• Rapid growth and canopy closure by conifers resulted in higher canopy interception rates than those observed in deciduous trees, both on reforested mine sites and in natural forests.

  • Results from this study show the importance of species survival rates on interception and that low survival rates could negatively impact, or at least delay, the hydrologic recovery of a reclaimed area and potentially threaten landscape stability.

  • Use of the forestry reclamation approach can help restore hydrologic function on reclaimed surface mines.

The Appalachian Region is a rich forested ecosystem that has been impacted by coal mining. The Surface Mining Control and Reclamation Act of 1977 was enacted to resolve many of the environmental problems caused by surface mining. Reclamation practices resulted in excessive soil compaction and use of nonnative grasses and shrubs that have altered hydrologic processes. The Forestry Reclamation Approach (FRA) is a best practice for reestablishing forested ecosystems on mined lands in Appalachia. This project evaluated precipitation throughfall in reforested 10- and 20-year-old FRA sites and unmined 100-year-old forest stands as a metric for evaluating the return of forest hydrologic function after reclamation. Stands of coniferous and deciduous trees were evaluated independently for each age class. Throughfall rates were significantly impacted by tree type and age. Throughfall in coniferous trees was less than in deciduous trees, and throughfall in the 10-year-old deciduous trees tended to be highest. Throughfall was also significantly impacted by storm characteristics. Higher rainfall depth and longer duration resulted in significantly larger throughfall depths under both coniferous and deciduous stands, whereas increased intensity increased throughfall depths for the 10- and 100-year-old plots, but not for the 20-year-old plots. As canopy closure occurs in young FRA forests, throughfall rates resemble those reported for young, naturally regenerating forests in the region. Results may help guide management of forested watershed strategies to reduce surface runoff and local flooding on reclaimed surface mined lands.

Mixed hardwood forests of Appalachia provide valuable ecosystem services, not the least of which is maintenance of freshwater resources. Unfortunately, Appalachian forests are threatened by a variety of pressures, including climate change, invasive species expansion, and resource extraction. Surface mining for coal is one of the most important drivers of land use change in the Appalachian Region, reducing native-forest cover, causing forest fragmentation, and eliminating intact soil (Wickham et al. 2013). In addition to these significant impacts to terrestrial ecosystems, surface mining impairs the region’s valuable water resources. A widening body of literature implicates surface mining in long-term water-quality impacts including elevated dissolved salt concentrations, changes in solution pH (Bernhardt et al. 2012; Lindberg et al. 2011), and loss of sensitive macroinvertebrate assemblages (Negley and Eshelman 2006; Pond 2010; Pond et al. 2008). Surface mining has also been linked to alterations in hydrologic function, typically related to a combination of drastically reduced infiltration rates (due to heavy soil compaction), decreased saturated hydraulic conductivity, reduced evapotranspiration rates, and increased peak flows (Negley and Eshelman 2006; Taylor et al. 2009a; Williamson & Barton 2020).

Numerous environmental problems, such as landslides, erosion, flooding, instream sedimentation, and poor water quality on surface mines, prompted the US Congress to enact the Surface Mining Control and Reclamation Act (SMCRA) of 1977. As a result of SMCRA, regulators focused on mechanical grading to achieve land stability and return to approximate original contour while reducing erosion and subsequent sedimentation of streams (Adams 2017). This resulted in highly compacted lands that were planted with fast-growing, and often nonnative, herbaceous vegetation (Graves et al. 2000; Zipper et al. 2011). Tree survival on compacted lands, and in the face of fierce competition from herbaceous vegetation, was poor (Skousen et al. 2009; Zipper et al. 2011). Grasses, which grow more quickly than trees, allowed operators to meet reclamation bond requirements within the required short term (e.g., 5-year period, 90% ground cover; Environmental Protection Agency [USEPA] 2011; Holl et al. 2009; Roberts et al. 1998). Since the passage of SMCRA, more than 600,000 ha (6,000 km2) of land in Appalachia has been mined (Pericak et al. 2018; Zipper et al. 2011), resulting in a greater than 9% change in land cover, predominately associated with an increase in surface coal mining, a decrease in forested area, and an increase in grasslands-shrubs (Sayler 2016; Townsend et al. 2008).

The solution to some of these problems may lie in improved mine reclamation techniques. The forestry reclamation approach (FRA) is a prescription for reclaiming mined lands in such a way as to promote tree growth while addressing concerns of land stability, erosion, and sedimentation (Adams 2017, and chapters therein). Research indicates that FRA is effective at promoting tree growth and survival (Dement et al. 2020; Sena et al. 2015; Skousen et al. 2009), accelerating soil development (Miller et al. 2012; Sena et al. 2018), improving water quality (Agouridis et al. 2012), and restoring hydrologic response in the form of reduced peak flows and runoff volumes (Sena et al. 2014; Taylor et al. 2009a, b).

Replacing forest with grass, even without the complicating factors of soil compaction, alters ecosystem functions such as nutrient cycling, water storage, carbon sequestration, habitat provision, and temperature moderation (Osborne & Kovacic 1993; Saunders et al. 1991; Zipper et al. 2011). Hydrologic processes such as interception, infiltration, saturated hydraulic conductivity, evapotranspiration, and runoff are altered due to changes in vegetation type (Mao & Cherkauer 2009; Negley & Eshelman 2006; Sena et al. 2014); these effects are exacerbated by soil compaction from heavy-equipment operations used to regrade the mine spoil and stabilize land to reduce landslides and erosion (Hoomehr et al. 2015; Jorgensen and Gardner 1987; Simmons et al. 2008). Forests play a critical role in managing how water moves through the landscape. A forest canopy has the potential to intercept nearly 25% of the precipitation associated with a storm depending on tree age and type; storm depth (precipitation in millimeters), duration, and intensity; and wind speed (Miralles et al. 2010). Llorens et al. (1997) found that a 33-year-old pine forest (Pinus sylvestris L.) in the Mediterranean mountains of Spain intercepted, on average, 24% of rainfall over a 30-month period, with a maximum rate of 49%. Bryant et al. (2005) measured interception rates ranging from 17 to 22% in five forest types (pine, mixed forest, lowland hardwood, pine plantation, and upland hardwood) in Georgia. Habib et al. (2012), using the Climate Prediction Center morphing technique precipitation product and the Gash analytical model, determined that needleleaf forests intercepted 22% of rainfall followed by broadleaf deciduous forests at 19% and broadleaf evergreen forests at 13%. In a review of the literature, Levia and Frost (2006) noted that 77–83% of rainfall becomes throughfall in deciduous forests, whereas 47–91% of rainfall becomes throughfall in coniferous forests.

Because of the short time that FRA has been in practice (<25 years), no information exists on the role that the closing of the canopy, as trees mature, may have on throughfall rates. At what point do trees planted in accordance with FRA guidelines begin to intercept rainfall at rates similar to mature forests? How does tree type, coniferous versus deciduous, influence canopy interception and throughfall on reforested mine lands in Appalachia? This latter question has important implications for watershed managers considering climate impacts such as the northern migration of trees species (Lafleur et al. 2010; Woodwall et al. 2009) and changing rainfall patterns (Dore 2005). For instance, in July 2022 a historic rain event (1,000-year event) in eastern Kentucky led to floods that resulted in the fatalities of 39 people and more than 10,000 homes were reported damaged or destroyed (National Public Radio 2022). The area impacted by the flooding had a high concentration of surface mines, and there were reports that the mining likely contributed to the flooding due to poor reclamation and lack of vegetation (NBC News 2022). If climate change is likely to increase the frequency of damaging floods, and mined landscapes are thought to contribute to the flooding response, additional information on the role of mining and reclamation on hydrologic responses to precipitation events is needed. As such, the objective of this study was to measure canopy interception and throughfall, as a component of the hydrologic cycle, in reforested stands established using FRA. We examined 10- and 20-year-old FRA stands and compared them to 100-year-old, unmined, forested control stands in coniferous- and deciduous-dominated plots to evaluate the influence of age and tree type on throughfall. Understanding how water disposition varies with tree age and type is critical to characterizing the environment created by individual reclamation strategies and to managing postmined landscapes on which watershed function determines water availability, water quality, and ecology.

Study Site

Research was conducted at the University of Kentucky’s Robinson Forest and the Starfire Mine, both of which are in southeastern Kentucky in the Cumberland Plateau section of the Appalachian Plateaus province of the Appalachian Highlands (Fig. 1; Woods et al. 2002). These adjacent study areas are situated in the Appalachian mixed mesophytic forest region (Braun 1950) that is characterized by hills and valleys, with elevation differences ranging from 385 to 610 m (Smalley 1986). From 1971 to 2018, the average annual rainfall at Robinson Forest was 1,164 mm (Sena et al. 2021). The climate at Robinson Forest is humid and temperate (Cherry 2006; US Department of Commerce [USDC] 2002). Weather is highly variable throughout the year, with long periods of low-intensity rainfall during the winter and high-intensity storm events that typically occur during the summer and early fall (Husic et al. 2019).

Fig. 1

Tree plots are in eastern Kentucky near the intersection of Breathitt, Perry, and Knott counties. Plots were located in the Starfire and Laurel Fork surface coal mines and Robinson Forest. The traditional grassland plot is adjacent to the 20-year-old plots. WS = weather station. Base map from US Geological Survey National Map with US Department of Agriculture National Agricultural Imagery Program land cover

Fig. 1

Tree plots are in eastern Kentucky near the intersection of Breathitt, Perry, and Knott counties. Plots were located in the Starfire and Laurel Fork surface coal mines and Robinson Forest. The traditional grassland plot is adjacent to the 20-year-old plots. WS = weather station. Base map from US Geological Survey National Map with US Department of Agriculture National Agricultural Imagery Program land cover

Close modal

Robinson Forest (37.450°N, 83.183°W) is an approximately 6,000-ha research and teaching forest that consists of a main block approximately 4,200 ha in size and seven other discontinuous areas. The forest is characterized by steep side slopes, averaging 45%, coupled with a hydrologically restrictive geologic layer consisting of interbedded sandstone, siltstone, shale, and coal (Hinrichs 1978). Robinson Forest was clear cut between 1890 and 1920, with subsequent forest regeneration dominated by oaks (Quercus spp.), hickories (Carya spp.), and yellow poplar (Liriodendron tulipifera L.; Overstreet 1984). During the early- to mid-1990s, a portion of Robinson Forest (ca. 810 ha) of the Laurel Fork Tract was surface mined for coal, resulting in the creation of considerable flat lands reclaimed as pasture with several valley fills. Mining of the site ceased in the early 2000s, and much of the site has received bond release. Since 2005, several research projects have been conducted on the mined tract to examine forestry reclamation techniques.

Starfire Mine (37.400°N, 83.117°W) is in Perry and Knott counties. Since the early 1980s, the mine has operated as a mountaintop removal operation whereby multiple coal seams were extracted using dragline and truck-shovel techniques (Barton et al. 2017). Most of this area has been reclaimed as hay and pastureland on compacted spoils by using traditional techniques.

Treatments

Treatments were composed of plots (10 m × 10 m) on 10-, 20-, and 100-year-old coniferous and deciduous forest types located at either Robinson Forest or Starfire Mine. Tree plots were established as monoculture plantations in FRA forests or in an area dominated by either a single conifer or deciduous species in the control forest. Trees in the plantations were planted on an approximately 2-m spacing. Because of the rough-and-rocky nature of the planting areas, planting density was not uniform. Differing conifer and deciduous species were used because single-species monocultures of the same species were not available across each age class. Replication was limited to the precipitation collectors within each plot (n = 12), and not to multiple plots. A “traditional-reclamation,” grass-only plot was also established at the Starfire Mine. Canopy density was measured in each plot in June 2017 by using a spherical densiometer (Model A, Forestry Suppliers, Jackson, MI). Measurements were made at the plot center facing four cardinal directions (N, E, S, and W) from that point and then averaged.

10-Year-Old Plots

The 10-year-old FRA tree plots (37.428°N, 83.177°W) were established on the Laurel Fork section of Robinson Forest that was mined during the mid-1990s. The relatively flat site was cross ripped with a bulldozer equipped with a 1-m ripping shank (plow) and subsequently planted in plots of either loblolly pine (Pinus taeda [L.]) or northern red oak (Quercus rubra L.) in 2007 (Barton et al. 2008). The coniferous monitoring plot at this site had 25 loblolly pines, whereas the deciduous plot had 10 northern red oaks (Table 1). Autumn olive (Elaeagnus umbellata Thunb.) has invaded the understory of the deciduous plot, but not the coniferous plot.

Table 1

Number of trees and equivalent stem density per plot

Number of trees and equivalent stem density per plot
Number of trees and equivalent stem density per plot

20-Year-Old Plots

In 1997, an experiment was established at the Starfire Mine where 1-ha reclamation cells were created overtop previously reclaimed mined land by using spoil (nonacidic shale and sandstone overburden) from the active mining operation (Angel et al. 2006). Spoil was placed in accordance with FRA by using the strike-off method: spoil is dumped in closely abutted piles approximately 2–2.5 m high and slightly leveled with two passes of a small bulldozer (Adams 2017). The site was planted in individual species plots of eastern white pine (Pinus strobus L.), white ash (Fraxinus americana L.), black walnut (Juglans nigra L.), yellow poplar, white oak (Quercus alba L.), and northern red oak with 1-year-old nursery stock bare root seedlings (Dement et al. 2020). The 20-year-old FRA tree plots (37.410°N, 83.118°W) were established and planted on the strike-off FRA plots at the Starfire Mine. The coniferous plot had 21 white pines, whereas the deciduous plot had 32 white oaks (Table 1).

100-Year-Old Plots

The 100-year-old, forested control plots (37.473°N, 83.153°W) were established in the Little Millseat watershed (77.9 ha), located in the main block of Robinson Forest. Stands in the watershed are typical of second-growth forest in the Cumberland Plateau. Elevation in the watershed ranges from 304 to 451 m. The watershed is characterized by steep slopes (25–60%), a dense drainage network (drainage density of 0.0038 m·m−2), and a narrow valley with well-drained soils (Cherry 2006; Hayes 1991). The coniferous plot was situated on a north-facing toeslope position with a slope of 25% and included 17 eastern hemlocks (Tsuga canadensis [L.] Carrière). The deciduous plot was located on a south-facing side slope position with a slope of 18% and contained 21 white oaks (Table 1).

Throughfall Measurement

Precipitation data were obtained from three University of Kentucky weather stations (Camp, Little Millseat, and Laurel Fork; https://doi.org/10.5066/P9FPLG1O) and the National Weather Service’s Jackson, Kentucky, weather station (JKL; 37.592°N, 83.316°W; https://www.weather.gov/jkl/). The Camp (CWS; 37.461°N, 83.159°W), Little Millseat (LMB; 37.473°N, 83.153°W), and Laurel Fork (LF; 37.413°N, 83.176°W) weather stations are all within a 7.5-km radius of all tree plots (Fig. 1; Sena et al. 2020a). The JKL weather station is 20–25 km from individual tree plots. At the CWS and LMB weather stations, precipitation data were recorded at 15-min intervals, whereas precipitation data were recorded at 60-min intervals at LF and at 5-min at JKL. All University of Kentucky weather stations used a tipping bucket rain gauge linked to a CR10X data logger (Campbell Scientific, Inc., Logan, UT) to collect precipitation data (Cherry 2006). To reduce the effect of local random errors associated with tipping bucket rain gauges, storm event data from all four weather stations were averaged (Ciach 2002). Rainfall events were considered independent storms when a 3-hr or greater separation time was observed.

Twelve platforms (46 cm × 46 cm) were randomly distributed in each 10-m × 10-m tree plot (Bryant et al. 2005; Helvey & Patric 1965; Levia & Frost 2006). Each platform stand was 61 cm above the ground surface. A Rain Collector II tipping bucket rain gauge (Davis Instruments, Haywood, CA) equipped with a Hobo Event Data Logger (Onset Computer Corporation, Cape Cod, MA) was located on the corner of each platform (Fig. 2). Rain gauges were used to record throughfall timing and depth. Every 2 weeks, rain gauges were rotated clockwise 90° around their platform to account for variations in canopy coverage and the potential for bias from individual gauges and to provide a more accurate representation of throughfall throughout the tree plots (Bryant et al. 2005; Helvey & Patric 1965; Levia & Frost 2006; Lloyd & Marques 1988). Although evaporation rates were not measured in this study, other researchers have found that this type of funneled rain gauge loses between 0 and 4% of captured throughfall to evaporation (Thimonier 1998).

Fig. 2

Example of throughfall collector and plot arrangement in a 10-year-old loblolly pine plot (10-year coniferous)

Fig. 2

Example of throughfall collector and plot arrangement in a 10-year-old loblolly pine plot (10-year coniferous)

Close modal

Throughfall data were collected for a 1-year period (from 19 May 2017 to 18 May 2018). For each storm event, throughfall depths from all 12 gauges within each plot were aggregated to determine plot-median throughfall depths. If there was evidence at the time of biweekly data download that a throughfall collector’s performance was compromised due to clogging (e.g., leaf litter accumulation or biofilm blockage of funnel tip) or via wildlife damage to the tipping bucket, data from that collector were not used in computing the plot-median throughfall value.

Canopy interception was calculated as the percent of annual precipitation that was not measured as throughfall (100 × [precipitation − throughfall]/precipitation). Annual precipitation and throughfall data were summed for each tipping bucket, and means for each plot and all weather stations were used for the analysis. Because of logistical constraints, stemflow was not measured.

Statistical Analysis

Repeated measures analyses of variance (ANOVAs) on ranks (a nonparametric analysis of variance using rank-transformed data) were used to compare storm event depths and durations between the weather stations as well as the average of all four weather stations (Johnson et al. 2002; SAS Institute, Inc. 2016; Singh et al. 2013). A generalized linear mixed model (GLMMIX) in Statistical Analysis Software 9.4 (SAS Institute, Inc. 2016) was used to test for differences in natural-log-transformed throughfall depths due to tree age (10, 20, and 100 years) and tree type (coniferous and deciduous). Storm intensity, storm duration, and leaf-on (May–September) and leaf-off (November–February) periods served as covariates. Storm depth was not included in the model because it was highly correlated with storm duration. Storm intensity and duration were modeled as random effects. The Tukey test was used for multiple comparisons between tree plot type and age when significant differences were present (α = .05). Data are available at https://uknowledge.uky.edu/bae_etds/65/.

Canopy Density

The estimated overstory density in the 10-year-old plots was 97% for the coniferous plot and 61% for the deciduous plot. Canopy density in the 20-year-old plots was 94% for the coniferous plot and 96% for the deciduous plot. The 100-year-old plots exhibited overstory densities of 93% for the coniferous plot and 95% for the deciduous plot.

Precipitation Characteristics

In total, 113 storms occurred during the study period. Total rainfall depths for the year ranged from 1,261 mm at LMB to 1,434 mm at LF, whereas rainfall durations ranged from 15 min at CWS to 37 hr at LF. Most storms were small and short (Figs. 3 and 4). Long-term mean annual precipitation for Robinson Forest (1971–2013) was reported to be 1,164 mm (Sena et al. 2021), suggesting that the study period was slightly wetter than normal. Across all weather stations, approximately 85% of the recorded storms had rainfall depths of 20 mm or less and durations of 12 hr or less. The two storms with the highest recurrence interval were close to the 5-year storm recurrence interval (Lexington-Fayette Urban County Government [LFUCG] 2009) at 41.4 and 39.3 mm for 1.11 and 0.96 hr, respectively. The third highest recurrence interval was for a 6.25-hr storm with 57.0 mm of rainfall. Most of the storms were below the 1-year recurrence interval (LFUCG 2009). Table 2 shows the comparison of median storm event depth and duration between the weather stations. The LF weather station had a significantly higher median storm event depth than LMB and a significantly longer median duration than all other weather stations individually and averaged. The LF weather station recorded data at the largest interval, 1 hr, which may account for its difference with regard to median storm duration. The significant linear relation between depth and duration at these weather stations was similar to that described by Austen and Claborn (1971), who noted a correlation of 0.409 between these two parameters for storm events in Lubbock, Texas.

Fig. 3

Distribution of rainfall depths as averaged over all four weather stations for the period from 19 May 2017 to 18 May 2018

Fig. 3

Distribution of rainfall depths as averaged over all four weather stations for the period from 19 May 2017 to 18 May 2018

Close modal
Fig. 4

Distribution of rainfall durations as averaged over all four weather stations for the period from 19 May 2017 to 18 May 2018

Fig. 4

Distribution of rainfall durations as averaged over all four weather stations for the period from 19 May 2017 to 18 May 2018

Close modal
Table 2

Results of comparing annual medium storm event depth and duration from weather stations by using repeated measures ANOVAs

Results of comparing annual medium storm event depth and duration from weather stations by using repeated measures ANOVAs
Results of comparing annual medium storm event depth and duration from weather stations by using repeated measures ANOVAs

Storm events displayed seasonal differences among winter (December–February), spring (March–May), summer (June–August), and autumn (September–November) for duration and average intensity, but not depth (Tables 3 and 4). Storm-event durations were significantly longer during the winter than during the summer. Summer storms were significantly more intense (greater than millimeters per hour) than winter storms. Long periods of low-intensity rainfall during the winter and high-intensity storm events in the summer and early fall are typical for this region (Husic et al. 2019).

Table 3

Results of comparing seasonal variations in storm event durations by using a one-way ANOVA. Values displayed are medians and units are in hours

Results of comparing seasonal variations in storm event durations by using a one-way ANOVA. Values displayed are medians and units are in hours
Results of comparing seasonal variations in storm event durations by using a one-way ANOVA. Values displayed are medians and units are in hours
Table 4

Results of comparing seasonal variations in storm event average intensities by using a one-way ANOVA. Values displayed are medians and units are in millimeter per hour

Results of comparing seasonal variations in storm event average intensities by using a one-way ANOVA. Values displayed are medians and units are in millimeter per hour
Results of comparing seasonal variations in storm event average intensities by using a one-way ANOVA. Values displayed are medians and units are in millimeter per hour

Throughfall Characteristics

Median values and rank-based statistics were used on natural-log-transformed data due to the heterogenous distribution of throughfall within the plots (Table 5), a phenomenon noted previously (Tanaka et al. 2015). All plots showed a significant difference among tipping buckets within the plot. The greatest variability within a single plot type was seen in the 10-year deciduous plots.

Table 5

Results of comparing tipping buckets, within plots, by using repeated measures ANOVAs. Values reported are annual median throughfall depths in millimeters

Results of comparing tipping buckets, within plots, by using repeated measures ANOVAs. Values reported are annual median throughfall depths in millimeters
Results of comparing tipping buckets, within plots, by using repeated measures ANOVAs. Values reported are annual median throughfall depths in millimeters

For the traditional site (grass only), throughfall data were collected for 99 storm events during the period of June 2017 through May 2018. During this period, 1,165 mm of rainfall occurred of which 1,088 mm, or approximately 93%, was recorded by the throughfall collectors.

Storm characteristics had an influence on throughfall volume. Increases in precipitation volume and duration resulted in significantly larger throughfall depths for all plots. An increase in precipitation intensity led to significant increases (p < 0.01) in throughfall depths in the 10- and 100-year plots, but not the 20-year plots.

Treatment Effects

Throughfall was significantly influenced by tree type (coniferous and deciduous) and age (10, 20, and 100 years). Considering all ages, coniferous trees had significantly lower throughfall depths (Table 6). The median throughfall depth (all ages combined) for coniferous trees was 3.8 mm and for deciduous trees was 5.8 mm. The effect of age depended on tree type. For coniferous trees, throughfall depth significantly differed between the 10- and 100-year plots as well as the 20- and 100-year plots, but not the 10- and 20-year plots. For deciduous trees, the 20-year plot significantly differed from both the 10- and 100-year plots; however, the 10- and 100-year plots did not differ.

Table 6

Influence of tree type and age on throughfall depth. Values displayed are annual median throughfall depths

Influence of tree type and age on throughfall depth. Values displayed are annual median throughfall depths
Influence of tree type and age on throughfall depth. Values displayed are annual median throughfall depths

Canopy interception varied with tree type and age (Fig. 5). The highest interception rates occurred with the 10- and 20-year-old coniferous plots, with 44 and 40% of rainfall intercepted, respectively. The lowest interception rate occurred with the 10-year-old deciduous plot that is mixed with a dense understory of autumn olive. At approximately 6%, this plot was equivalent to the grassland plot. Although stemflow was not measured, it has been observed in eastern hardwood forests at rates from less than 1% of total precipitation in an old growth forest (>200 years old) to as high as 8% in a high stem-density-regenerating forest (12 years old; Brantley et al. 2019). As such, true interception rates would be lower.

Fig. 5

Annual percent interception rates for each plot. The numbers indicate tree age in years. C = coniferous and D = deciduous

Fig. 5

Annual percent interception rates for each plot. The numbers indicate tree age in years. C = coniferous and D = deciduous

Close modal

A similar canopy cover (93–97%) was exhibited on all plots except for the 10-year deciduous plot (61%). High mortality of the northern red oak (62%) versus the loblolly pine (37%) likely contributed to the differences in canopy cover observed at that site. Tree volume index (height × diameter2) further highlights the dramatic species differences, with loblolly pine volume (150,000 cm3) more than 18 times greater than northern red oak volume (8,200 cm3) in 2014 (Sena et al. 2020b). This also likely explains why the 10-year deciduous plot exhibited the higher variability of throughfall depths of any plot. Unlike the other plots, the 10-year deciduous plot contained large numbers of invasive shrubs, such as autumn olive, that can impact tree growth and hence canopy cover (Dukes et al. 2009; Orr et al. 2005). Tree density (Table 1) also appeared to have an influence on throughfall depth (Table 6): the deciduous plot with the highest density (20-year deciduous) had significantly lower throughfall than the other deciduous plots, whereas the conifer plot with the lowest density (100-year coniferous) had significantly higher throughfall than the other conifer plots.

Levia and Frost (2006) found storm magnitude and duration were important factors affecting throughfall. Throughfall was significantly affected by storm characteristics as a function of both tree type and forest age. Increases in depth (millimeters of precipitation) and duration resulted in significantly larger throughfall depths under both coniferous and deciduous stands, whereas increased intensity affected the plots differently based on age. Significant increases (p < 0.01) in throughfall depths related to intensity were only observed in the 10- and 100-year plots. Prior research has shown that throughfall rates are related to storm duration and intensity (Levia & Frost 2006). Crockford and Richardson (2000) noted that interception was strongly related to rainfall duration and intensity as higher interception rates were generally associated with low-intensity, long-duration events and low interception rates were often linked to high-intensity, short-duration events. Bryant et al. (2005) noted that short-duration, low-level intensity storms were associated with lower throughfall depths (i.e., higher interception rates) for five forest communities in western Georgia. Miralles et al. (2010) noted that regions such as Scandinavia and northern Canada that exhibit long-duration storm events had higher interception losses than areas dominated by short-duration convective storms.

Throughfall collectors on the traditional grassland site collected 93% of the total precipitation at that location. It is possible that some rainfall was intercepted by the grasses in the summer when vegetation was slightly higher than the top of the collector, although it is also likely that the difference between precipitation and throughfall depths may be attributable to localized (micro)variations in rainfall patterns and losses from evaporation, splash, and wetting of the collector surface (Levia & Frost 2006; Thimonier 1998). In some instances, negative throughfall values were recorded (e.g., throughfall depths exceeding rainfall depths). Similar occurrences were noted in prior throughfall studies with some authors attributing this occurrence to wind (Crockford & Richardson 2000; Levia & Frost 2006; Robson et al. 1994).

Based upon constraints of the experimental design that required the comparison of multiple species of trees, we separated our treatments into a general forest type of either conifer or deciduous. Barbier et al. (2009) reviewed studies that compared rainfall disposition among tree species in temperate and boreal forests to examine tree traits on water budgets. Results from that review showed that evergreens had a lower throughfall than deciduous trees, with an annual difference of 13.9% less total precipitation. They also showed that throughfall declined with the successional status from pioneer to late successional tree species. Our results showed similar trends with regard to throughfall in conifer versus deciduous plots, with a 43% reduction in throughfall in 10-year plots versus a 22% reduction in 100-year conifer plots. However, we found that throughfall volumes were highest in both the 100-year conifer and deciduous plots. The increase in throughfall depth over time and the reduction in differences between vegetation types with age may be attributable in part to self-thinning (McCarthy et al. 1991).

Recognizing that throughfall depths decrease with age could have watershed management implications, particularly in instances where tree stands are not allowed to fully mature, such as with woody biomass production where stands may be harvested between 10 and 20 years of age (Caldwell et al. 2018). Such strategies of sustaining a young tree population may have implications for flood mitigation in disturbed landscapes such as mined lands that are often reclaimed as grasslands on soils that exhibit limited infiltration (Zipper et al. 2011) and high surface runoff peaks and volumes (Guebert & Gardner 2001; Jorgensen & Gardner 1987; Skukla et al. 2004). As demonstrated by Kuczera (1987) and Haydon et al. (1996), interception and water yield are not a linear relationships with forest age. Interception increases rapidly, peaking between 20 and 40 years, before steadily declining. Similarly, water yields are nonlinear and instead have a reversed curve because yields rapidly decrease between 20 and 40 years before steadily increasing.

FRA stands are regrowing similarly to a natural forest stand. For the deciduous plots, low canopy density in the 10-year plot resulted in greater throughfall depths; interestingly, interception rates for the 20-year plot were higher, reflecting more mature forest and canopy closure. Although it is not clear whether invasive species caused poor survival of the 10-year oak trees or whether poor survival allowed incursion of the invasive species, the results from this study show the importance of species survival rates on interception and that low survival rates could negatively impact, or at least delay, the hydrologic recovery of the area, perhaps threatening landscape stability. Bell et al. (2017) concluded that coniferous trees (shortleaf and loblolly pine) experience more rapid growth than deciduous trees (northern red oak, white oak, and chestnut oak) on reclaimed mined lands, a benefit for outcompeting invasive grasses that are common to such areas. Hansen et al. (2015) observed a similar trend at the site where the 10-year plots were established.

FRA is a proven method for reestablishing forested ecosystems on mined lands (Adams 2017). Since 2004, nearly 60,000 ha of mine lands has been reclaimed to forest after FRA (Angel et al. 2015). Results from this study indicate that the trees grown using FRA are following an expected pattern with regard to throughfall (Haydon et al. 1996; Kuczera 1987) and complement prior FRA-focused runoff and infiltration studies (Sena et al. 2015; Taylor et al. 2009a, b). Because mean annual streamflow is inversely related to interception, it is hypothesized that mined watersheds reforested using FRA will exhibit decreased water yields as the forest matures. These patterns suggest that using FRA will aid in the restoration of natural hydrologic function.

Funding for the project was provided by the US Department of the Interior Office of Surface Mining Reclamation and Enforcement’s Applied Science Program (grant S16AC20056). This work was partially supported by the National Institute of Food and Agriculture, US Department of Agriculture, McIntire-Stennis Research Program (accession 1005547). We thank Dwayne Edwards, Sam Austen, Somsubhra Chattopadhyay, and the staff at Robinson Forest for field assistance. Any use of trade, firm, or product names is for descriptive purposes only and does not imply endorsement by the US Government.

Declaration of Interest Statement The authors declare that they have no known competing financial interests, conflicts, or personal relationships that could have appeared to influence the work reported in this paper.

Adams
MB
(
2017
)
The forestry reclamation approach: Guide to successful reforestation of mined lands
(
General Technical Report, Northern Research Station 169
).
US Department of Agriculture, Forest Service
,
Northern Research Station, Newtown Square, PA
. https://doi.org/10.2737/NRS-GTR-169
Agouridis
CT,
Angel
PN,
Taylor
TJ,
Barton
CD,
Warner
RC,
Yu
X,
Wood
C
(
2012
)
Water quality characteristics of discharge from reforested loose-dumped mine spoil in eastern Kentucky
.
Journal of Environmental Quality
41
:
454
468
.
Angel
HZ,
Barton
CD,
French
ME,
Angel
PN
(
2015
)
The Appalachian Regional Reforestation Initiative and Green Forests Work: Bringing back the forest on surface coal mines in Appalachia
.
Journal of the American Society of Mining and Reclamation
4
:
91
101
.
Angel
PA,
Graves
DH,
Barton
CD,
Warner
RC,
Conrad
PW,
Sweigard
RJ,
Agouridis
CT
(
2006
)
Surface mine reforestation research: Evaluation of tree response to low compaction reclamation techniques
.
Pages X–X
. In:
Barnheisel
RI
(ed)
Proceedings, National Meeting of the American Society of Mining and Reclamation
,
7th International Conference on Acid Rock Drainage
,
St. Louis, MO
,
26–30 March 2008
.
American Society of Mining and Reclamation, Lexington, KY pp. 45-58
.
Austen
TA,
Claborn
BJ
(
1971
)
Statistical models of short-duration precipitation events
. Pages
356
365
.
In:
Proceedings of the Symposium on Statistical Hydrology
,
Tucson, AZ
,
31 August–2 September 1971
,
Miscellaneous Publication No. 1275. US Department of Agriculture–Agricultural Research Service
Barbier
S,
Balandier
P,
Gosselin
F
(
2009
)
Influence of several tree traits on rainfall partitioning in temperate and boreal forests: A review
.
Annals of Forest Science
66
:
602
602
.
Barton
CD,
Marx
D,
Sweigard
R,
Barton
W
(
2008
)
Evaluating spoil amendment use and mycorrhizal inoculation on reforestation success in the eastern and western Kentucky coalfields
. Pages
98
111
.
In:
Barnhisel
RI
(ed)
Proceedings, National Meeting of the American Society of Mining and Reclamation
,
Richmond, VA
,
14–19 June 2008
.
American Society of Mining and Reclamation
,
Lexington, KY
.
Barton
CD,
Sena
K,
Dolan
T,
Angel
P,
Zipper
C
(
2017
)
Restoring forests on surface coal mines in Appalachia: A regional reforestation approach with global application
. Pp.
124
-
145
. In:
Bolan
NS,
Kirkham
MB,
Ok
YS
(eds)
Spoil to soil: Mine site rehabilitation and revegetation
.
Taylor & Francis
.
Bell
G,
Sena
KL,
Barton
CD,
French
ME
(
2017
)
Establishing pine monocultures and mixed pine-hardwood stands on reclaimed surface mined land in eastern Kentucky: Implications for forest resilience in a changing climate
.
Forests
8
:
375
.
Bernhardt
ES,
Lutz
BD,
King
RS,
Fay
JP,
Carter
CE,
Helton
AM,
Campagna
D,
Amos
J
(
2012
)
How many mountains can we mine? Assessing the regional degradation of central Appalachian rivers by surface coal mining
.
Environmental Science and Technology
46
:
115
122
Brantley
ST,
Miniat
CF,
Bolstad
PV
(
2019
)
Rainfall partitioning varies across a forest age chronosequence in the southern Appalachian Mountains
.
Ecohydrology
12
:
e2081
.
Braun
EL
(
1950
)
Deciduous forests of eastern North America
.
Hafner Press
,
New York
Bryant
ML,
Bhat
S,
Jacobs
JM
(
2005
)
Measurements and modeling of throughfall variability for five forest communities in the southeastern US
.
Journal of Hydrology
312
:
95
108
.
Caldwell
PV,
Jackson
CR,
Miniat
CF,
Younger
SE,
Vining
JA,
McDonnell
JJ,
Aubrey
DP
(
2018
)
Woody bioenergy crop selection can have large effects on water yield: A southeastern United States case study
.
Biomass and Bioenergy
117
:
180
189
.
Cherry
MA
(
2006
)
Hydrochemical characterization of ten headwater catchments in eastern Kentucky
(unpublished MS thesis)
,
University of Kentucky
,
Lexington
Ciach
GJ
(
2002
)
Local random errors in tipping-bucket rain gauge measurements
.
Journal of Atmospheric and Oceanic Technology
20
:
752
759
.
Crockford
RH,
Richardson
DP
(
2000
)
Partitioning of rainfall into throughfall, stemflow and interception: Effect of forest type, ground cover and climate
.
Hydrological Processes
14
:
2903
2920
.
Dement
WT,
Hackworth
ZJ,
Lhotka
LM,
Barton
CD
(
2020
)
Plantation development and colonization of woody species in response to post-mining spoil preparation methods
.
New Forests
51
:
965
984
.
Dore
MHI
(
2005
)
Climate change and changes in global precipitation patterns: What do we know?
Environment International
31
:
1167
1181
.
Dukes
JS,
Pontius
J,
Orwig
D,
Garnas
JR,
Rodgers
VL,
Brazee
N,
Cooke
B,
Theoharides
KA,
Stange
EE,
Harrington
R,
Ehrenfeld
J,
Gurevitch
J,
Lerdau
M,
Stinson
K,
Wick
R,
Ayres
M
(
2009
)
Response insect pests, pathogens, and invasive plant species to climate change in the forests of northeastern North America: What can we predict
?
Canadian Journal of Forest Research
39
:
231
248
.
Graves
DH,
Ringe
JM,
Pelkki
MH,
Sweigard
RJ,
Warner
RC
(
2000
)
High value tree reclamation research
. Pages
413
421
. In:
Proceedings, Sixth International Conference on Environmental Issues and Management of Waste in Energy and Mineral Production
,
30 May–2 June 2000
,
Calgary, Alberta, Canada
.
CRC Press
Guebert
MD,
Gardner
TW
(
2001
)
Macropore flow on a reclaimed surface mine: Infiltration and hillslope hydrology
.
Geomorphology
39
:
151
169
.
Habib
E,
Haile
AT,
Tian
Y,
Joyce
RJ
(
2012
)
Evaluation of the high-resolution CMORPH satellite rainfall product using dense rain gauge observations and radar-based estimates
.
Journal of Hydrometeorology
13
:
1784
1798
.
Haydon
SR,
Benyon
RG,
Lewis
R
(
1996
)
Variation in sapwood area and throughfall with forest age in mountain ash (Eucalyptus regnans F. Muell.)
.
Journal of Hydrology
187
:
351
366
.
Hayes
RA
(
1991
)
Soil survey of Breathitt County, Kentucky
.
US Department of Agriculture–Natural Resources Conservation Service
,
Washington, DC
. https://www.nrcs.usda.gov/Internet/FSE_MANUSCRIPTS/kentucky/breathittKY1998/breathittKY1998.pdf
Helvey
JD,
Patric
JH
(
1965
)
Canopy and litter interception of rainfall by hardwoods of eastern United States
.
Water Resources Research
1
:
193
206
.
Hinrichs
EN
(
1978
)
Geologic map of the Noble quadrangle, eastern Kentucky: Geologic Quadrangle 1476
.
Holl
KD,
Zipper
CE,
Burger
JA
(
2009
)
Recovery of native plant communities after mining
.
Publication 460-140
.
Virginia Cooperative Extension
,
Blacksburg
Hoomehr
S,
Schwartz
JS,
Yoder
D,
Drum
EC,
Wright
W
(
2015
)
Erodibility of low-compaction steep-sloped reclaimed surface mine lands in the southern Appalachian Region, USA
.
Hydrological Processes
29
:
321
338
.
Husic
A,
Fox
J,
Adams
E,
Backus
J,
Pollock
E,
Ford
W,
Agouridis
C
(
2019
)
Inland impacts of atmospheric river and tropical cyclone extremes on nitrate transport and stable isotope measurements
.
Environmental Earth Science
78
:
36
.
Johnson
DW,
Hanson
PJ,
Todd
PJ
(
2002
)
The effects of throughfall manipulation on soil leaching in a deciduous forest
.
Journal of Environmental Quality
31
:
204
216
.
Jorgensen
DW,
Gardner
TW
(
1987
)
Infiltration capacity of disturbed soils: Temporal change and lithologic control
.
Water Resources Bulletin
23
:
116
1172
Kuczera
G
(
1987
)
Prediction of water yield reductions following a bushfire in ash-mixed species eucalypt forest
.
Journal of Hydrology
94
:
215
236
.
Lafleur
B,
Paré
D,
Munson
AD,
Bergeron
Y
(
2010
)
Response of northeastern North American forests to climate change: Will soil conditions constrain tree species migration
?
Environmental Reviews
18
:
279
289
.
Levia
DF,
Frost
EE
(
2006
)
Variability of throughfall volume and solute inputs in wooded ecosystems
.
Progress in Physical Geography
30
:
605
632
.
Lexington-Fayette Urban County Government [LFUCG]
(
2009
)
Stormwater manual
.
Lexington-Fayette Urban County Government
,
Lexington
. https://www.lexingtonky.gov/sites/default/files/2020-10/Stormwater%20Manual%2010-1-2020.pdf
Lindberg
T,
Bernhardt
ES,
Bier
R,
Helton
AM,
Merola
RB,
Vengosh
A,
Di Giulio
RT
(
2011
)
Cumulative impacts of mountaintop mining on an Appalachian watershed
.
Proceedings of the National Academy of Sciences of the United States of America
108
:
20929
20934
.
Llorens
P,
Poch
R,
Latron
L,
Gallart
F
(
1997
)
Rainfall interception by a Pinus sylvestris forest patch overgrown in a Mediterranean mountainous abandoned area. I. Monitoring design and results down to the event scale
.
Journal of Hydrology
199
:
331
345
.
Lloyd
CR,
Marques
ADO
(
1988
)
Spatial variability of throughfall and stemflow measurements in Amazonian rainforest
.
Agricultural and Forest Meteorology
42
:
63
73
.
Mao
D,
Cherkauer
KA
(
2009
)
Impacts of land-use change on hydrologic responses in the Great Lakes Region
.
Journal of Hydrology
374
:
71
82
.
McCarthy
EJ,
Skaggs
RW,
Famum
P
(
1991
)
Experimental determination of the hydrologic components of a drained forest watershed
.
Transactions of the ASAE
34
:
2031
2039
.
Miller
J,
Barton
C,
Agouridis
C,
Fogle
A,
Dowdy
T,
Angel
P
(
2012
)
Evaluating soil genesis and reforestation success on a surface coal mine in Appalachia
.
Soil Science Society of America Journal
76
:
950
960
.
Miralles
DG,
Gash
JH,
Holmes
TRH,
De Jeu
RAM,
Dolman
AJ
(
2010
)
Global canopy interception from satellite observations
.
Journal of Geophysical Research
115
:
D16122
. https://doi.org/10.1029/2009JD013530
National Public Radio
(
2022
)
Kentucky’s flooding victims face years of rebuilding efforts
.
National Public Radio
. https://www.npr.org/2022/09/09/1121745589/kentuckys-flooding-victims-face-years-of-rebuilding-efforts
NBC News
(
2022
)
Abandoned mines and poor oversight worsened Kentucky flooding, attorneys say
. https://www.nbcnews.com/news/us-news/lack-regulation-abandoned-mines-worsened-kentucky-flooding-attorneys-s-rcna41716
Negley
TL,
Eshelman
KN
(
2006
)
Comparison of stormflow responses of surface-mined and forested watersheds in the Appalachian Mountains, USA
.
Hydrological Processes
20
:
3467
3483
.
Orr
SP,
Rudgers
JS,
Clay
K
(
2005
)
Invasive plants can inhibit native tree seedlings: Testing potential allelopathic mechanisms
.
Plant Ecology
181
:
153
165
.
Osborne
LL,
Kovacic
DL
(
1993
)
Riparian vegetated buffer strips in water-quality restoration and stream management
.
Freshwater Biology
29
:
243
258
.
Overstreet
JC
(
1984
)
Robinson Forest Inventory. 1980–1982
.
Department of Forestry, College of Agriculture, University of Kentucky
,
Lexington
Pericak
AA,
Thomas
CJ,
Kroodsma
DA,
Wasson
WF,
Ross
MRV,
Clinton
NE,
Campagna
DJ,
Franklin
Y,
Bernhardt
ES,
Amos
JF
(
2018
)
Mapping the yearly extent of surface coal mining in central Appalachia using Landsat and Google Earth Engine
.
PLOS One
13
:
e0197758
.
Pond
GJ
(
2010
)
Patterns of Ephemeroptera taxa loss in Appalachian headwater streams (Kentucky, USA)
.
Hydrobiologia
641
:
185
201
.
Pond
GJ,
Passmore
ME,
Borsuk
FA,
Reynolds
L,
Rose
CJ
(
2008
)
Downstream effects of mountaintop coal mining: Comparing biological conditions using family- and genus-level macroinvertebrate bioassessment tools
.
Journal of the North American Benthological Society
27
:
717
737
.
Roberts
JA,
Daniels
WL,
Bell
JC,
Martens
DC
(
1998
)
Tall fescue production and nutrient status on southwest Virginia mine soils
.
Journal of Environmental Quality
17
:
55
62
.
Robson
AJ,
Neal
C,
Ryland
GP,
Harrow
M
(
1994
)
Spatial variations in throughfall chemistry at small plot scale
.
Journal of Hydrology
158
:
107
122
.
SAS Institute, Inc
. (
2016
)
SAS/CONNECT® 9.4 user’s guide
(4th ed).
SAS Institute Inc
.
Saunders
DA,
Hobbs
RJ,
Margules
CR
(
1991
)
Biological consequences of ecosystem fragmentation: A review
.
Conservation Biology
5
:
18
32
.
Sayler
KL
(
2016
) Central Appalachian Ecoregion. In:
Sayler
KL,
Acevedo
W,
Taylor
JL
(eds),
Status and trends of land change in the eastern United States–1973–2000
.
Professional Paper 1794-D. US Geological Survey
.
Sena
K,
Agouridis
C,
Miller
J,
Barton
C
(
2018
)
Spoil type influences soil genesis and forest development on an Appalachian surface coal mine ten years after placement
.
Forests
9
:
780
.
Sena
K,
Barton
C,
Angel
P,
Agouridis
C,
Warner
R
(
2014
)
Influence of spoil type on chemistry and hydrology of interflow on a surface coal mine in the eastern US coalfield
.
Water, Air and Soil Pollution
225
:
2171
.
Sena
K,
Barton
C,
Hall
S,
Angel
P,
Agouridis
C,
Warner
R
(
2015
)
Influence of spoil type on afforestation success and natural vegetative recolonization on a surface coal mine in eastern Kentucky, USA
.
Restoration Ecology
23
:
131
138
.
Sena
KL,
Barton
CD,
Williamson
TN
(
2020a
) Water quality of precipitation and streamflow, with air temperature data, in four Kentucky, Appalachian watersheds–1971 to 2018.
ScienceBase. US Geological Survey data release
.
Sena
K.,
Metzmeier
J,
Smith
B,
Hansen
B,
Barton
C
(
2020b
)
Climate change and invasive species: Challenges and opportunities for forest establishment on Appalachian surface mines
.
Journal of Sustainable Forestry
41
:
573
587
.
Sena
KL,
Williamson
TN,
Barton
CD
(
2021
)
The Robinson Forest environmental monitoring network: Long-term evaluation of streamflow and precipitation quantity and stream-water and bulk deposition chemistry in eastern Kentucky watersheds
.
Hydrological Processes
35
:
e14133
.
Simmons
J,
Currie
W,
Eshelman
K,
Kuers
K,
Monteleone
S,
Negley
T,
Phlad
B,
Thomas
C
(
2008
)
Forest to reclaimed mine land use change leads to altered ecosystem structure and function
.
Ecological Applications
18
:
104
118
.
Singh
V,
Rana
RK,
Singhal
R
(
2013
)
Analysis of repeated measurement data in the clinical trials
.
Journal of the Ayrurveda and Integrative Medicine
4
:
77
81
.
Skousen
J,
Gorman
J,
Pena-Yewtukhiw
E,
King
J,
Stewart
J,
Emerson
P,
DeLong
C
(
2009
)
Hardwood tree survival in heavy ground cover on reclaimed land in West Virginia: Mowing and ripping effects
.
Journal of Environmental Quality
38
:
1400
1409
.
Smalley
GW
(
1986
) Classification and evaluation of forest sites in the northern Cumberland Plateau. General Technical Report GTP-SO-50.
Southern Forest Experiment Station, US Department of Agriculture–Forest Service
,
New Orleans, LA
.
Tanaka
N,
Levia
D,
Igarashi
Y,
Nanko
N,
Yoshifuji
N,
Tanaka
K,
Tantasirin
C,
Suzuki
M,
Kumagai
T
(
2015
)
Throughfall under a teak plantation in Thailand: A multifactorial analysis on the effects of canopy phenology and meteorological conditions
.
International Journal of Biometeorology
59
:
1145
1156
.
Taylor
T,
Agouridis
CT,
Warner
RC,
Barton
CD,
Angel
PN
(
2009a
)
Hydrologic characteristics of Appalachian loose-dumped spoil in the Cumberland Plateau of eastern Kentucky
.
Hydrologic Processes
23
:
3372
3381
.
Taylor
T,
Agouridis
CT,
Warner
RC,
Barton
CD,
Angel
PN
(
2009b
)
Runoff curve numbers for loose-dumped spoil in the Cumberland Plateau of eastern Kentucky
.
International Journal of Mining, Reclamation, and Environment
23
:
103
120
.
Thimonier
A
(
1998
)
Measurement of atmospheric deposition under forest canopies: Some recommendations for equipment and sampling design
.
Environmental Monitoring and Assessment
52
:
353
387
.
Townsend
PA,
Helmers
DP,
Kingdon
CC,
McNeil
BE,
de Beurs
KM,
Eshelman
KN
(
2008
)
Changes in the extent of surface mining and reclamation in the Central Appalachians detected using a 1976–2006 Lands Landsat Time Series
.
Remote Sensing of Environment
113
:
62
72
.
US Department of Commerce [USDC]
(
2002
) Climatography of the United States: Monthly station normals of temperature, precipitation, and heating and cooling degree days 1971–2000,
Weather Bureau No. 81(15)
.
US Department of Commerce
,
Asheville, NC
US Environmental Protection Agency [USEPA]
(
2011
) The effects of mountaintop mines and valley fills on aquatic ecosystems on the Central Appalachian Coalfields
(EPA/600/R09/138F)
.
Office of Research and Development, National Center for Environmental Assessment
Wickham
J,
Wood
PB,
Nicholson
MC,
Jenkins
W,
Druckenbrod
D,
Suter
GW,
Strager
MP,
Mazzarella
C,
Galloway
W,
Amos
J
(
2013
)
The overlooked terrestrial impacts of mountaintop mining
.
BioScience
63
:
335
348
.
Williamson
TN,
Barton
CD
(
2020
)
Hydrologic modeling to examine the influence of the forestry reclamation approach and climate change on mineland hydrology
.
Science of the Total Environment
743
:
140605
.
Woods
AJ,
Omernik
JM,
Martin
WH,
Pond
GJ,
Andrews
WM,
Call
SM
(
2002
)
Ecoregions of Kentucky (color poster with map, descriptive text, summary tables, and photographs)
.
US Geological Survey
,
Reston, VA
. https://www.epa.gov/eco-research/ecoregion-download-files-state-region-4#pane-15
Woodwall
CW,
Oswalt
CM,
Westfall
JA,
Perry
CH,
Nelson
MD,
Finley
AO
(
2009
)
An indicator of tree migration in forests of the eastern United States
.
Forest Ecology and Management
257
:
434
1444
.
Zipper
CE,
Burger
JA,
Skousen
JG,
Angel
PN,
Barton
CD,
Davis
V,
Franklin
JA
(
2011
)
Restoring forests and associated ecosystem services on Appalachian coal surface mines
.
Environmental Management
47
:
751
765
.