Blog Archives
Planting mini-forests in eastern Pennsylvania
Collaborative summary of a project undertaken by Esther Cisneros-Ramos, Tessa Dougan, Lauren Hagerty, Isabella Jacavone, Louise Bugbee, Ben Felzer, and Bob Booth
Development of forest and agricultural land is likely unavoidable in some regions. There is a clear need to advance our understanding of ways to minimize the negative impacts of this development on wildlife and ecosystem services. For example, in the Lehigh Valley of eastern Pennsylvania, farmland is rapidly being bought by developers and converted into manufacturing and distributing facilities. Much of this growth is being fueled by e-commerce, as large warehouses are needed for distribution hubs to allow companies like Amazon to deliver products to our doorsteps in only a few days. Land around new construction sites is often landscaped with grass and sparsely planted trees, and these plantings are typically non-native, ornamental species. This approach to landscaping requires continual mowing, provides little wildlife value, and likely provides few other ecosystem services.
In the fall of 2022, we embarked on a reforestation project that attempts to demonstrate a potential alternative approach to traditional landscaping. Our project combines principles of the Miyawaki method of forestry with an interest in agroforestry, and aims to assess the benefits of planting small, high-density, native agroforests around warehouses and other structures. The Miyawaki method has been used to successfully produce productive and functional small forests in a relatively short time in other regions (see review in Lewis, 2022), and mini-forests like these would allow for warehouses and other facilities to be better hidden from the public eye and arguably have greater aesthetic appeal than traditional landscaping. In addition to providing a demonstration of the Miyawaki approach, our goal was to explore the creation of forests with high production of nuts, berries, and other edible foods for people and wildlife. Furthermore, we hoped to establish plots using various approaches, so that future students, researchers, and the Lehigh Valley community could assess and visualize how these different approaches might affect outcomes related to habitat quality, aesthetics, biodiversity, and ecosystem services like carbon sequestration.
What is the Miyawaki method?
The Miyawaki method of forestry is a planting technique first introduced by Akira Myawaki, a botanist and plant ecologist interested in the restoration of degraded land. In this method of planting, species of the canopy, subcanopy, and lower vertical layers of a forest are planted simultaneously, and selected intentionally to reflect their frequency in native forests of a region and their specific traits. The condition of the soil is first assessed and improved if necessary, and seedlings are then planted in very high density, oftentimes 2 to 7 individuals per square meter (Lewis, 2022). This high density of planting is vastly different from traditional landscaping and most forest restoration methods. A hectare of Miyawaki forest would likely include 25,000 individuals while traditional planting might contain 1,000 individuals! However, usually only small areas are planted using the Miyawaki method. Ultimately, after about three years of relatively low-maintenance weeding and watering, the method has produced small, self-sustaining forests even in the middle of urban areas.
General questions guiding our project
- Are the principles of the Myawaki method an effective and efficient planting technique for nut and fruit-bearing agroforests at small scales?
- Are agroforests planted using the principles of the Miyawaki method a practical alternative to traditional landscaping methods?
- What ecosystem services does a Miyawaki-style forest provide in comparison to traditional landscaping, or in comparison to just letting secondary succession take place?
- Does planting seeds versus seedlings/saplings have any effect on the composition and ecosystem services of an eventual mini-forest?
Design and establishment of demonstration plots
We established four, 10 x 10 meter plots on recently abandoned agricultural (corn) land in Wayne Grube Memorial Park in Northampton County, Pennsylvania. Each plot was cleared of vegetation, and different approaches to revegetation were applied in each.
A land developer might perform traditional landscaping around a warehouse or structure, or might leave the area to naturally revegetate; therefore we established two plots to reflect these possible scenarios (traditional landscaping, natural succession). Ready-to-plant seed options might be convenient for landscapers, although planting seeds is not the typical Miyawaki approach. However, to assess the potential of planting lower-cost and potentially more genetically diverse seeds, we established one of our densely planted plots using seedballs. In the final plot we planted saplings at the high densities used in the Miyawaki method. The plots are detailed below:
Traditional landscaping (Plot 1)
This plot was meant to reflect the typical style of landscaping used by developers and commercial landscapers around areas such as warehouses. The plot was planted with a blue spruce and a red maple sapling, and ornamental grass. About 4 inches (~10cm) of mulch was spread on the entire plot prior to planting.
Miyawaki-inspired with saplings (Plot 2)
This plot was planted with eleven species using saplings arranged in high-density (8 individuals per 2×2 meter subplot). A total of 200 saplings were planted, and the precise arrangement of species is documented in the next section of this blog post. About 4 inches (~10cm) of mulch was spread on the entire plot prior to planting.
Miyawaki-inspired with seedballs (Plot 3)
We made seed balls by placing 2-5 seeds (depending on size) in a mixture of natural rodent repellent, cat litter, clay, and potting soil. Each seed ball contained seeds of one species. These balls were planted in the same pattern seen in the Miyawaki-inspired sapling plot (detailed in next section). About 4 inches (~10cm) of mulch was spread on the entire plot prior to planting.
Natural succession (Plot 4)
We cleared this plot but did not plant anything, and it will be allowed to revegetate without intervention.
Planting configuration of the Miyawaki-inspired plots
Our Miyawaki-inspired plots were divided into 2 x 2 meter subplots to help organize and arrange the planting. Although Miyawaki plots are usually planted using a random distribution of native plants, we guided our planting arrangement based on plant characteristics. To inform our discussion of what plant species to use and how to arrange them, we first compiled information on the traits of native shrub and tree species. For example, knowledge of adult tree height, shade tolerance, deer palatability, growth rate, and other factors helped us identify species that might do well together and provided information on how to potentially arrange them within the plots. Some species that produce edible nuts, like black walnuts, were eliminated from our list because of toxicity or incompatibility with other plants. Although there were other species that we could have planted, in the end the final species inventory was shaped by what was donated, available at nurseries, and acquired in time for planting.
Our final species list included common persimmon (Diospyros virginiana), shagbark and shellbark hickory (Carya ovata, Carya laciniosa) hazelnut (Corylus americana), black gum (Nyssa sylvatica), red mulberry (Morus rubra), arrowwood (Viburnum dentatum), nannyberry (Viburnum lentago), redbud (Cercis canadensis), hawthorn (Crataegus sp.), spicebush (Lindera benzoin), and black currants (Ribes americanum). Hickories and persimmons (the main canopy species) were positioned in the center of the plot at roughly equal distances apart to ensure the highest rate of survival. Species less palatable to deer such as arrowwood, currants, and spicebush were primarily located around the perimeter to deter deer from entering the plots. Each 2 x 2 meter subplot contained a diversity of species with a range of potential adult heights, and each species was generally spread evenly throughout the entire plot in case it was impacted by disease or experienced mortality from other causes.
A major theme of our discussions was protection from deer herbivory, a huge problem in Pennsylvania forests. However, we wanted to keep our demonstration as realistic as possible, and we doubted that developers interested in applying the Miyawaki method would be willing to invest in temporary fencing to keep deer out. Therefore, in addition to planting the edges of the plots with deer resistant or less palatable species, we also lined the perimeter of the plots with piles of woody debris, something that developers would likely have ready access to. We hope that this wall of debris will keep deer herbivory low while the plants get established; however, we will be monitoring the plots and may need to develop other solutions if the deer pressure becomes too great.
Planting and baseline data collection
We planted the Miyawaki-inspired sapling plot on October 22nd 2022, and the seedball and traditional landscaping plots were planted a couple weeks later. Plots were carefully measured and marked with permanent stakes, and string was used to temporarily delineate the 2 x 2 meter subplots to assist with planting the Miyawaki-inspired plots. Both planting days were gorgeous fall days and we had a great time planting and thinking about what our mini-forests might look like in 10 or 15 years.
Prior to mulching and planting we had soil tests performed to obtain baseline information on pH, nutrients, and carbon composition. We also collected soil samples for potential future analyses and measured soil respiration rates. All measurements were made on each of the four plots. The length of each planted sapling was also measured so that future estimates of growth rates would be possible. Trail cameras were mounted behind each plot to capture wildlife activity (and assess deer herbivory!).
Check out the slide show below for a photographic tour through our project!
Lauren Hagerty, Isabella Jacavone, Tessa Dougan, and Esther Cisneros-Ramos getting ready to establish the plots 
Delineating the plots 
Collecting soil samples 
Collecting soil samples 
Collecting soil samples 
Collecting soil samples and measuring soil respiration 
Ready to measure soil respiration 
Measuring soil respiration 
Measuring soil respiration 
Measuring soil respiration 
Collecting soil 
Measuring soil respiration 
Mulch delivery 
Spreading the mulch. So much work…and we smelled for the rest of the day. 
Spreading mulch 
Discussing our planting strategy 
Spreading mulch 
Delineation of the subplots 
Subplot delineation and plant inventory. 
Saplings 
Saplings 
Saplings 
Measuring the seedlings 
Measuring the seedlings 
Measuring the seedlings 
Arranging the miyawaki-style plot with seedlings 
Planting the miyawaki-style plot with seedlings 
Planting the miyawaki-style plot with seedlings 
A sapling in ground! 
Saplings in the Miyawaki-style plot 
Planting completed. Louise starts to work on the deer barrier. 
Saplings in the Miyawaki-style plot 
Excited to make seed balls! 
Making the seedballs 
Making the seed balls 
Making the seed balls 
Seed balls color-coded by species 
Arranging the seed balls for planting 
Arranging the seed balls for planting 
Arranging the seed balls for planting 
Planting seed balls 
The traditional planting plot after planting 
Installing a camera 
Celebrating our hard work with a nut-sampling party
What’s next?
We wait patiently! Hopefully this spring we will observe our planted forests come to life. Once the plots start growing, data collection comparing tree growth rates, plant diversity, wildlife usage, soil composition, and carbon sequestration will begin.
Acknowledgements
Financial support for this project came from Northhampton County, via a consulting collaboration with Louise Bugbee, Imsuppan, LLC. Additional support came from the Earth & Environmental Science Department at Lehigh University. We had considerable help from Bryan Cope (Superintendent, Northampton County Parks and Recreation) and the Park Staff at Wayne Grube Memorial Park in the preparation of the plots and the transport of mulch. Several other volunteers helped with the spreading of mulch and the planting, including Jim Wilson (Recreation Specialist, Northampton County Parks and Recreation), Eyler Booth, and others. Free seedlings were obtained from the 10 Million Trees PA Partnership and the remainder were sourced from Zach Elfers, Future Forest Plants.
State of the Lehigh Experimental Forest, 2017
General ecology (EES-152) students have finished resurveying a portion of the Lehigh Experimental Forest, assessing changes in species mortality and recruitment since 2013. A total of 1174 trees were inventoried and measured from across the forest the last two years, representing more than 1/2 of all trees originally tagged in 2013. In the four years since 2013, 167 of these 1174 trees have died (~14%) and only eleven new trees have established in the study area (<1%). Data for the dominant tree species are shown in the plot below.
Abundance, mortality, recruitment, and the net percentage change of tree/shrub species in the Lehigh University Experimental Forest, 2013-2017. Relative frequency data are from 2013 (M. Spicer, MS thesis 2014) and indicate the percent of each species present (based on a total of 1174 trees). Total mortality and recruitment for each species with greater than 10 individuals are shown as percentages. Species are arranged from those undergoing substantial declines in abundance at the top to those that have increased in abundance on the bottom.
We will use these data to discuss processes controlling forest dynamics as the semester progresses. However, for now, students should answer the following questions:
- What factors might have caused the differences in mortality among species?
- Develop a hypothesis to explain the lack of recruitment for most tree/shrub species. Then do some research on the two tree species that have successfully recruited and those species that have not. Are there species traits that are common to successful and unsuccessful recruiters? Are these traits consistent (or inconsistent) with what you might predict from your hypothesis?
- What does the pattern of mortality and recruitment suggest about the future of the Lehigh Experimental Forest? Assuming the rates of total tree recruitment and mortality are representative of future years, when will there be less than 100 trees in this forest? In 2013, there were ~2000 trees in the forest so you can use that as your starting number. Show your work and describe how you arrived at your estimate. Do you think this scenario is likely? Why or why not?
State of the forest, 2016
General ecology (EES-152) students have finished resurveying a portion of the Lehigh Experimental Forest, with the goal of assessing changes in tree growth, mortality, and recruitment since 2013. A total of 690 trees were measured from across the forest, representing more than a 1/4 of all trees. In the three years since 2013, 70 of these 690 trees have died and only three new trees have established in the study area. Data for the dominant tree species are shown in the plot below.
Tree abundance, mortality, recruitment, and growth rates in the Lehigh University Experimental Forest, 2013-2016. Relative frequency data are from 2013 (M. Spicer, MS thesis 2014) and indicate the percent of each species present (based on a total of 690 trees). Total mortality and recruitment across the time period are shown as percentages. The average increase in basal area of individuals of each species is shown, with the mean value for all species indicated with the vertical dashed line. Total change in basal area for each species, incorporating mortality losses and basal-area gains, is also shown.
We will use these data to discuss the processes controlling forest dynamics as the semester progresses. However, for now, students should answer the following questions:
- The dbh measurements were converted into estimates of area, assuming that each tree was a perfect circle in cross-section. Why do you think basal area was used to compare growth rates among the different species? Why was this expressed as the average change in basal area per tree? What factors might have caused the observed differences in radial growth among species?
- What does the pattern of mortality and recruitment suggest about the future of the Lehigh Experimental Forest? What factors might have caused the differences in mortality among species during these two years? What factors might be contributing to the lack of new tree recruitment in the forest?
- Assuming the rates of total tree recruitment and mortality are representative of future years, when will there be no trees left in this forest? In 2013, there were ~2000 trees in the forest. Show your work and describe how you arrived at your estimate. Do you think it is likely that the trees will really be gone by this time? Why or why not?
- Which species had both very high mortality and very low growth during this time period? Do some research on current threats to this particular species, and summarize your research in a short paragraph.
First resurvey of the Lehigh Experimental Forest
Growth, mortality, and recruitment (shown in red) of dominant tree species in the Lehigh Experimental Forest from 2013-2015. Average tree size and numbers of indivduals included in the survey shown in blue.
Taking inventory of the forest, 2015.
Students in general ecology (EES-152) resurveyed a portion of the Lehigh Experimental Forest, to assess changes in tree growth, mortality, and recruitment since 2013. No new trees greater than 1.4 m high were documented, and both growth and mortality varied considerably among species. Over 500 trees were measured, and the plot above shows data for the dominant trees (those with >15 individuals included in the survey).
We will use these data as a springboard for discussion of processes controlling forest dynamics, and will examine some of these issues in greater depth during our discussions and future lab activities.
For now, students should answer the following questions:
1. The dbh measurements were converted into estimates of area, assuming that each tree was a perfect circle in cross-section. Why do you think basal area was used to compare growth among the different species? Why was this expressed as the average change in basal area per tree, as opposed to the total change in basal area for all individuals of the species? What factors might have caused the observed differences in radial growth among species?
2. What does the pattern of mortality and recruitment suggest about the future of the Lehigh Experimental Forest? What factors might have caused the differences in mortality among species during these two years? What factors might be contributing to the lack of new tree recruitment in the forest?
3. Which species had both very high mortality and very low growth during this time period? Do some research on current threats to this particular species, and summarize your research in a short paragraph.
Presque Isle Exploration (Pymatuning Wetlands 2015, Day 14)
Marching out to Gull Point on Presque Isle. The trail was a bit washed out, but not an obstacle for these wetlanders.
The Pymatuning Wetlanders visited Presque Isle today, where we observed coastal processes and successional change. After a stop at the Tom Ridge Environmental Center, explored the peninsula to observe coastal wetlands and processes. This included a hike out to Gull point, located at the tip of the peninsula, to observe the youngest landscape and wetlands. We did some wading in Lake Erie to cool off, had lunch on the beach, on our way home we stopped for the long-promised ice cream. A fun day before tomorrow’s final exam.
Some young ponds on Gull Point, the youngest portion of Presque Isle.
It is vey hard to determine which one does not belong….
Dead Phragmites on Presque Isle. They are trying hard to get rid of it.
Among The Stately Trees 