The Effect of Wildfire on Forest Composition

Figures:

Results:

In Figure 1, we counted the number of saplings at each location and found that the control group had the least amount of saplings, 6, while the location 24 years after the fire, had the most with 39. The other two testing sites, the 15 and 20 year locations, had 30 and 21, respectively. We see a relatively linear relationship between the number of saplings in a location and the time elapsed since a forest fire where, when more time has passed, there are more saplings growing in each location.

In Figure 2, we found that our control group had the highest density soil samples, 95.9%, while the 15 year location had the lowest density of the group at 63.5%. The 24 year location had the second lowest density at 82.1%, and the 20 year group had the second highest density of 85.9%. Graphically, we believe the control location could be distinct from the three locations affected by fire, as it has slightly more dense soil than the other three sites. However, with the size of the error bars and the relatively small difference between the four sites, we also believe that we could also reasonably assume no relationship between soil density and the recency of forest fire.

In Figure 3, we determined that the 20 year location and the control location had very similar canopy cover percentages 63.8% and 63.2%, respectively. The 24 year location had a 44.6% canopy cover, while the 15 year group had no trees and thus, a 0% canopy cover. Clearly, the site which burned 15 years ago, is significantly different from the three other sites as it had no canopy coverage. The other sites, however, we believe are indistinct from one another.

In Figure 4, we found that the 24 year location had far and away the most trees overall with 19 group 1 trees. In contrast, our control location has only 4 trees, but those trees are larger and more mature having only one group 1 tree, two group 2 trees and one group 3 tree. The 15 and 20 year site are sparsely covered with small trees with the 15 year site having three group one trees and the 20 having three group one trees and one group 3 tree. The control group and the 24 year group are distinct from the rest of the data, however the 15 and 20 year sites are indistinct from one another. Overall, the data seems to trend in the direction of more and smaller trees the less recent a fire site is.

Fig 5: A map of the Blue Hills Reservation and parts of the surrounding town of Milton. Studied sites are circled in red.

Discussion:

Both the sapling data and the tree number and size data exhibit the same trend of the number of trees increasing with the time since the fire. (Fig 1, Fig 4) In both cases, the 24 year site exhibits the densest, smallest vegetation, while the control site has fewer but larger and more mature trees. That relationship fits well into the initial stages of the secondary succession model. Secondary succession is the process which an environment recovers after an event kills all the organic material, but leaves the soil intact. Before the fires, the forest would have been, for the most part, open and populated by large, mature trees, which is what we see in the control group (Fig 4). In the case of our forest fires, the blaze would have swept through the sites and burned all organic material there except for seeds buried underground. Without the canopy shielding the forest floor, smaller, sunlight loving trees like conifers would sprout saplings. The 15 year site is exhibiting that stage as it has the least trees of any of the sites,which are all group 1, and therefore it had 0% canopy cover as well (Fig 3, Fig 4). Then, those saplings mature into smaller trees resulting in the conditions we see at the 24 year site. Eventually, shade tolerant deciduous trees will probably sprout under the shade of the conifers and take the forest back over, but that will take quite some time.

The two possible trends in the soil sample data also could possibly fit into the secondary succession model. In the case that we proceed with the interpretation that the soil density has no relationship to the years since a forest fire occurred, (Fig 3) that could support the concept that fire doesn’t physically change the soil. Other conclusions that fit this interpretation could be either that the fires in the area were not severe enough to alter the soil density or that sufficient time has passed for the soil density to recover. Therefore it doesn’t change the soil’s suitability for plant growth, and the sites are all still in the secondary succession phase. If we interpret the data to be that the control site has more dense soil than the other three sites, that could suggest that the increased growth of all smaller shrubs, saplings, and trees could be decompacting the dirt with their roots.

One source of error came with our soil samples. Because the samples had slightly fallen apart when we returned to the lab, we took only a fraction of the sample into account. Though we aimed to take a 3cm core sample from the top of each sample, some cores had lost enough on top that we had to take the measurement a couple of centimeters away from the top, which could result in slightly denser than average soil samples. Another source of error came from the Biltmore stick. Though it is a very useful tool, as we mentioned in the methods video, there is a certain amount of imprecision that comes with using the Biltmore stick. The free-hand nature of the tool could have led to some minor differences in the size of the trees. The final source of error was exhibited in our canopy cover. Because the mirror forced us to take the pictures at an angle, photoshopping the images into a perfect circle became more difficult. Because of the ovular picture of the mirror, we could not edit out some small pieces outside of the mirror, resulting in a very slight change in the percentages given by the color extractor.

In the future, we could expand upon this study by collecting more in depth data on the soil. One of our initial plans with the soil had been to test the amount of certain nutrients with them and compare that to what normal levels of those nutrients would be. However, due to both the time constraints and the complexity that this research would add to the study, we decided not to collect this data. Additionally, going back and testing the canopy cover in the later months of spring and summer would be helpful since when we went there, there were no leaves that provided additional cover.

Due to the ravaging effects of forest fires along the west coast of the United States, research into the long term effects of forest fires has become increasingly important. With all of the data collected related to trees, we can make the conclusion that these locations are all at various stages in the process of secondary succession and we can begin to establish a timeline for that process. However, between the limited amount of data and the imprecision of our soil testing, we unfortunately cannot make any substantive conclusion about the effects of forest fires on the soil's composition.

Sources:

Skyline Trail in Blue Hills. (n.d.). [Photograph]. Trail Mob. http://trailmob.com/trail/skyline-trail-in-the-blue-hills/

Dublin, R. D., Lew, K. L., Bartley, B. B., Rhode, E. R., Williams, J. L. W., & Honea, D. H. (Eds.). (2001). Succession - Changing Forest Habitats. In Alaska’s Forests and Wildlife - Alaska Wildlife Curriculum Teacher’s Guide (pp. 43–52). Alaska Department of Fish and Game Division of Wildlife Conservation.

Stoof, C (2011). Fire effects on soil and hydrology [Doctoral dissertation, Wageningen University