Abstract

The twentieth century confluence of clear-cutting, deer overabundance, and rising nitrogen deposition favored dominance by the shade-intolerant, unpalatable, and nitrogen-demanding black cherry (Prunus serotina) throughout the Allegheny Plateau of the eastern United States. The abundance of this species conferred unique and valuable ecological and economic benefits that shaped regional biodiversity and societies. Sustaining these values is increasingly difficult because black cherry, seemingly inexplicably, has experienced diminished establishment, growth, and survival in the twenty-first century. In the present article, we chronicle the change and assess underlying drivers through a literature review and new analyses. We found negative plant–soil microbial feedback loops and lowered nitrogen deposition are biologically, temporally, and geographically consistent with observed declines. The evidence suggests that black cherry dynamics are the unintended consequence of actions and policies ostensibly unconnected to forests. We suggest that these shifts are a bellwether of impending changes to forests, economies, and ownership patterns regionally and beyond.

The structure and composition of the eastern deciduous forests of North America shifted markedly from the postcolonial to the modern era (Thompson et al. 2013, Nowacki and Abrams 2014). Although the drivers of these changes are often intentional and their impacts generally known (e.g., intensive harvests promote early successional species dominance), additional factors that are often unrelated to land management activities can exert unintended and often adverse outcomes on plant communities (for a review, see Tilman and Lehman 2001). For example, ornamental horticulture and international trade inadvertently introduced exotic pests and pathogens that caused collapses in American chestnut (Castanea dentata), elm (Ulmus spp.), and ash (Fraxinus spp.) populations (Gibbs 1978, Anagnostakis 1987, Herms and McCullough 2014). Aside from these notorious cases, there are myriad examples in which anthropogenic actions exert unforeseen effects on plant populations such as acid deposition predisposing sugar maple (Acer saccharum) and red spruce (Picea rubens) to mortality or reductions in fire frequency constraining oak (Quercus spp.) recruitment (Brose et al. 2001, Long et al. 2009, Adams et al. 2012). Collectively, these changes diminish diversity, threaten the provisioning of ecosystem services and, potentially, alter ecosystem functioning.

Throughout the northeastern United States, extensive exploitative clear-cutting radically altered forest composition and structure (Greeley 1925). In the Allegheny Plateau region of the eastern United States (figures 1 and 2), old-growth hemlock–beech–maple forests were cut around the turn of the twentieth century, creating the present-day second-growth stands dominated by early to mid-successional species (Hough and Forbes 1943, Whitney 1990). Turn of the twentieth century harvesting occurred in phases with white pine removed first (ship masts) where it was present, followed by hemlock removal (tanning industry) and ultimately a complete overstory removal (chemical wood cut), with each entry increasing the black cherry proportion (Marquis 1992). Consequently, black cherry (Prunus serotina) was among the principal beneficiaries, increasing in relative abundance by an order of magnitude from precolonial levels across the Northeast (Thompson et al. 2013, Nowacki and Abrams 2014) and becoming one of the dominant species across the 6.5 million hectares (ha), representing the core of its distribution on the Allegheny Plateau of the eastern United States (figures 1 and 2; Prasad et al. 2014). The regional dominance of black cherry was so pronounced that the Allegheny hardwoods forest type was defined to differentiate it from the broader northern hardwood forest type. Classification as an Allegheny hardwood type requires black cherry, tulip poplar (Liriodendron tulipifera), and white ash (Fraxinus americana) to jointly represent at least 50% of the total basal area (Eyre 1980); however, in the core of the Allegheny Plateau, inventory data find black cherry alone often represents 31% of the basal area overall, and 52% of the basal area in stands of the Allegheny hardwood forest type (Allegheny National Forest 2019).

Figure 1.

Map of black cherry (Prunus serotina) distribution and abundance |$(}}_}}} = }\frac} \times }}_}}}}}}}_}}}}} + \frac} \times }}_}}}}}}}_}}}}})$| throughout the eastern United States based on current Forest Inventory and Analysis (FIA) data (www.fs.fed.us/nrs/atlas/tree/762; see also Prasad et al. 2014). Little's range (1971) is outlined in red. The Pan-Allegheny region is outlined in black. For a delineation of the five ecoregions contained within the Pan-Allegheny region, see figure 2.

Figure 1.

Map of black cherry (Prunus serotina) distribution and abundance |$(}}_}}} = }\frac} \times }}_}}}}}}}_}}}}} + \frac} \times }}_}}}}}}}_}}}}})$| throughout the eastern United States based on current Forest Inventory and Analysis (FIA) data (www.fs.fed.us/nrs/atlas/tree/762; see also Prasad et al. 2014). Little's range (1971) is outlined in red. The Pan-Allegheny region is outlined in black. For a delineation of the five ecoregions contained within the Pan-Allegheny region, see figure 2.

Figure 2.

Map depicting the five ecological sections within the broader Allegheny Plateau (i.e., the Pan-Allegheny region) region that form the commercial core of black cherry (Prunus serotina) and the Allegheny hardwood forest type. The bar graphs depict the percentage change in average wet nitrogen deposition (ammonium [NH4+] and nitrate [NH3–]) annual rates in the 28-year period preceding (1972–1990) and following (1990–2018) the Clean Air Act Amendments. Source: The deposition data were retrieved from the National Atmospheric Deposition Data/National Trends Network (http://nadp.slh.wisc.edu/NADP). See supplemental material S3 for specific site locations.

Figure 2.

Map depicting the five ecological sections within the broader Allegheny Plateau (i.e., the Pan-Allegheny region) region that form the commercial core of black cherry (Prunus serotina) and the Allegheny hardwood forest type. The bar graphs depict the percentage change in average wet nitrogen deposition (ammonium [NH4+] and nitrate [NH3–]) annual rates in the 28-year period preceding (1972–1990) and following (1990–2018) the Clean Air Act Amendments. Source: The deposition data were retrieved from the National Atmospheric Deposition Data/National Trends Network (http://nadp.slh.wisc.edu/NADP). See supplemental material S3 for specific site locations.

The regional success of this shade-intolerant species in the twentieth century is only partly explained by the high-light environment clear-cutting created at the dawn of the century. Competitive dynamics during succession were further inadvertently shaped by two additional anthropogenic drivers in ways that heavily favored black cherry. Namely, game management policies that promoted high white-tailed deer (Odocoileus virginianus) densities benefitted black cherry, a low preference browse, over more browse-sensitive species, and anthropogenic increases in nitrogen availability enhanced black cherry survival and more than doubled its relative growth rates (Thomas et al. 2010, Canham and Murphy 2017).

Although Allegheny hardwood forests are the artifact of exploitative clear-cuts, browsing, and nitrogen deposition, the vast and nearly contiguous block of forest cover provides a range of unique, diverse, and valuable ecological and economic services. Forests in the region have low incidence and prevalence of exotic invaders (Iannone et al. 2015) and provide important breeding habitat for forest interior birds (Robertson and Rosenberg 2003). Black cherry produces the most predictable and energy-dense soft mast available to birds and mammals in northeastern forests and its foliage hosts an abundant, distinct, and diverse lepidopteran larvae community (Ryan et al. 2004, Tallamy and Shropshire 2009, Rose et al. 2014). Until recently, black cherry was the highest value timber species regionally, with average prices as much as 78% greater than the second most valuable species (Ray 2018), making these forests among the most profitable forests, on a per-board-foot basis, nationally (USDA Forest Service 2019b). This value forms the backbone of a multimillion dollar forest products industry that, in just a portion of the overall region (26 counties in northwest Pennsylvania and southwest New York: 62,094 square kilometers [km2]), contributes an estimated $34 ­million to local economies (USDA Forest Service 2016).

Unlike the twentieth century when these ecological services developed reliably, sustaining the values and services of Allegheny hardwood forests is increasingly challenged in the twenty-first century because black cherry, the archetypal species of this forest type, is currently in decline. On the Allegheny Plateau, Forest Inventory and Analysis (FIA) data indicates black cherry mortality rates (the percentage of trees per ha per year; diameter at breast height [dbh] ≥ 12.7 centimeters [cm]) that were, on average, initially lower than that of all other species combined, increased throughout the Pan-Allegheny region by 35% over the past 15 years, with a 73% increase in the northern unglaciated Allegheny Plateau ecoregion specifically (figure 3). In contrast, the mortality rates of all other species remained relatively consistent over that time period (figure 3). Similarly, black cherry tree growth, as measured by basal area increment (in square centimeters per year per tree), has fallen by an average of 29% across all diameter classes, whereas that of all other species has declined only modestly (approximately 6%; supplemental material S1). Therefore, from a relative perspective, black cherry is waning compared with other species.

Figure 3.

Annual mortality rates (percentage of trees per hectare per year; dbh ≥ 12.7 cm) for (a) black cherry (Prunus serotina; top panels) and (b) all other species (bottom panels) from 2004 to 2019 as calculated from USDA Forest Inventory and Analyses plots (USDA Forest Service 2019a, Stanke et al. 2020). The points represent annualized estimates from FIA inventories, and the solid line represents a nonparametric smoothed trend line. The grey line represents the average response (i.e., Pan-Allegheny) across all five ecological sections within the Allegheny Plateau. The colored lines for the individual sections are identical to those in figure 2. For a complete list of all included species, see supplemental material S5.

Figure 3.

Annual mortality rates (percentage of trees per hectare per year; dbh ≥ 12.7 cm) for (a) black cherry (Prunus serotina; top panels) and (b) all other species (bottom panels) from 2004 to 2019 as calculated from USDA Forest Inventory and Analyses plots (USDA Forest Service 2019a, Stanke et al. 2020). The points represent annualized estimates from FIA inventories, and the solid line represents a nonparametric smoothed trend line. The grey line represents the average response (i.e., Pan-Allegheny) across all five ecological sections within the Allegheny Plateau. The colored lines for the individual sections are identical to those in figure 2. For a complete list of all included species, see supplemental material S5.

Black cherry crown health measured in 102 plots of the US Forest Service Forest Health Monitoring Program distributed across the Allegheny National Forest indicate that crown density (the amount of light blocked by leaves and branches) and the live crown ratio have both decreased substantially in recent (2014–2016) surveys compared with earlier (1998–2006) surveys (Long et al. 2017). Historic versus current seed production surveys reveal mast crops have become more erratic and infrequent (Long and Ristau 2020). Reduced seed production has resulted in a concomitant decline in seedling densities. Indeed, FIA records reveal a 60% decline in established seedling densities (height ≥ 30 cm, dbh less than 2.5 cm) across the entire Allegheny Plateau region and near 76% declines in the northern unglaciated Allegheny Plateau since 2000 (figure 4; see also USDA Forest Service 2019a). The declining seedling establishment numbers have caused black cherry to drop in relative abundance from approximately 25% in 2004 to a mere 4% in 2019.

Figure 4.

Seedling densities (height ≥ 30 cm, dbh < 2.5 cm) for (a) black cherry (Prunus serotina; top panels) and (b) all other species (bottom panels) from 2004 to 2019 as calculated from USDA Forest Inventory and Analyses plots (USDA Forest Service 2019a, Stanke et al. 2020). The points represent annualized estimates from FIA inventories, and the solid line represents a nonparametric smoothed trend line. The grey line represents the average response (i.e., Pan-Allegheny) across all five ecological sections within the Allegheny Plateau. The color lines for the individual sections are identical to those in figure 2. Note the scale differences between the top and bottom panels. For a complete list of all included species, see supplemental material S5.

Figure 4.

Seedling densities (height ≥ 30 cm, dbh < 2.5 cm) for (a) black cherry (Prunus serotina; top panels) and (b) all other species (bottom panels) from 2004 to 2019 as calculated from USDA Forest Inventory and Analyses plots (USDA Forest Service 2019a, Stanke et al. 2020). The points represent annualized estimates from FIA inventories, and the solid line represents a nonparametric smoothed trend line. The grey line represents the average response (i.e., Pan-Allegheny) across all five ecological sections within the Allegheny Plateau. The color lines for the individual sections are identical to those in figure 2. Note the scale differences between the top and bottom panels. For a complete list of all included species, see supplemental material S5.

The observed declines in black cherry regeneration cannot simply be attributed to successional dynamics that predict increasingly shade tolerant species dominance over time. Although succession may play a role, three lines of evidence suggest something more pernicious underlies the contemporary declines. First, widespread reports from the twentieth century describe black cherry as “ubiquitous” in the regeneration layer (Collins and Pickett 1982) in undisturbed forests, including the low-light conditions of old growth stands (Maguire and Forman 1983, Collins 1990, Smith and Vankat 1991, Horsley et al. 2003). Second, this recently observed pattern of a sharp decline in black cherry densities in a region overwhelmingly dominated by mid-successional forests (less than 100 years; Pan et al. 2011) is inconsistent with the generally steady and gradual species replacement patterns that occur during succession (e.g., Xi et al. 2019). Third and most remarkably, even when natural or anthropogenic stand replacing disturbances reinitiate succession, the once copious regeneration of this iconic species (e.g., Husch 1954, Fredericksen et al. 1998, Vickers and Fox 2015) is no longer observed (e.g., Krueger et al. 2009, Royo et al. 2016). In fact, in a region in which even-aged silviculture is widely practiced by public and private industrial land managers as a tool to sustain the high-value timber and additional ecosystem services provided by early successional forests, surveys of regenerating stands reveal the magnitude of the change. For example, research from the 1960s and 1980s demonstrated that black cherry was abundant and successfully recruited into the sapling (more than 1.52m) size classes forming a major (more than 31%) fraction of the regenerating tree community (supplemental material S2). In contrast, contemporary studies show that few stems recruit into the larger size classes and form only a miniscule portion (approximately 3%) of the regenerating tree community (supplemental material S2). Overall, the declines in black cherry health and regeneration are so pronounced that regional stakeholders have identified the challenges as a priority threat to the sustainability of the Allegheny hardwood forests in the region, coequal with losses in other tree species resulting from pests and pathogens (Stout et al. 2019).

Initially, managers across the region believed these changes were temporary anomalies. Only over the past ∼15 years have all stakeholders fully grasped their scope and extent. Therefore, the science to determine the underlying causal factors is ongoing and definitive answers remain elusive. Below, we outline a set of hypothesized causal mechanisms for the observed black cherry health and regeneration declines and use a combination of synthesis of existing work and new analyses to examine the support for these hypotheses. We focus on five factors known or assumed to influence forest health broadly and black cherry dynamics specifically: climate variability, deer browsing, senescence, negative plant–soil microbe feedback loops (i.e., pathogens), and reduced nitrogen availability. Although other punctuated and often spatially heterogeneous components of declines can and do exist (e.g., pest outbreaks, drought), our focus is on factors that are well established, chronic, and geographically widespread.

Climate

Plant recruitment, growth, and survival rates are generally temperature and precipitation dependent with species distributions matched to and constrained by their climatic envelope or niche (Tilman and Lehman 2001). For many plant taxa, the current pace of climate change may increasingly decouple the optimal environmental conditions from the species’ current distribution potentially decreasing tree survival and recruitment and causing notable shifts in species ranges (e.g., Ibáñez et al. 2007, Loarie et al. 2009). Given that tree species ranges are defined by the presence and abundance of mature, long-lived individuals, range shifts may exhibit considerable temporal lags. However, demographic parameters such as tree mortality and seedling recruitment and survival may serve as early barometers of climate-driven changes (Vanderwel et al. 2013). Therefore, one might posit that marked deviations from the long-term (i.e., twentieth century) average climatic conditions and a shifting climate envelope may detrimentally affect black cherry overstory health and seedling recruitment.

Multiple lines of evidence, however, suggest climatic departures from long-term trends are not the primary factor underlying recent changes in black cherry demographic parameters. First, declines in black cherry health and reproduction are strongest within the northern unglaciated Allegheny Plateau, near the geographic core of the species’ range, rather than at its range boundaries, where climate-driven impacts may be most detectable (e.g., Hanberry and Hansen 2015). Moreover, the geographic core of the species is only predicted to shift modestly northeastward by about 35–53 kilometers under various climate scenarios (Peters et al. 2019). Notably, vegetation monitoring data from the last 10–28 years do not reveal substantial shifts in black cherry seedling or sapling distributions (Zhu et al. 2012, Woodall et al. 2018). If anything, the data suggest a minor range expansion, particularly at the southern edge (Hanberry and Hansen 2015).

We tested for any such hypothetical climate-driven changes by compiling data from 93 weather stations within the Pan-Allegheny region and a focused look within the 13 stations located within the northern unglaciated Allegheny Plateau (supplemental material S3) with continuous records since 1950 (www.ncdc.noaa.gov). Those data were used to test for deterministic changes in July temperatures and growing season (May–September) precipitation over the 70-year period from 1950 to 2019. We focused on these two variables as niche-based habitat suitability models find both variables are the predominant climatic factors shaping black cherry habitat suitability (Prasad et al. 2014). In addition, we tested for increases in minimum temperatures during the spring months of March through May, to assess whether warmer, earlier growing seasons may have occurred that might facilitate expansion of pathogens.

We performed a regression analysis on the average July temperature and the year from 1950 to 2019 for the 93 sites in the Pan-Allegheny region but found no evidence for any change in average temperatures over the past 70 years (r2 = .005, F = 0.58, p = .31). However, restricting the same analysis to the 13 stations within the core northern unglaciated Allegheny Plateau range revealed a significant increase in July minimum temperatures over the past 70 years (r2 = .19, F = 16.26, p < .001) of 0.255 degrees Celsius (°C; 0.46°Fahrenheit [F]) per decade, on average (figure 5a). In addition, the average minimum temperature from March through May for the core 13 sites increased significantly over time (r2 = .17, F = 13.9, p < .001) at an average rate of 0.16°C (0.29°F) per decade (figure 5b). The average cumulative precipitation received from May through September increased significantly since 1950 for the entire Pan-Allegheny region (r2 = .11, F = 8.1, p = .006) at an average rate of 1.35 cm (0.53 inches) per decade and in the core northern unglaciated Allegheny Plateau (r2 = .12, F = 9.4, p = .003) the increase was at a slightly higher rate of 1.81 cm (0.71 inches) per decade (figure 5c). Therefore, the climate throughout the Pan-Allegheny region has become progressively warmer and wetter in the past 70 years, especially so in the northern unglaciated Allegheny Plateau. However, given the broad geographical and climatic range occupied by black cherry, it seems unlikely that these slight changes in climate have been the primary driver of observed declines; nevertheless, changing climatic conditions may interact with other factors (e.g., fungal pathogens; see below).

Figure 5.

Seventy year change (1950–2019) in climatic variables on the northern unglaciated Allegheny Plateau show (a) increased average July temperature (r2 = .19, F1,68 = 16.26, p < .001), (b) increased March–May average minimum (r2 = .17, F1,68 = 13.9, p < .001), and (c) greater average cumulative precipitation from May to September (r2 = .12, F1,67 = 9.4, p = .003).

Figure 5.

Seventy year change (1950–2019) in climatic variables on the northern unglaciated Allegheny Plateau show (a) increased average July temperature (r2 = .19, F1,68 = 16.26, p < .001), (b) increased March–May average minimum (r2 = .17, F1,68 = 13.9, p < .001), and (c) greater average cumulative precipitation from May to September (r2 = .12, F1,67 = 9.4, p = .003).

White-tailed deer browsing

Overbrowsing is often cited as a major contributing factor to tree regeneration failures in forests worldwide (Rooney and Waller 2003, Côté et al. 2004). Given that chronically high deer populations persisted throughout the Allegheny Plateau in the twentieth century (Royo and Stout 2019), it is possible variation in browse pressure may account for observed changes in black cherry regeneration dynamics. Specifically, higher browse pressure may directly limit black cherry establishment and growth through tissue consumption or, alternatively, lower browse pressure may indirectly limit cherry establishment and growth by intensifying interspecific competition (Horsley et al. 2003).

Support for a direct adverse browsing effect on recruitment is weakened by the fact black cherry is often considered a low to moderately preferred browse species, relative to most other co-occurring tree species (Healy 1971, Latham et al. 2005). Low foliage, stems and fruit are unpalatable because of the presence of cyanogenic glycosides prunasin and amygdalin, which are toxic (Bischoff and Smith 2011). These palatability differences relative to co-occurring species clarify how black cherry densities can be reduced when browse pressure is high but still increase in relative abundance (e.g., Marquis 1974, Horsley et al. 2003; see also supplemental material S2). More importantly, the argument that browsing severely limits black cherry regeneration in the twenty-first century is unreasonable, given that the deer densities within the core of the Pan-Allegheny region, where seedling numbers have declined the most, have declined by as much as 45% over the past two decades relative to the late twentieth century (Royo and Stout 2019). This suggests that the diminished browse pressure and concomitant increase of browse-sensitive species (Royo et al. 2010) may intensify competition, thereby limiting black cherry establishment and survival. Indeed, a controlled browsing experiment conducted in the 1980s found lower deer densities (3.9–7.8 deer per km2) promoted recruitment of co-occurring hardwood species thereby reducing the absolute and relative abundance of black cherry (Horsley et al. 2003). Nevertheless, even when low deer abundance favored denser and more speciose competitive neighborhoods, black cherry establishment and growth remained high and the species ultimately represented a third of the basal area, firmly establishing these forests as Allegheny hardwoods (see also supplemental material S2).

Contemporary studies at both low and moderately high deer densities (4.46–13.4 deer per km2) have shown a deeply altered dynamic. Cherry seedling banks have declined by as much as 76% within the northern unglaciated Allegheny Plateau (figure 4), and relative growth rates have dropped by as much as 50% (see Gottschalk 1985, Krueger et al. 2009), thereby diminishing black cherry relative abundance to a mere 3%–14.6% (e.g., Royo et al. 2016; supplemental material S2). Therefore, although deer browsing was a major determinant of tree recruitment success regionally over the past century (Royo and Stout 2019) and might still limit black cherry in portions of the species’ range in which white-tailed deer densities remain very high (McWilliams et al. 2018), we suspect other factors are more directly tied to the observed black cherry declines of the twenty-first century in the core of its range.

Senescence

Evolutionary theory predicts that senescence, a decline in fecundity and increased mortality risk with age, should be a near universal phenomenon (Kirkwood 1977). Although empirical examinations of tree fecundity schedules with age are rare (Harper and White 1974), demographic analyses indicate a majority of trees species (81%) exhibit no to negligible senescence (Silvertown et al. 2001, Baudisch et al. 2013). As was noted above, the vast proportion of trees in the Allegheny hardwood type originated from a short, intense period of exploitative harvesting late in the nineteenth century and early twentieth century. Therefore, trees making up the canopy of these stands now are largely mature. Black cherry is an early successional species that exhibits growth declines and increased mortality beginning at age 80 and an estimated mean lifespan of 100 years, with some individuals achieving upward of 250 years (Hough and Forbes 1943). Therefore, it is reasonable to hypothesize that senescence is responsible for the decreased survival and fecundity.

Analyses of tree mortality using repeatedly measured plots throughout the eastern United States find stand age is not a strong predictor of mortality for hardwood species, including black cherry (Dietze and Moorcroft 2011, Morin et al. 2015). These results are unsurprising because forests throughout the region are relatively young with a majority clustered in the 40–80 year age class and because of the inherent variability in ages of individual trees within stands, particularly in multiple-age stands (Pan et al. 2011). However, even on the Allegheny Plateau, where average stand ages for black cherry are greater, we find little support for the senescence hypothesis. A recent study comparing canopy health and seed production in 70 versus more than 110-year-old forest stands showed that tree health, as assessed by the percentage of standing dead, live crown ratio, dieback, and density metrics, was better in the older stands than in younger stands (Long and Ristau 2020). Moreover, comparisons of seed production in stands measured in the 1970s and remeasured in the 2010s revealed that although the periodicity in bumper crops years became more temporally and spatially idiosyncratic in the twenty-first century, older black cherry dominated stands (more than 110 years old) produced as much or more seed than younger (less than 80 year old) stands (Bjorkbom 1979, Long and Ristau 2020). This follows the results from Martin and Canham (2010) showing that seed production is limited in smaller size classes but nearly constant across size after trees reach 34 cm dbh. Relatedly, Royo and Ristau (2013) extensively surveyed seed banks in 39 stands spanning a 46–106 year stand age gradient. Despite black cherry seed's ability to remain viable for multiple years in the soil (i.e., seed bank), they found viable black cherry in only three of the 39 stands (65, 67, and 91 years old) and even there, only at low abundance (167 seeds per ha). To the degree seed banks correlate with seed production, these findings are in stark contrast to work in the 1970s by Bjorkbom (1979) who found the black cherry seed bank averaged 642 seeds per ha. Overall, the data indicate some factor other than age-related senescence is causing the deterioration.

Negative plant–soil microbial community feedback loops

An alternative explanation views tree species as active participants changing their environment through chemical, physical, and biological interactions. Mature forest trees shape their associated forest soil microbial communities and, in some cases, amass microbial antagonists that ultimately hinder plant recruitment (Mills and Bever 1998). These negative plant–soil microbial community feedback loops can lead to conspecific negative density or distance dependence (CNDD) and limit tree recruitment in tropical and temperate forests (Hyatt et al. 2003). Therefore, it is reasonable to hypothesize that the high relative abundance of black cherry in the Allegheny hardwood forests may cause high associated microbial antagonist (e.g., pathogenic fungi) densities that limit black cherry establishment, growth, and survival.

Robust support exists for this hypothesis. For example, infection risk on and mortality to black cherry seedlings by the foliar ascomycete pathogen Blumeriella jaapii (cherry leaf spot) positively correlates with seedling density in nurseries (Stanosz 1992). More notably, Packer and Clay (2000, 2003) first documented negative plant–soil microbe feedback loops with temperate forests in the black cherry–Pythium spp. (root rot) system. This work, along with subsequent corroboration, demonstrates strong negative plant–soil microbial feedback loops on black cherry elsewhere in its native range (e.g., Reinhart et al. 2005). Indeed, assessments of the pervasiveness and strength of CNDD in North American temperate trees find arbuscular mycorrhizal (AM) species, including black cherry, suffer disproportionately from CNDD effects with black cherry exhibiting the strongest or near strongest negative feedback loops of all tested species (55 species, Bennett et al. 2017; 151 species, Johnson et al. 2012). Multiple studies have shown that these plant–soil microbial interactions cause high black cherry seedling mortality close to mature conspecifics, thereby shifting seedling distributions away from mature trees (e.g., Packer and Clay 2003, Martin and Canham 2010).

Over time, these antagonistic microbial effects are hypothesized to intensify as host tree populations mature and their root systems expand, thereby extending the zone of potential pathogen influence (Packer and Clay 2004). An examination of the overstory tree community from 1990 to the present shows that the black cherry population structure of individuals at least 2.54 cm throughout the Pan-Allegheny region is shifting toward fewer but decidedly larger trees. Regionally, declines in tree density are primarily driven by losses in smaller diameter stems with a concomitant increase of 34% in the quadratic mean diameter and a 26% increase in basal area (supplemental material S4, figure 6). Among the five ecoregions, the northern unglaciated Allegheny Plateau has the highest basal area of black cherry (3.0 square meters [m2] per ha in 2019), but it experienced the steepest (76%) decline in seedling densities over time (figures 4 and 6). In contrast, black cherry basal area in the northern glaciated Allegheny Plateau ecoregion is the lowest of all five ecoregions (1.4 m2 per ha in 2019) and, in the present study, cherry seedling densities have remained consistently low over time (figure 4). Although further examination of the variation in pathogen loads and their effects across a host tree density gradient would help support or refute the negative plant–soil microbial feedback hypothesis, the accruing mature tree versus declining seedling population dynamics documented in the present article are consistent with Packer and Clay's (2004) prediction that antagonistic microbial effects may intensify as host trees grow larger.

Figure 6.

Basal area (in square meters per hectare) diameter distributions for (a) black cherry (Prunus serotina; top panels) and (b) all other species (bottom panels) in each of the five ecoregions contained within the Pan-Allegheny region in four time periods spanning 1990–2019 as calculated from USDA Forest Inventory and Analyses plots (USDA Forest Service 2019a, Stanke et al. 2020).