six inches of the linear path. Twenty-five points total were recorded in the twelve feet of the linear path. This was done twice for each side, north, and south, of the plot for a total of four times per plot. In order to calculate the percentage of plant coverage, the average number of points that were “cover” was divided by twenty-five, then multiplied by 100 (Table 3 and Fig. 3). A chi-square test was performed on the existing data collected from this exercise to determine if there was a statistical significance in increasing percentage cover.

A separate set of data involved measuring out a 12 x 12 ft quadrat on the south side of each plot site. Within the quadrat, all trees and size classes were identified and recorded. The circumference at breast height was also recorded for the tallest tree within the quadrat. Three quadrats were created per plot site. This was repeated for all five plot sites. The data gathered from this exercise was combined with the data from the second exercise for analysis. We analyzed three species of grasses and three species of trees that occurred in the five plot sites. The density of each species was averaged, and a chi-square of "goodness of fit" was run on each species average.

Results:

In the analysis of organic matter from the soil samples (Table 1 and Fig 1.), the largest amount of organic matter was found within the samples from the black oak savannah at 87.5%. Dune one contained the smallest recorded percentage of organic matter at 9.76%. The linear graph (Fig. 1) trends higher at the savannah than either dune one or dune five.

The Simpson diversity index measured highest at the savannah with an average density of 4.01 (Fig. 1). The lowest densities were seen at dunes one and two with densities of 2.54 and 2.53, respectively. The plots showed a gradual increase from plot 3 through plot 5 (Fig. 2).

Percentage ground cover of plants showed a gradual increase of coverage from the plot one through plot four (dune 5), which measured the highest at 63% coverage (Fig. 3). There was a slight decrease in coverage from plot four (dune 5) to plot 5 (the savannah), which fell to 56%. Plot 1 (dune 1) had the lowest coverage at 29% (Fig. 3). A chi-square of the data showed a P-value of <.0001 at 4 degrees of freedom (Table 7).

The average ratio of grasses to forbs showed a linear decrease as the plots moved further away from the shoreline. The lowest average ratio was the savannah (plot 5) at 0.41 (Table. 4). The highest ratio was recorded by dune 1 (plot 1) at 0.93. A downward trend is seen in Fig. 4.

Grass density per dune age was calculated for comparison (Table 5). The total average densities of all plot sites, expected density, of marram grass, bluegrass, and forbs was 4.33, 1.6, and 1.73, respectively. An average density for each individual plot site was analyzed to create observed densities (Fig 5). Observed densities were then compared to the expected densities in the chi-square “goodness of fit” test. The p-value for all three species of grass was above 0.05, with 4 degrees of freedom (Table 8). 

Tree density per dune age was calculated for comparison. The total average densities of all plot sites, the expected density, of Cottonwood, jack pine, and black oak was 0.73, 1.4, and 18.2, respectively (Table 6). An average density for each individual plot site was analyzed to create observed densities (Fig. 6). Observed densities were then compared to the expected densities in the chi-square “goodness of fit” test. The p-value for jack pine was above 0.05. However, the p-value for cottonwood was <.05. The p-value for black oak was <.0001 (Table 8).

Discussion:

The percentage of organic matter that was collected from three plot sites supported our hypothesis that organic matter would increase with dune age (Fig. 1). The black oak savannah contained a much higher amount of organic matter. This was an expected finding. Though it is heavily studied that facilitation may be a culprit of this process of organic matter, Johnson (2008) argues that there are many dynamics to succession occurring simultaneously. We do not contest Johnson’s (2018) argument due to the possibility that multiple dynamics are occurring; in addition to facilitation. However, we do believe that facilitation plays a large role in the change of organic composition within the Indiana Dunes. As discussed later on, we find patterns of higher diversity with dune age, which lends support that an increase in organic material may be facilitated by early pioneer species, such as marram grass.

The result of the Simpson diversity index supported our hypothesis that diversity would increase with dune age. In the data collected, dune five had the highest amount of diversity per the index (Fig. 2). Olson (1958) also found a higher species diversity in older dunes as he suggests a reason for this is more invading species from thickets of the forest colonize areas dominated by grasses and forbs.  

The results of the plant coverage percentages lend support to our hypothesis that ground coverage of vegetation would increase with dune age. The data trended with a gradual increase from the first plot to the last plot (Fig. 3). There was a slight drop from dune five to the black oak savannah, from 63% down to 56% (Fig. 3). This slight drop suggests that is possible that exploitative competition is occurring in the savannah. Competition may be an element of a climax equilibrium, where the late-successional species are maturing and possibly competing for canopy space, as a distinct pattern is taking place (Noble and Slatyer 1980). The P-value for the chi-square test, at <.0001, also suggested that there was a statistical significance of change coverage percent occurring with dune age. Based on this, we reject the null hypothesis that there is no change in percent cover.

The average ratio of grasses to forbs showed a steady decrease with dune age, as suggested by our data (Fig. 4). This supported our hypothesis that the average ratio of grasses to forbs would decrease with dune age. We surmise that as the diversity of species increases and the organic matter in the soil increases, it would be possible for a multitude of other vegetation to colonize and decrease the density of grass but increase the density of forbs. Under the belief that facilitation is occurring due to our data and previous studies (Olson 1958), it is possible that early species, grasses, are facilitating for later species, forbs.

Statistically, we failed to reject our null hypothesis that no change was occurring in grass density with dune age; as the P-value from the data of all three kinds of grass run in the chi-square “goodness for fit” test were above 0.05. Our hypothesis that grass density would decrease with dune age was not supported by our data either. Rather, the grass seemed to be closer to a constant occurrence at all plot sites as seen in Fig. 5. The p-value from the chi-square test that was run on the data from tree density varied. The p-value for cottonwood was above 0.05. For Cottonwood, we failed to reject the null hypothesis that no change was occurring in the average density of cottonwood. However, for the species jack pine and black oak, we found the p-value to be <0.05. In these two cases, we rejected the null hypothesis that no change in tree density was occurring with dune age. Our data supported our hypothesis, though only for black oak, that the density of tree occurrence would increase with dune age. Black pine left an anomaly as the growth did not match a full array of growth at all plots, but steady growth at only plots 3 - 5. The occurrence of increasing tree species showing growth in the later dune ages is a sign of later succession taking place; arguably due to facilitation (Huston and Smith 1987).

In this study, multiple models and methods of succession are taken into account. Though we conclude that succession is occurring, based on our data, we realize that many methods of succession are taking place. Some of the methods of succession were not specifically investigated, such as inhibition and tolerance. Several more studies can take place in order to further investigate. Future studies may include looking into the composition of nitrogen fixation and comparing it to vegetation that requires small and high amounts. Also, the testing of pH levels in the soil may be a suitable test for interference. As several researchers have noted, (Chapin and Walker 1987; Johnson and Miyanishi 2008), there is not one model that can give a definitive answer. Therefore, a continued compilation of investigations is a better way to test for succession.

Literature Cited

Connell, J. H., and R. O. Slatyer. 1977. Mechanisms of Succession in Natural Communities and Their Role in Community Stability and Organization. The American Naturalist 111:1119–1144.

Huston, M., and T. Smith. 1987. Plant Succession: Life History and Competition. The American Naturalist 130:168–198.

Indiana DNR. Indiana Dunes State Park. 2019. https://www.in.gov/dnr/parklake/2980.htm.

Johnson, E. A., and K. Miyanishi. 2008. Testing the assumptions of chronosequences in succession. Ecology Letters 11:419–431.

Noble, I. R., and R. O. Slatyer. 1980. The Use of Vital Attributes to Predict Successional Changes in Plant Communities Subject to Recurrent Disturbances. Succession 3:5–21.

Olson, J. S. 1958. Rates of Succession and Soil Changes on Southern Lake Michigan Sand Dunes. Botanical Gazette 119:125–170.

Walker, L. R., and F. S. Chapin. 1987. Interactions among Processes Controlling Successional Change. Oikos 50:131.

 Average Percent Cover S.E. Dune 1 29 6.608076 Dune 2 52 6.608076 Dune 3 56 9.092121 Dune 5 63 9.574271 Savanna 56 5.416026

 Appendix

Figure 1 Percentage of organic material in dune 1, dune 5, and the savannah

Table 1 Measurements of sand and organic matter in dune 1, dune 5, Savannah

Figure 2 Species Diversity vs Dune Age

Table 2 Avg Density per Dune Age and Standard Error

Figure 3 Percent Coverage of plants in each plot

Table 3 Average Plant Percent Cover and Standard Error

Figure 4 Ratio of Grass to Forbs per Plot

Table 4 Average ratio of grass to forb per dune age

Figure 5 Density (#/m^2) vs Dune Age for grass species

Table 5 Avg Densities of Grass species per plot age

Figure 6 Density (#/m^2) vs Dune Age for tree species

Table 6 Avg Densities of Tree Species per plot age

Table 7 Chi-square of ground coverage percentage

Table 8 Chi-squares of grass and tree species