Parks Write-up

Analyzing the effects of Douglas fir as a keystone species on prevalence of invasive
species
Introduction:
Douglas fir is a keystone species, which means it has a disproportionate effect on
ecosystem function relative to its abundance. (Flynn). It is important for species diversity
and organization in an ecosystem. Keystone species help make an optimal environment
for the ecosystem to function. A loss of Douglas fir trees in an ecosystem leads to a loss
of native understory plants, which results in changes in the makeup of the ecosystem.
These changes have impacts on insect and plant species that rely on environmental
conditions that are generated by Douglas firs. For example, Douglas fir trees change the
chemistry of soil, which enables these species to survive. (Flynn).
The question we will attempt to answer is whether or not presence of Douglas fir
has an effect on prevalence of invasive species nearby. We propose that it may function
as a sort of “inverse invasional meltdown” in which the presence of more keystone
Douglas fir trees (greater biomass of Douglas fir) may alter the environment in a way that
decreases the likelihood of invasion by non-native species (Krosby). Our hypothesis is
that an increase in Douglas fir will increase native plant biomass and decrease invasive
biomass. Conversely, our null hypothesis is that the negative correlation between
Douglas fir biomass and invasive biomass is due to chance. Given the role of Douglas fir,
we predict that an increase in biomass of the Douglas fir would help other native species
that have evolved to live in ecosystems with the Douglas fir grow, and thus make it
harder for invasive species to grow. To test this we will compare Douglas fir biomass
with biomass of six invasive species we have learned to identify (Table 1) to see if there
is a decrease in invasive biomass with an increase in Douglas fir biomass. Specifically we
will be looking at the effects that the Douglas fir has on the invasives across all parks and
all years for which data is available.
Methods:
As part of our methods for performing the analysis, we followed the general field
protocol for Conservation Biology 476 (Tewksbury 2009), which should be on file with
the Seattle Parks. In analyzing our data we performed an unpaired t-test using the
program R and Microsoft Excel because we wanted to test the significance of the
correlation between Douglas fir and invasives. We used the t-test because we only
wanted to include two groups: invasive biomass with Douglas fir and invasive biomass
without Douglas fir. We ran an unpaired t-test because we had a different number of data
points with and without Douglas fir to compare, so we weren’t able to compare by point.
We took out the plots in Seward that didn’t have data to prevent irrelevant points from
confounding our results. Additionally, we excluded t-test data from Interlaken Park
because there was only one point recorded with Douglas fir and thus not enough data to
run a t-test. We presented our data in a bar graph format with “Invasives Present” on the
x-axis, with and without the presence of Douglas fir, and “Invasive Percent Coverage” on
the y-axis. We then made standard error bars for each to determine the relative variability
in our results. We did not include points where there was no data on invasive species.
Scatter plots were not feasible in the absence of sufficient data. Calculations performed
included averages, standard error, and a t-test for each park.

Results:
Our data shows the combined invasive and Douglas fir percent cover in the years
2007, 2009, 2013, 2014 and 2015 in six different Seattle Parks. The x-axis is labeled with
the abbreviations of the invasive species (Table 1). The bar graphs for each park give
conflicting results in relation to our hypothesis (Figures 1-6). As shown in the error bars
in our graphs, the standard error was found to be highly variable in each park. In pooling
all of the data from all parks over all years, we constructed a bar graph showing the
correlation between presence or absence of Douglas fir and percent coverage of each of
the twelve local invasive species. Table 2 shows the results of our t-tests for each park. A
p-value of 0.05 or lower traditionally corresponds to rejection of the null hypothesis,
whereas a higher p-value does not support rejection of the null hypothesis. For Carkeek
and Ravenna Park, a low p-value was found, suggesting we can reject the null hypothesis.
However for Discovery and Seward Park, a high p-value was determined, and we cannot
reject the null hypothesis. Moreover, the p-value for our final graph, the combined data
from all of the parks, gave a high p-value of 0.8366 to further contradict our hypothesis.
Discussion
Initially we wanted to establish several ranges of Douglas Fir cover (i.e.: no
Douglas Fir, low, medium, and high) to look at how percent coverage of Douglas Fir
affects invasive species cover, but unfortunately we didn’t have enough points with
Douglas Fir to create multiple levels of the independent variable. As a result we had to
group it into sites with and without Douglas fir. We also wanted to look at data in each
park year-to-year to see how percent Douglas Fir increased or decreased over time and
what its effect is on invasives, but we didn’t have enough points with Douglas Fir so
decided it would be more effective to just group all of the years together and analyze each
park’s total over the years.
Additionally, we wanted to compare “cm squared of Douglas Fir” to the “cm
squared of invasives (specifically the English Laurel)”. After combining the data and
creating a scatterplot, we determined that there was not enough information present in the
graphs or consistency in points at which the data was collected to infer anything from a
scatter plot. Consequently, we decided to stick to our initial strategy of the bar graphs and
did not take up the scatterplot approach. As shown in the bar graphs, our standard error
lines overlap to a fairly large degree, which can be attributed to limited data points and
wide variability. As stated before, we excluded Interlaken Park in the t-test because there
was only one point with Douglas fir, so we were unable to run a t-test for it.
One thing to note: the results of our t-test for Ravenna Park could be slightly
skewed due to the available data points. Ravenna Park only had three data points for
invasives with Douglas fir, but all three happened to have no invasive species. While this
gives a tremendously low p-value, the data is not necessarily reflective of the rest of the
points in the park with Douglas fir and could easily be a result of sampling bias.
Another issue with data collection is a limited knowledge of and ability to identify
different invasive species. It’s very possible that invasive cover was underreported, and
only taking into account twelve invasive species has its own shortcomings. There is also

a possibility of other native keystone species that were not measured interfering with
results. In the future, collecting data from more points will be very helpful, and collecting
data on all species present could also help to avoid problems like the ones described
above.
In our data analysis we looked at the effects of Douglas fir as a keystone species
on invasive species biomass. However, other factors such as complex interspecific
interactions may also play a role in the effect of Douglas fir on invasive species biomass.
It is worth investigating factors such as interspecific interactions and perhaps the effect of
other keystone species on invasive biomass to determine the implications for endemic
species conservation.
Recommendations:
Our p-value from Carkeek Park allows us to accept our hypothesis that Douglas
fir does, in fact, decrease invasive species biomass. Figure 1 shows that the percent cover
of Himalayan Blackberry greatly decreased with Douglas fir present. There was also a
decrease in English Ivy, Evergreen Blackberry, and Stinky Bob. We recommend planting
more Douglas firs to continue counteracting the invasive species population.
Our p-value from Ravenna Park allows us to accept our hypothesis. Figure 4
shows that all invasive species biomass decreased with the presence of Douglas fir. In
order to continue to have a decreased percent cover of English Ivy, Himalayan
Blackberry, Evergreen Blackberry, English Holly, English Laurel, and Stinky Bob we
suggest that more Douglas firs be planted.
For Interlaken Park there was insufficient data to run a t-test. We recommend
further researching the park and collecting data to help reach a solution to decreasing
invasive biomass. Our p-values from both Discovery Park and Seward Park were not
consistent with our hypothesis. However, we recommend looking further into other
factors, including but not limited to, soil conditions and other species interactions. More
data will also help future conservation biologists to suggest ideas to decrease invasive
biomass.
We suggest gearing the Douglas fir planting projects in Carkeek and Ravenna in a
fun way that incorporates community involvement. Family events such as “Plant-A-
Doug-Fir” help raise awareness as well as raise community morale.
References:
Lecture notes:
 Krosby, M. 2015. Species Invasions [Lecture notes]. Seattle, WA: University of
Washington, Department of Biology.
Website:
 Flynn, Samantha. "Coastal Douglas-fir Ecosystems." Ministry of Environment,
Lands and Parks. BC Conservation Data Center, Mar. 1999. Web.
<http://www.env.gov.bc.ca/wld/documents/douglasfir.pdf>.
Journal article:
 Howe, Gleen T., Et. Al.
"Breeding Douglas Fir." Plant Breeding Reviews 27 (2006): 245-
353. Web. 02 Feb. 2015.

 Munscher, Eric. "Understanding the Role of Keystone Species in Their
Ecosystems." SWCA Environmental Consultants. N.p., Sept. 2013. Web. 02 Feb.
2015.
Tables/Figures:
Table 1. Invasive Species Abbreviations Used
HEHE English Ivy
RUDI Himalayan Blackberry
RULA Evergreen Blackberry
CLVI Traveler’s Joy
DALA Spurge Laurel
CYSC Scot’s Broom
SODU Bittersweet Nightshade
FAJA Japanese Knotweed
ILAQ English Holly
PRLA English Laurel
TAOF Dandelion
GERO Stinky Bob
Table 2. Results of t-test by park
Park P-value Degrees of freedom
Carkeek 0.01826 19.106
Discovery 0.2566 46.822
Interlaken N/A N/A
Ravenna 0.00002911 27
Seward 0.8093 21.385
Overall 0.8366 5.563

Figure 1.
Figure 1. Percent Cover of different species of Invasives with and without Douglas fir in
Carkeek Park 2007. 2009, 2012, 2013, 2014 and 2015.
Figure 2.
Figure 2. Percent Cover of Invasives with and without Douglas Fir in Discovery Park
2007, 2009, 2012, 2013, 2014 and 2015

Figure 3.
Figure 3. Percent Cover of Invasives with and without Douglas Fir in Interlaken Park
2007, 2009, 2012, 2013, 2014 and 2015
Figure 4.
Figure 4. Percent Cover of Invasives with and without Douglas Fir in Ravenna Park
2007, 2009, 2012, 2013, 2014 and 2015

Figure 5.
Figure 5. Percent Cover of Invasives with and without Douglas Fir in Seward Park 2007,
2009, 2012, 2013, 2014 and 2015
Figure 6.

Figure 6. Percent Cover of Invasives with and without Douglas Fir in Carkeek,
Discovery, Interlaken, Seward and Ravenna Parks 2007 ,2009, 2012, 2013, 2014 and
2015

Parks Write-up

More Related Content

Similar to Parks Write-up (20)

Parks Write-up