click here for printable version

Site Assessment

Site Description
The site for our Garry oak woodland restoration is located at the corner of Wahkiakum Lane and another gravel footpath.  The site has a trapezoidal shape which measures approximately 40' x 60' in size (see Shrub Plan).  The site is slightly sloped with a southwest facing aspect.  Last fall, wood chips were dumped on the site by local arborists and spread to cover the entire area.  Although the mulch will be referred to in this report as "wood chips" for convenience, the chippings also included a high concentration of conifer needles and other leaves.  By the beginning of this quarter, a thick layer of wood chips covered the entire site, reaching depths of up to four feet on the east side of the plot. 

Existing Vegetation
Only after moving several cubic yards of mulch to install our treatments did we discover the original vegetation on the site.  Invasive perennial pasture grasses had dominated the site.  Some of these grasses managed to persist, although in etiolated form, even after being buried by a thick layer of mulch for over two months.  We also came across the cut bases of Himalayan blackberry plants in a few locations on the site.  Our most surprising discovery was a recently cut 4" diameter stump of what had probably been an apple or hawthorn tree.

Soil
A soil profile is useful in assessing site condition and gaining an understanding of what one has to work with. A core sample was taken at the intercept of two diagonals, each connecting opposite corners of the plot. What was found was; a 2 ft O horizon of Douglas fir clippings set apart from A by a bold interface, an  A horizon of 2-3 inches with aggregates beginning to form, <Strong interface> a B horizon of amorphous clay 5-8 inches thick, overlying a dense layer of coarse sand, the bottom of which was not found in the nearly 2 meter core sample. These findings were expected due to the nature of the site. The sample is indicative of a soil cap.

Due to the low organic matter and pore space in the A and B horizons a high bulk density can be expected. The high clay content of the upper soil suggests CEC will also be high providing a helpful buffer against the strong pH gradients anticipated within the sandy layer due to microbial decomposition of the landfill contents and the high fluctuation of ground water level. Also, the soil and groundwater temperature can be counted upon to be a few degrees warmer than adjacent areas, due to microbial activities. The clay cap acts like a barrier between the surface and ground water so candidate species for restoration should be both drought and flood tolerant. However, the underlying sand is currently acting as an aquifer so less drought resistant but deeper rooting species may be possible. A more thorough understanding of the site's hydrology should be pursued.

A concern on the site may be the effect that the conifer needles in the mulch will have upon lowering the pH of the soil.

Wind and Moisture
Desiccation and wind damage are further concerns for plot 7. Exposed with fair fetches in all directions gusting winds should be taken into account when selecting species. Furthermore, the winds will tend to come from the northwest in the summer bringing dry air from the city and robbing the planted community of moisture. Coupled with the accompanying drop of groundwater level and the fully exposed location, transpiration should be a concern.

Restoration Goals and Rationale

Our class was assigned the task of planting a shrub buffer around the mounded prairie site to facilitate restoration of the Lake Union Natural Area. Planting a shrub system around the mounded prairie restoration site is intended to accomplish four goals:
1)  Expand the prairie restoration project by planting natural systems normally found adjacent to prairies in Washington State.
2)  Provide a buffer against weedy windborne seeds that might blow into the site. 
3)  Provide a vegetative boundary against human intrusion into the site.
4)  Provide more diverse habitat for animals in the Lake Union Natural Area.
5) Test the effects of cardboard on weed suppression and native plant establishment.

Our group decided to fulfill the goals of this project by planting a Garry oak (Quercus garryana)/snowberry (Symphoricarpos albus var. albus) community as the mounded prairie buffer. Although the establishment of a Garry oak/snowberry buffer may take years to develop, we believe this system is the most appropriate buffer for the goals as previously stated. This section will explain the rationale of our decision and provide an initial assessment of how well we feel we have accomplished our goals.

Expanding the prairie system
According to Franklin and Dyrness (1973) Douglas-fir (Pseudotsuga menziesii) and Garry oak have invaded prairies in recent years due to human practices of fire exclusion. These tree species have pioneered into the prairies from adjacent forests and woodlands that have existed for thousands of years in a dynamic state with the prairies. Although both conifer and oak woodland systems abut the prairies, the process that maintained prairie systems (i.e., fire) has also maintained oak woodlands. Because fire has been important to both systems, the herbaceous component of oak woodlands is very similar to that seen in the prairies. Therefore, when determining what system we would use to buffer the mounded prairie system, our group decided that an oak woodland system would be more appropriate than a Douglas-fir forest.

When considering what species to plant, we also considered the edaphic conditions of our site. As was stated in the site analysis section, our site is heavily influenced by a thick clay layer in the soil that creates very wet conditions for the majority of the year. Taylor and Boss (1975) state that the ecological success of Garry oak in Washington State "depends in large part upon this tree's rather broad edaphic tolerance and ability to grow on 'poor' sites." Garry oak can be found in xeric, mesic, and aquic soil conditions, although competition by Douglas-fir in more mesic sites is in large part responsible for the current reduction in oak habitats. Considering the soil conditions, Garry oak may have the best chance of survival on the site of all species we have to plant. 

In her research on classifying South Puget Sound prairie plant communities, Jeanette Dorner (1999) found one shrub-dominated community.  She describes this as a Symphoricarpos albus dominated community, in which Garry oak, Viola nuttalli, and Rubus ursinus also occur.  Of the four main plant associations of Quercus garryana that Franklin and Dyrness (1973 after Thilenius 1964, 1968) report within the Willamette Valley, the least disturbed sites have an understory dominated by Amelanchier alnifolia - Symphoricarpos albus. Snowberry plantings have already been successfully established in other parts of the Union Bay Natural Area, indicating that this species is tolerant of the unusual soil conditions on the site.  We have chosen to plant snowberry because it is a common element in the South Puget Sound prairies and adjacent oak woodlands and because it has been proven to succeed in the Natural Area.

Provide buffer against windborne weedy seeds
The denser the planting within our shrub buffer area, the better will be the buffer against windborne weedy seeds. Realistically, the success of our shrub buffer area will not be realized for a number of years. However, once the shrubs take hold, the snowberries should provide a very dense layer that will reduce the amount of windborne weeds entering the site. Snowberry is known to produce dense mats beneath the canopy in oak woodland habitats, and is therefore the appropriate choice for a weed buffer.  In addition to the Garry oak/snowberry community restoration, an ornamental planting of shrubs along Wahkiakum Lane on will also buffer the north side of the prairie from weeds.  Furthermore, to reduce weed establishment within the shrub buffer area we will retain a thick layer of mulch on the site.

Provide a vegetative boundary against human intrusion
As was stated for the buffer section above, the success of our shrub buffer area will not be realized for a number of years. However, once the shrubs take hold, the snowberries should provide a very dense layer that will reduce the amount of intrusion into the site. The trick will be to keep people out of the site until the shrubs take hold. Three sides of the prairie site are already protected (i.e., at least in spring). The parking lot edge is signed and protected by standing water. The south side of our plot is adjacent to the next group's plot, and has little appeal for the general public. And, the north side of our plot is maintained by a temporary fence and signage as well as with mulch and last year's plantings.

We plan to accomplish the task of protecting our shrub buffer by erecting temporary fencing on our site with fence stakes and string, in a similar fashion to the north side of our prairie site. We will fence the north and east sections of our plot. As our plot is located in the junction of Wahkiakum Lane and a side trail, the installation of a fence will be necessary to protect our site from visitors prone to cutting trails. The fences will be signed to alert the public of the ongoing restoration activities and urge them not to disturb the site. However, because of the temporary nature of the fence and signs (due to our limited budget), continuing maintenance will be necessary to preserve the effectiveness of crowd control until the shrubs take hold. This topic will be discussed in greater detail in the management section of this report.

Provide more diverse habitat for animals
The importance of oak habitat for wildlife cannot be understated. The structural diversity, mix of plant associations, processes of decay of the Garry oak, and the large food supply (e.g., acorns, leaves, fungi, insects, grasses and herbs) within Garry oak habitat combine to create a habitat that supports a diverse group of wildlife (Larsen and Morgan 1998). In fact, Larsen and Morgan (1998, after Christisen and Koschgen 1955, USFS 1969) suggest that acorns influence more wildlife species than any other single kind of natural food, reporting 186 species dependent on acorns. In addition to their edible berries, the dense, twiggy growth of snowberries also provides good cover for birds.  Snowberries replace the shrub habitat that has been removed from the site by mowing blackberries.  Taking this into consideration, the best way to provide habitat for a diverse group of wildlife would be to plant an oak/snowberry buffer.

Test cardboard for weed suppression
We will test the use of cardboard barriers as a treatment in our shrub plots. According to Harris, et al. (1999), mulch and cardboard together can be an effective tool to control weeds. Most sources suggest removing weeds prior to mulching if mulching is used alone. The addition of cardboard (or newspaper) is suggested as a substitute for weeding.  According to the Pacific Northwest Weed Control Handbook, "during emergence, rhizomes of some perennials become disoriented when penetrating 6 to 16 layers of overlapped newspaper" (William et al. 2000).  Similar effects should be found with corrugated cardboard.

As stated previously, although we believe the plantings will take years to develop, the oak/snowberry plant association will best serve the goals of our restoration project. Only time will tell if the sites edaphic conditions are appropriate for our design. We believe, however, that with careful planning and using the best restoration procedures available for our work will help to minimize stress to the planting stock and insure high survivability.   

New Plantings and Experiments

New plantings for the shrub buffer will take place in two main areas: treatment plots and along Wahkiakum Lane. Based on the rationale explained above, our treatment plots will be planted to replicate a Garry oak/snowberry plant association. However, because we don't want to waste available planting stock, all bare root plants will be used in our work. All other plants will be planted along Wahkiakum Lane to provide an ornamental, native shrub buffer to the prairie and shrub sites.

Treatment plots
The main experiment within our plot is to test the efficacy of weed barrier treatments on the survival and growth of native shrubs and the exclusion of non-native species from our plots. For the experiment we will contrast the use of mulch alone against the use of mulch with a double layer of cardboard.  Because mulch is proven to suppress weeds and help retain soil moisture (Harris et al. 1999), we are using mulch rather than bare ground as a control.  We are interested in testing the effects of cardboard on further improving weed suppression and native plant survival.  There are 8 replicates of the two treatments within our plot  (see Shrub Plan). The treatments were randomly assigned to the different plots using a random number generator within Microsoft Excel.

Treatment plots were planted with a central Garry oak surrounded by eighteen snowberry starts (see Shrub Planting Diagram). Site preparation entailed clearing the mulch from each treatment plot and digging out any blackberries.  Herbaceous weeds were not removed and the soil was not dug or amended.  Oak seedlings were planted first and growth tubes were inserted. Two layers of corrugated cardboard were then applied to half of the plots and staked using turf staples. To determine the appropriate plant spacing within our plots, a pattern cut of clear plastic was overlain on each plot and each planting was marked before the pattern was removed. Mulch only treatments were marked with a flag (photo), and a small hole was excavated at each planting location to mark the cardboard treatments (photo).

All starts were planted by excavating holes large enough for the individual root stock and back filling with native soils (photos 1, 2, 3, 4). If soils were clumped (as they often were due to the clay properties found at our site) the clumps were broken up before back filling. All starts were watered immediately upon planting.  Following planting, all treatment plots were covered with a four inch layer of mulch (photo).

To facilitate the growth of Garry oak, we placed the seedlings in growth tubes. Growth tubes will effectively shade the oaks and provide some protection against browsing. Papanikolas (1997) found that shading increased oak seedling survival by 74%. Shading will help prevent desiccation in the hot dry summers of the Pacific Northwest, especially in the open conditions of our study site.

Wahkiakum Lane planting
Several species of shrubs were made available to our group for this restoration project.  In order to simplify our experimental design and more accurately recreate a Garry oak/snowberry plant association, we chose to plant only snowberries in the majority of our shrub plot.  With the extra plant stock we have for our work, we planted an ornamental, native shrub buffer along Wahkiakum Lane (See Shrub Plan and Prairie Plan). Previous groups have started planting shrubs adjacent to the prairie area; however, planting densities are relatively low.

We arranged the species along a moisture gradient from relatively dry in the east to wet in the west.  Serviceberry (Amelanchier alnifolia) was planted on the northern borders of the shrub restoration plot.  Oceanspray (Holodiscus discolor), mock orange (Philadelphus lewisii), and red flowering currant (Ribes sanguineum) were intermixed on the east side of the border by the prairie plot.  Rose (Rosa pisocarpa) and ninebark (Physocarpus capitatus) were planted on the wet western part of the border and in the wet part of  the prairie plot.  A wet spot in the northeast corner of the shrub plot was also planted with rose.  The soil throughout this border planting had a high proportion of coarse gravel in the upper soil horizons.  The gravel probably originated both from Wahkiakum Lane and the former parking lot.

Although they are not being used to restore a specific native plant community, these shrubs will protect the north side of our prairie plot from weed invasions, provide habitat for birds, increase native plant biodiversity, and improve the area's aesthetic value.  For example, we planted red flowering currant (Ribes sanguineum) by the path to provide a nectar source for hummingbirds and delight passersby with its showy red flowers. 

Management Prescriptions

To continue the restoration of our site, we recommend that next year's group:
1)  Evaluate the success of our planting and modify our techniques as appropriate.
2)  Analyze the effectiveness of the treatments and determine if cardboard should be added to the mulch only treatment.
3)  Determine the extent of weed invasion and take appropriate measures to ameliorate the influence of weeds on our site.
4) Increase the density of shrub plantings.
5)  Reinforce our site's boundaries for crowd control. (This activity really should take place year round.)

Success of planting
Our group was concerned by the poor soil conditions existing on our site, but didn't know we had the resources to amend the soils until we were far into our restoration activities. If next year's group identifies very poor survival of our plants, the group may want to add topsoil to our site to provide a better growing medium for the shrub plants. Also of concern is the amount of mulch existing on our site. Although our group displaced an incredible amount of the mulch, there was more than could be reasonably removed. Because of the deep mulch, some of the snowberry plants are in danger of being buried. However, our group determined this to be part of the overall experiment.

Cardboard addition
Next year's measurements should include survival of plantings, mean height of snowberry, and non-native percent cover. Height of snowberry is measured from soil to apex of plant. Because it would be near impossible to compare changes in height between individual plants, we have measured all plants within each plot  but suggest that treatment means be used to perform statistical tests (see Table 4).  Since there are only two treatments to compare in this experiment, mean percent survival, mean height, and mean percent cover should be calculated for each treatment and compared using t-tests to test the following hypotheses:

H_O: There is no difference in non-native percent cover between the cardboard and the mulch only treatment.
H_A: Non-native percent cover differs between treatments.

H_O: There is no difference in native plant species survival between the cardboard and mulch only treatments.
H_A: Survival differs between the two treatments.

H_O: There is no difference in snowberry height between the cardboard and the mulch only treatment
H_A: Snowberry height differs between treatments

Harris, et al. (1999) suggest that mulch is effective at maintaining soil moisture, reducing stress on plants, reducing overall weed growth, and stopping clay soils from cracking during dry periods. They go on to suggest that although mulch is effective at reducing weed growth, the addition of a barrier is much better at weed exclusion. If next year's group determines that our treatments also reflect the literature findings, they may elect to add cardboard to the mulch only plots and/or use cardboard for new plantings.

Weed invasion 
During plot preparation and planting, our group noticed and removed a number of blackberry burls from the site. Because it appears the site was infested with blackberries before mulching, ongoing maintenance and removal will be important to extirpate blackberries from the site. Furthermore, until the shrub community is established there is great potential for annual plant invasion.  If the mulch is not effective at preventing weed invasions, we recommend planting more native plants and/or using a selective herbicide to reduce overall weed cover.

Denser shrub plantings
We planted approximately 290 bare root snowberry starts on our site, but the areas between are treatment plots are still bare. If more snowberries are available for planting next year, we recommend using this species to create a greater density of shrubs in between the plots (i.e., if soil conditions do not prove to be prohibitive). 

Boundary conditions
We have placed signs and temporary fencing around our site along the most traversed perimeters. However, because signs and fencing can be torn down, it will be important to periodically check the signs and fencing throughout the year and replace as needed. We recommend that this be done more periodic during times when extra-curricular activities (e.g., sporting events) draw large crowds to the Union Bay Natural Area.