Survival and growth of 20 species of trees and shrubs on Appalachian surface mines

The Forestry Reclamation Approach recommends that 2 tree types be planted during reclamation reforestation projects to reclaim the site to a forest ecosystem. Late succession trees such as oaks (Quercus), maples (Acer), poplars (Liriodendron), and pines (Pinus) are planted as crop trees, but understory trees and shrubs of native forests are not often planted or planted in fewer numbers, so less is known of their survival and growth in mine soils. This study evaluated survival and growth of 20 species of small trees and shrubs planted on 4 surface mines in Appalachia. Survival and height were determined at the end of the first year after planting, and then again after 7 or 9 years depending on the site's planting date. Soil properties varied across sites with pH ranging from 3.8 to 7.5 and percent soil material (≤2 mm) ranging from 58% to 82%. Extracted nutrient concentrations (P, K, Ca, Mg, and S) were also variable across sites. Five of the 20 species had survival ≥50% (black chokeberry [Aronia melanocarpa Michx.], black cherry [Prunus serotina Ehrh.], Washington hawthorn [Crataegus phaenopyrum L. f.], nannyberry [Viburnum lentago L.], and hazelnut [Corylus avellana L.]). Eleven species survived from 37% to 47% and 4 species experienced ≤30% survival. Almost all species at least doubled their height and many increased 3 to 4 times above their initial planted seedling height. The species that survived ≥50% are recommended for forestry plantings and those that survived ≥37% may be considered as candidates for planting during forestry reclamation.

. The Forestry Reclamation Approach (FRA) was developed by a multidisciplinary group of researchers, industry, and regulatory personnel (foresters, soil scientists, ecologists, hydrologists, and economists) with the goal to provide best practices that will successfully establish commercially valuable tree species on mined sites and to assure their rapid growth and development.
The FRA is described in five steps .
Step 1 of the FRA encourages creating a suitable rooting medium that is no less than 1.2 m deep and composed of topsoil, weathered sandstone, and/or the best available material.
Growth media with low to moderate levels of soluble salts, equilibrium pH of 5.0 to 7.0, low pyritic sulfur content, and textures conducive to proper drainage and nutrient holding capacity are preferred . Coarse fragment content should be less than 50% with the majority of the soil material passing through a ≤2 mm sieve. A mixture of brown (weathered) and gray (unweathered) sandstones, when loosely graded, can form a soil medium suitable for trees Zipper, Burger, Barton, & Skousen, 2013).
Step 2 of the FRA proposes that operators should loosely grade the topsoil or topsoil substitutes because excessive soil compaction can have a major negative effect on survival and growth of trees (Sweigard et al., 2007).
Step 3 recommends using ground covers that are compatible with growing trees. Ground cover vegetation used in reforestation must control erosion but not out-compete tree seedlings for light, water, and space. Ground covers should include grasses and legumes that are slow-growing, have sprawling growth forms, have low nutrient requirements, and are tolerant of a wide range of soil conditions (Burger et al., 2009;Franklin, Zipper, Burger, Skousen, & Jacobs, 2012). Steps 4 and 5 endorse planting early succession species and commercially valuable crop trees in the correct manner on reclaimed sites (Davis, Burger, Rathfon, Zipper, & Miller, 2012;Davis, Franklin, Zipper, & Angel, 2010).
Small trees and shrubs that are present in understories of Appalachian forests and that provide important wildlife benefits and ecosystem services are not often planted in surface mine reforestation projects or only sparsely planted. Reclamation planners in the past have focused on planting commercially valuable, late successional trees on mined lands such as oaks (Quercus), maples (Acer), poplars (Liriodendron), and pines (Pinus), with less emphasis on forest ecosystem restoration that may include planting a wide variety of understory species. The late succession crop trees generally comprise more than 90% of the planting stock. The remaining 10% of the trees planted are often black cherry (Prunus serotina Ehrh.), dogwood (Cornus) eastern redbud (Cercis canadensis L.), and black locust (Robinia psuedoacacia L.). Davis et al. (2012) acknowledge that more than 100 tree and shrub species grow in Appalachian forests and they recommend planting a variety of trees and shrubs in mined land reforestation projects in addition to crop trees. Among the tree and shrub species recommended by these scientists are eastern redbud, gray or flowering dogwood (Cornus), American hazelnut (Corylus americana L.), green hawthorn (Crataegus viridis L.), common persimmon (Diospyros virginiana L. f.), and elderberry (Sambucus canadensis L.). Because these small understory tree and shrub species are not commonly planted on reclaimed surface mines, less is known about their seedling survival and growth. Thus, information is needed about which early succession or wildlife habitat woody species should be planted with commercially valuable crop trees in Step 4 of the FRA.
Early succession trees and shrubs are ecological enhancers that provide many benefits to the ecosystem. When soil or soil substitutes have been placed in a noncompacted manner and the herbaceous ground cover has been reduced during reclamation, native and unplanted species from surrounding forests will slowly colonize and reinhabit the site. But foresters acknowledge that planting a diversity of tree types at the start of forest re-establishment will more rapidly develop the functional and structural diversity of the ecosystem (Aerts & Honnay, 2011;Cardinale, Palmer, & Collins, 2001). The intent of FRA reclamation is to develop a forest plant community with all the ecological components necessary for maximizing multiple use of the reclaimed area (MacDonald et al., 2015;Zipper et al., 2011). A diverse plant community composed of early succession and late succession species enhances the wildlife habitat potential, recreational, aesthetic, and productive value of the reclaimed land (Burger, 2011).
The objective of this study was to evaluate the growth and survival of 20 small tree and shrub species planted on four surface mines in 2008 and 2010. Site characteristics and soil properties at each site were also determined. The results of this research will help foresters and reclamation planners select additional woody species that could be planted during reclamation in order to enhance ecosystem development and services.

| Experimental design
The experiment was a completely randomized block design. At each site, four blocks, each measuring 0.42 ha (4,160 m 2 ), were established ( Figure 2). Two blocks were located on east-facing slopes and the other two blocks on west-facing slopes. In each block, 20 plots, each measuring 208 m 2 (14.4 m × 14.4 m), were delineated and one of 20 different small tree or shrub species (Table 1)  The species selected for transplanting included understory species in Appalachian forests, and all provide important ecosystem services such as wildlife food (browse, nuts and seeds, insect pollinators) and habitat (bird nesting, and cover and shelter), and structural, functional, and aesthetic diversity in forests (Table 1).  Slope steepness was determined in each plot with a handheld clinometer and averaged for each site. Herbaceous ground cover percentage was visually estimated in ¼-m 2 quadrats. The quadrat was randomly placed in three places in each species plot, and the herbaceous species and their cover in the quadrat were recorded.

| Planting and first year monitoring
Soil sampling and analysis were performed to determine mine soil properties. Soil samples were collected to a depth of 15 cm at three random locations in each plot and composited into one sample to represent the soil characteristics in each species plot. Therefore, 20 soil samples were collected for each block, and 80 samples were collected per site for a total of 320 soil samples for analysis. Soil samples were air-dried, weighed, and sieved into coarse and fine soil fractions (material passing through a 2 mm sieve), then weighed to determine the soil percentage. A subsample of the soil fraction was analyzed for chemical properties including soil pH using a 1:1 soil : water ratio and soluble salts (as electrical conductivity using 2:1 soil : water ratio). Phosphorus (P), potassium (K), magnesium (Mg), calcium (Ca), sodium (Na), and FIGURE 2 An example of the randomized complete block design used at each site. The block design here is that used at ICG. Numbers within plots for each block represent different tree and shrub species (Table 1). Twenty-five individuals of each species were planted in each plot on 2.4 × 2.4 m spacing

| Statistical analysis
The survival data were left-skewed, and transformations did not correct the lack of normality in the data. Therefore, the data were treated as categorical, and Mantel-Haenszel frequency analysis was used. Chi-square analysis of survival by species was used to compare survival by year (2015 and 2016). The effect of aspect on survival was explored through frequency analysis (chi-square) for all species combined and separately by species and by site.
Repeated frequency analysis using the ridit scores of the Mantel-Haenszel procedure measuring the nonzero correlation of age and survival was done to test the effect of age on survival.
For plant height, a repeated measures ANOVA was performed.
Age was used as the repeated measure factor in a model with age, site, aspect, and their interactions as independent variables on height. After finding significant main effect(s) or interaction(s) in the repeated measure ANOVAs, Tukey-Kramer adjustments were applied to multiple comparisons to control type 1 errors. Statistical analyses were performed using SAS 9.4 software (Statistical Analysis System, 2011). Significance for all tests was an α of ≤0.05.

| Site and soil properties
Mine soil composition at these sites was a mixture of brown and gray sandstone substrate. The brown sandstone material was found from the surface to a depth of about 10 m, whereas gray sandstone was found below 10 m. The sandstone mixtures were developed when materials from both depths were used to rebuild the landscape and graded at the surface with bulldozers for reclamation. Site and soil properties varied across the four sites (Table 2) due to differences in geology, mining and reclamation techniques, and mixing of materials. Elk Run and Fola had a lower average slope (15%) compared with about 22% at Hobet and ICG. The research plots had a wide range of herbaceous ground cover from 0% to 100%, which averaged 27% at Hobet to 67% at Fola. Ground cover species that were commonly found on these sites were a mix of seeded and ICG had the lowest percent soil material (≤2 mm) at 58%, whereas the other sites had higher levels from 66% to 82% (Table 2). A higher level of soil material usually improves water relations and nutrient holding capacity for plants (Haering, Daniels, & Galbraith, 2004 (Table 3).
Significant differences in survival were found among species across the four sites ( Although tree and shrub survival was generally greater on west-facing compared with east-facing aspects in this study, the results were highly influenced by four species with much higher survival on the   Seedling survival of small trees and shrubs in this study were slightly lower than seedling survival of oaks (Quercus spp.), poplars (Liriodendron and Populus spp.), and maples (Acer spp.) in other reforestation studies. In this region, survival of planted oak and poplar seedlings is usually between 80% and 90% after the first year (Casselman, Fox, Burger, Jones, & Galbraith, 2006;Chazdon, 2008;Koropchak, Zipper, Burger, & Evans, 2013;Skousen et al., 2009). Survival after 7 or 9 years across all species in this study was 40%, which was generally lower (Emerson, Skousen, & Ziemkiewicz, 2009;Skousen et al., 2009)   Mean survival after the first growing season (Age 1) and after 9 years (Age 9) on the sites planted in 2008 (Elk Run and Hobet), and after the first growing season (Age 1) and after 7 years (Age 7) on the sites planted in 2010 (ICG and Fola) Note. The probability of significant difference (p value ≤ .05) is shown for mean survival for each species between Ages 1 and 9 for Elk Run and Hobet, and between Ages 1 and 7 for ICG and Fola.
*Data denote significant mean survival differences for that species between Ages 1 and 9 or 7.
The absence of a tree canopy in the study areas, which is an important growth requirement for some small tree and shrub species, appeared to affect survival in this study. Species having a partial-shade requirement such as pawpaw and flowering dogwood had poor survival. Shading would be beneficial to many of these species by decreasing the amount of solar radiation received, decreasing soil temperature, and increasing the amount of soil moisture. Most species needing shade also require moist soil conditions and organic matter in soils to thrive, both of which are deficient in Appalachian mine soils.
Three species planted in this study were also monitored for survival in a study by Wilson-Kokes et al. (2013) on reclaimed mines close to this study. They found after 8 years that black cherry survival was 11%, dogwood was 44%, and eastern redbud was 33%, all of which were lower than the 60% average survival of red oak (Quercus rubra L.), tulip-poplar (Liriodendron tuliperfa L.), white ash (Fraxinus americana L.), and white oak (Quercus alba L.) in their study. In this study, black cherry survival was 55%, dogwood was 10%, and eastern redbud was 45%, which indicates the wide variation in survival of trees and shrubs on reclaimed sites in Appalachia.

| Tree and shrub height
Significant differences in height across sites in 2016 were found for 13 species ( The effect of aspect on average height of these species was mostly insignificant (data are not shown but are available in Monteleone, 2017). Three species (common apple, flowering dogwood, and Washington hawthorn) had significantly greater height on west-facing aspects than east-facing aspects, and no species grew significantly taller on east-facing aspects. As mentioned, the soil conditions of east-facing aspects are generally acknowledged to be better for survival and growth for most woody species because these aspects normally result in cooler and wetter soil conditions than west-facing aspects. But this aspect effect was not reflected by growth of these species on our sites.
As expected, trees and shrubs on these sites showed substantial growth between the first year after planting (Age 1) and the next sampling time in 2016 ( Generally, the growth of most of the species was not associated with the silvical characteristics reported here (Table 1) or in the literature (e.g., Burns & Honkala, 1990). For example, blueberry prefers acidic soils and does not thrive in alkaline conditions (Haynes & Swift, 1985). Interestingly, the most acidic site, Fola, had the poorest blueberry survival. Similarly, some of the species that tolerate drier soil conditions should have grown well in this study but grew poorly or moderately well (i.e., eastern redbud). Drought tolerance is often reflected by a conservative growth strategy, where tolerance provides improved survival over less tolerant species, and may be partially responsible for the limited number of height differences among species.
Our objective was to evaluate which species from a list of 20 would perform well when planted on four reclaimed mine sites 7 and 9 years after planting. This study demonstrated a range of survival from 0% to almost 60% for this broad assortment of trees and shrubs.
Fourteen of the 20 species grew three to four times over their initial planted seedling height. The results of a study like this provide important information about collective success of planted trees and shrubs on several sites with varying site and soil conditions. Those species that averaged good survival and growth across these four sites provides some confidence that they may be successful in establishing and growing when planted at other reclaimed sites. These results will help reclamation planners select appropriate small trees and shrubs for reforestation projects on reclaimed lands. Further monitoring of survival and growth of these small trees and shrubs here and at other reclaimed reforestation projects will provide more information to refine the selection of species.

| SUMMARY AND CONCLUSIONS
The survival of small trees and shrubs on surface mines in this study averaged 40% after 7 and 9 years, which was lower than that of reforestation plantings with commercially valuable trees on reclaimed land (~60%) after 8 years (Wilson-Kokes et al., 2013). Five species (black cherry, Washington hawthorn, black chokeberry, hazelnut, and nannyberry) out of the 20 included in this study had ≥50% survival 7 or 9 years after planting. These species successfully established and persisted in this growing environment and are recommended for future reclamation reforestation plantings. They all produce benefits for wildlife food and habitat. Eleven other species survived from 37% to 47% and could be considered as candidates for reforestation plantings. Four species in this study had poor survival (≤30%) and included two species with ≤10% survival (e.g., pawpaw and flowering dogwood).
In reclamation plantings, mortality of planted seedlings is often very high due to planting stress, harsh soil conditions, and herbivory.
But after the first year, seedling mortality rates tend to decrease from that experienced during the first year . Therefore, selection of high-surviving plants for reforestation projects would provide better success and greater benefits long-term. The results of this study will aid reclamation planners in selecting species for future reforestation plantings. Planting of the better performing species in this study including black cherry, choke cherry, eastern redbud, serviceberry, Washington hawthorn, wild plum, black chokeberry, gray dogwood, hazelnut, highbush cranberry, and nannyberry would potentially encourage and accelerate ecological succession and benefit wildlife. Many of the more successful species were from the Rosaceae family that produce food and cover for wildlife. Some species had poor survival after 7 or 9 years on these sites that indicated that planting these species may not be practical and a waste of resources. Additionally, to increase the survival of the species that are more adapted to shade and other more mature soil conditions and to save money, it may be advisable to delay transplanting of these species or to simply wait for natural recruitment and succession after a canopy has been established from the planted trees. Species that prefer moist soil conditions and shade, such as pawpaw, should not be included in these plantings.

ACKNOWLEDGMENTS
Appreciation is expressed to Williams Forestry and Associates includ-