American Chestnut AND Effect of Tree-Influenced Soil at Fitzgerald Lake
Katherine Halstead Plant Ecology Jesse Bellemare December 2009 Herbivory on a Former Foundation Species: A Comparison of Herbivore Damage on American Chestnut and Red Oak Sprouts Introduction
The American chestnut (Castanea dentate) was formerly a foundation species in eastern North America. Chestnut and oak co-dominated forests were common across the eastern United States,
reaching to areas of southern New England (Ellison et al 2005). Castanea denate is thought to have played a defining role in influencing the ecosystem structure of these forests. Chestnut
wood decomposes more slowly than other hardwoods because of its high tannin content; in addition, the tree has high sprouting ability, fast growth and leaves of fairly high nitrogen
content (Ellison et al 2005). Dead chestnut wood provided habitat for a variety of wildlife, while nutritious chestnut leaves falling into forest streams were the main nutrient input
for aqueous ecosystems (Ellison et al 2005). For these reasons, the disappearance of chestnut probably changed ecosystem dynamics significantly (Ellison et al 2005). Since the introduction
of chestnut blight (Endothia parasitica) to eastern North America in 1900 (brought over on foreign wood), the species experienced a dramatic crash in abundance and within 50 years, only
small saplings remained (Ellison et al 2005; Pailet 1984). Chestnut blight is a fungus which kills saplings long before they reaches adult size by destroying the phloem, eventually cutting
off nutrient flow in the trunk (Pailet 1984). The blight eliminates sexual reproduction, but chestnut stumps and saplings continue to reproduce vegetatively by sprouting from the root
crowns (Pailet 1984). Attempts to create resistant hybrids by backcrossing American chestnut with resistant European species are ongoing, but it seems
1 Alison Montgomery BIO 364 – Plant Ecology J. Bellemare 11/1/09 The Effect of Tree-Influenced Soil on Understory Biomass Introduction A seed’s ability to germinate and survive is influenced
by a range of abiotic and biotic factors present in the environment. Abiotic factors include the presence and absence of nutrients, light, and water (Fenner and Thompson 2005). From
these alone, there are already factors that could prevent a seed’s success in a new environment. The biotic factors include neighboring plant competition for these resources, resulting
in behaviors such as shading, varying growth rates, and allelopathy (Grime 1977). These factors are all intertwined, as competitive behaviors displayed by any organism are a result of
the availability, or lack thereof, of abiotic resources. These competitive behaviors in turn influence the availability of abiotic factors in the environment. The result is continual
adaptation by the organisms in an environment, in order to better survive the stresses they face in their particular habitat and beyond (Keddy 2007). The abiotic factor of soil chemistry
and nutrient availability is one of the most important when it comes to plant survival (Finzi et al 1998). Certain trees (the biotic factor) affect the soil chemistry and therefore impact
other plants that may try to grow around the tree. Tree species examined here that seem to inhibit understory include the Eastern Hemlock (Tsuga canadensis) and Red Oak (Quercus rubra).
The leaf litter of these trees is not easily decomposed and is highly acidic, which may aid in the prevention of understory growth, and therefore competition (Finzi et al 1998). While
a seedling may be able to begin germination beneath these trees, the acidic, low-nutrient soil caused by
Montgomery 2 the leaf litter of these trees can prevent a seedling from surviving once the resources in the cotyledon are used up. It has been demonstrated that trees whose leaf litter
is not easily decomposed, such as the ones listed here, prevent other plants from obtaining necessary nitrogen and carbon by sequestering it for themselves. This is due to the organic
acids that build up from the slow decomposition of their leaf litter. Red Maple (Acer rubrum) is an intermediate species, as it causes less severe soil acidification than Hemlock or
Red Oak (Finzi et al 1998). In contrast, there may be instances of plants unintentionally aiding other species in making soil chemistry more favorable for growth. This is the potential
case of the Speckled Alder (Alnus incana), which is one of the few non-legume plant species to maintain a symbiotic relationship with nitrogen-fixing bacteria. This relationship provides
the bacteria with a source of nutrients that allow it to consume nitrogen, to the benefit of the plant host. The bacteria, Frankia spp. in the case of Alder, are able to take atmospheric
nitrogen, N2, and convert it into forms usable by plants such as ammonium (NH+4) (Binkley et al 1992). With this ability to harness atmospheric nitrogen, Speckled Alder is being examined
as a potential soil restorer of areas that have been overfarmed or polluted (Wheeler et al 2001). This relies on the idea that the alder does not retain all of the nitrogen it gains
from the bacteria, and ends up releasing some back into the environment, either via leaf litter or from the roots themselves. This study hopes to examine the relationship between soil
and seedlings, utilizing soil that comes from underneath of different tree species. The range of soils tested here, based on the evidence above, should ideally provide a gradient of
soil types. This gradient should lead to different rates of biomass accumulation in the herbaceous,
Montgomery 3 understory growth, represented in this experiment by Wisconsin Fast Plants (Brassica rapa). This species was selected due to its quick germination and growth for the timeline
of this experiment. Based on soil pH and nutrient studies by Finzi et al (1998), and the ability of alder to fix nitrogen, the gradient from best soil to worst soil should be as follows:
Alnus incana > Acer rubrum > Quercus rubra > Tsuga canadensis It is proposed that plants grown in Speckled Alder soil will have the most biomass at the end of the growing period, followed
by a notable decrease and the lowest levels of biomass in the plants grown in soil from underneath Eastern Hemlocks. Materials and Methods Soil samples were collected from two different
sites in western Massachusetts: Fitzgerald Lake Conservation Area in Northampton, MA, and along Nashill Road, Williamsburg, MA. Soil was collected from alders and hemlocks at both sites,
with additional soil from under red maple in Williamsburg and under red oak at Fitzgerald Lake. The soil was processed to remove any large stones, twigs, and roots before being potted
in small pots (~4 oz). For the six soil types, there were 15 pots each, resulting in 90 pots for the experiment. Two Wisconsin Fast Plant (Brassica rapa) seeds were planted in each pot,
according to protocol provided by Carolina Biological Supply Company. The pots were colored coded according to plant and site, randomly arranged on trays, numbered (1 through 90), watered
to saturation, and placed in a growth chamber on October 31st, 2009. To ensure there would be germination, the fast plants were kept on a 24-hour light cycle at 24oC. The trays were
rotated to minimize light bias, and watered, every other day
Montgomery 4 or as needed to maintain soil moisture. Once it was certain each pot had at least one seedling, pots with two seedlings had one removed to eliminate competition. The plants
were grown for 17 days. Any pots with no seed germination had the failed seedlings examined and were given another seed on Nov. 6th, 2009, and allowed to grow a few days longer. On the
17th day, Nov 16th, 2009, the plants were carefully snipped at soil level, cleaned to remove dirt and debris, and weighed. The data collection resulted in germination percentages by
soil type, presence of inflorescence, and weights for each soil type, and the data were analyzed in Excel. Results As this study aimed to show that the tree species altered the soil
regardless of the other soil chemistry factors, location ideally should not be a contributing factor on a regional scale. Therefore the locations where the soils were collected were
initially used only for comparing the tree-related soils. As no direct testing of nitrogen concentrations concentrations or soil pH occurred, however, it was difficult to separate out
the soil chemistry as influenced by the tree versus general soil conditions, as the general soil chemistry could be good enough or poor enough to possibly override a tree’s ability to
alter the soil. This denotes the interconnectedness of the biotic and abiotic factors in a community. These interconnected factors are seen most pronouncedly with the hemlock soil, where
the Williamsburg hemlock soil produced nothing, and the Fitzgerald Lake hemlock soil produced more biomass than either the oak or maple soil (Figure 1). The oak and maple data were combined,
in order to produce averages, and this worked well due to similar seedling results for both soils. This is noted in Figure 1, where both oak and maple soil showed no significant difference
in seedling germination (Oak: 78.57±SE 8.16, Maple:
Montgomery 5 83.33±SE 6.29). Alder, regardless of location, had the highest germination percentages and inflorescence amounts (Figure 1; Figure 2). To examine the overall health of the
seedlings, the germination percentage of was recorded, as 0%, 50%, or 100% as per the number of seedlings that appeared in each pot. Eastern Hemlock soil had significantly less germination
than the alder soil or oak and maple soil, and alder had the most, though not significantly more than the oak and maple (Figure 2). On Nov. 6th, 2009, the seeds from pots with 0% germination
were examined, and these seeds were observed to have begun germination, but poor soil conditions prevented many from reaching the surface. This occurred primarily with the hemlock, but
also with the maple and oak soils to a lesser degree. For the seedlings that did survive, the other measure of fitness observed was inflorescence. This can be used as an indicator of
whether a seedling or plant is simply surviving, or thriving and doing well enough to reach its reproduction stage. Seedlings in alder soil, on average, had the highest numbers of flowers
and buds (Will: 1.86±0.17, Fitz: 4.53±0.55), followed by those in hemlock soil (Fitz: 2.07±0.21, Will: none), and finally the oak and maple (Oak: 0.27± 0.49, Maple: None) (Figure 4).
It was interesting to observe that Fitzgerald Lake soil had inflorescence in all the soil types, where only the seedlings in Williamsburg alder soil produced inflorescence (Figure 3).
The final measurements taken were the aboveground biomass, which revealed that Speckled Alder soil had the most biomass at each location (Figure 5). Regardless of the other soil conditions,
alder soil was the best for the Wisconsin Fast Plants compared to the other tree-modified soils, even if the overall soil at Williamsburg was less conducive to growth than the soil at
Fitzgerald Lake. This is reinforced with the averaged
Montgomery 6 biomass by species (Figure 6). The alder-influenced soil had significantly greater biomass than either hemlock or the oak and maple. Though the Williamsburg alder biomass
was nowhere near the levels of the Fitzgerald Lake alder biomass, it was the highest for its site. Discussion Based on the data provided, it seems that the overall soil chemistry of
the Williamsburg site was less conducive to seedling growth than the soil from Fitzgerald Lake. This is most significantly displayed in the case of the hemlock soil from both sites.
Hemlock soil is known to be acidic, yet for one area to do so well over another suggests the overall soil at Fitzgerald Lake is of better quality than that at the Williamsburg site.
The combination of hemlock leaf litter with the already poor soil of Williamsburg led to a completely uninhabitable environment for the Fast Plants (Brassica rapa), where the presence
of alder trees in Williamsburg made the soil somewhat habitable for the plants. Lake Fitzgerald soil was much more favorable to the germination and growth of the Fast Plants. As the
results showed, all of the soil types from Fitzgerald Lake had germination, and the alder soil from Fitzgerald Lake was the best soil out of all the categories. With the growth of these
seedlings in soil collected from these two sites, an ecosystem hierarchy is revealed. There is the umbrella of the abiotic setting, with its own overall cycling of nutrients and resources,
modified with individual biotic factors having their own “pockets” of impact on the soil in which they are growing (Fenner and Thompson 2005). The aboveground biomass of the fast plants
strongly suggests the general soil chemistry plays a significant role in a seedling’s development. Yet as both field sites
Montgomery 7 currently have biomass present, the resulting experimental biomass is not the only measure of plant health. For future studies, it would be more precise to weigh the seedlings
individually, to obtain standard error and establish the exact significance of the biomass. The fitness of a plant is also closely related to soil chemistry, and the availability of
nutrients will ultimately impact a plant’s ability to not only germinate and grow to maturity, but to reproduce. Everything an organism does costs some amount of energy, and evolution
and adaptation lead to how an organism’s energy is allotted. Reproduction is one of these costly activities, and only if there are adequate resources will a plant be able to flower and
go to seed. Due to energy costs, it would not be advantageous to send seeds into an environment that the parent struggled to survive in in the first place, and so logically it is better
to allot resources elsewhere (Cousens et al 2008). The fact that seedlings in all Fitzgerald Lake soil types had inflorescence suggests that even if the leaf litter had a negative impact
on the soil, the soil was good enough that the seedlings had a high fitness level. Only the alder from Williamsburg produced flowers, suggesting that the soil chemistry was not favorable
for seedling survival, and the alder was capable of making the soil better for seedling growth and fitness. It seems that the soil from both sites is capable of supporting seed germination,
but the soil from the Williamsburg site may not allow plants to thrive and reproduce as well as those in Fitzgerald Lake soil. This strongly suggests that overall soil chemistry plays
a larger role than the individual trees, even if the areas are in close proximity. With the environment being laid out as a hierarchy, with the large abiotic scale and smaller biotic
factors within, it is clear that the organisms that grow in the soil alter the chemistry of it. It has been demonstrated that hemlock makes soil more acidic and
Montgomery 8 therefore less hospitable to seedlings, and underneath an oak or maple are not ideal locations for a seedling to settle. It was interesting that though oak and maple were
supposed to be better than hemlock, the results were not significant and did not support this hypothesis (Finzi et al 1998). A larger scale study of more sites, and better pairing of
these trees could change these results. Though the oak and maple had better germination than the hemlock, they had significantly less inflorescence (Figure 2; Figure 4). This may denote
the soil chemistry of the sites versus the impacts these trees were having on the soil, though as mentioned, further examination is necessary. Future studies, aside from including more
sites to determine tree-related soil modification versus the general soil chemistry, could also include chemical analysis of the soil. This could aid in determining the differences between
the sites, and tell us what in the soil is leading to better biomass and fitness at Fitzgerald Lake, or what is so inhibitive to the plants at Williamsburg. By increasing the site number
and doing more exact chemical analysis, the significance of the overall site chemistry in conjunction with a tree’s ability to modify soil could be determined to provide a more precise
view of abiotic and biotic interactions. Speckled Alder (Alnus incana), however, was significantly better at soil enrichment. The alder soil from both sites was much more favorable to
fast plant growth than any of the other tree-modified soils, even if the two sites had significantly different amounts of biomass and inflorescence. This suggests that its potential
use as a bioremediating plant is not unfounded. As the alder itself was able to establish at both sites, it may be capable of growing in a range of soil types. This is primarily due
to its symbiotic relationship with the bacteria Frankia spp., which becomes a nutrient gatherer
Montgomery 9 for the plant. Demonstrated here as being able to enrich the soil, and with further study on its ability to grow in a variety of soil types, Speckled Alder may be a candidate
for enriching overused or polluted land. Areas that have been used repeatedly for farming, as well as for different kinds of mining, are often depleted of nutrients required by plants
for survival (Wheeler et al 2001). If alder were to display significant proof of its ability to add nitrogen to the soil in future studies, the field of bioremediation would be able
to add another useful species to its repertoire. In summary, the connections between the interactions of the abiotic and the biotic factors in the environment are complex and somewhat
geographically localized. These results require further chemical analysis on a broader geographical scale to determine the relationship more exactly. It is clear, nonetheless, that Speckled
Alder (Alnus incana) creates more favorable soil conditions than the Eastern Hemlock (Tsuga canadensis), and underneath both Red Oak (Quercus rubra) and Red Maple (Acer rubrum) are not
ideal locations for seedling growth. This ability to improve soil may make the Speckled Alder a serious candidate in the bioremediation of overused soils.
Montgomery 10 References Binkley D, P Sollins, R Bell, D Sachs, D Myrold 1992. “Biogeochemistry of Adjacent Conifer and Alder-Conifer Stands.” Ecology 73(6): 2022-2033. Cousens R, C
Dytham, and R Law 2008. Chpt 8 “The Evolution of Dispersal”. In Dispersal in Plants: A Population Perspective. Oxford: Oxford University Press. Fenner M and K Thompson 2005. The Ecology
of Seeds. Cambridge: Cambridge University Press. Finzi AC, N Van Breemen, CD Canham 1998. “Canopy Tree-Soil Interactions Within Temperate Forests: Species Effects on Soil Carbon and
Nitrogen.” Ecological Applications 8(2): 440-446. Grime JP 1977. “Evidence for the existence of three primary strategies in plants and its relevance to ecological and evolutionary theory.”
The American Naturalist 111(982): 1169-1194. Keddy PA 2007. Chpt 5 “Competition.” In Plants and Vegetation: Origins, Processes, Consequences. Cambridge: Cambridge University Press. Wheeler
CT, LT Hughes, J Oldroyd, ID Pulford 2001. “Effects of nickel on Frankia and its symbiosis with Alnus glutinosa (L.) Gaertn.” Plant and Soil 231(1): 81-90.
Montgomery 11 Appendix Fig. 1: Average germination percentages for each tree species with location noted. Key: (F or W = Fitzgerald Lake or Williamsburg, respectively, A=Alder, M=Maple,
O=Oak, H=Hemlock). Fig. 2: Average germination percentage in soil type, regardless of location.
Montgomery 12 Fig. 3: Average inflorescence of seedlings grown in each soil type, by location. Inflorescence includes buds and flowers. See Figure 1 for key to labels. Fig. 4: Average
inflorescence of seedlings in each soil type, regardless of location. Inflorescence includes flowers and buds.
Montgomery 13 Fig. 5: Total weights of the aboveground biomass for each soil type, according to location. Key for labels in Figure 1. Fig. 6: Average aboveground biomass, regardless
of location, for each tree species.
improbable that the hybrid could reforest the vast area of chestnut that has been lost (Ellison et al 2005). The dramatic decline of C. denate and its subsequent prolonged absence from
the canopy probably changed the nature of the biological interactions between the tree and the species which formerly relied on it. In particular, insect herbivores which specialized
on chestnut might have been negatively affected. For this reason, chestnut might currently be experiencing an escape from its natural herbivores. This dynamic is similar to how exotic
species tend to lose their natural pests and pathogens during the invasion process, known as the enemy release hypothesis. Multiple studies on herbivory on exotic species have provided
evidence for enemy release (Carpenter and Cappucino 2005; Lewis et al 2006). In a geographic study Kristen Lewis and colleagues found that the invasive biennial Alliaria petiolana suffered
greater herbivory in its native European range than in its introduced North American region (Lewis et al 2006). In addition, a 2005 survey of 39 exotic and 30 native plant species in
Ottawa, Canada found that alien plants experienced less herbivore damage than natives (Carpenter and Cappucino 2005). To investigate if the enemy release hypothesis might be in play,
insect herbivory on chestnut could be compared with herbivory on a close relative. Red oak (Quercus rubra) and chestnut both belong to the family Fagaceae, so one could expect them to
experience similar levels of herbivore damage. Oak and chestnut co-dominated forests for centuries prior the onset of chestnut blight; though adult chestnut is absent in present times,
oak remains a common component of the southern New England forest canopy. This study quantified and compared leaf damage on C. canadensis and Q. rubra to test whether chestnut might
be experiencing an escape from its natural herbivores. Methods Leaves from chestnut and oak sprouts were collected on October 23, 2009 at Fitzgerald Lake Conservation Area in Northampton,
Massachusetts. Massachusetts. Ten leaves were randomly selected
from each sapling. To assess relative herbivory, damage was quantified as percentage of leaf area lost using a leaf scanner. Leaves from each sapling were sorted into two types: undamaged
and damaged by herbivores. Undamaged leaves were recorded as 0% damaged and discarded; herbivore-damaged leaves were brought to the lab for scanning analysis. Each leaf was scanned twice
(to account for scanning variability) prior to covering its holes with tape. Scanning was repeated with each patched leaf to determine the percentage of leaf tissue lost. The mean percent
leaf herbivory for each sprout was determined. The standard error between the means was calculated for the chestnut and oak sapling groups, respectively. To assess whether the difference
in leaf tissue loss between the chestnut and oak groups was significant, a two sample t-test assuming unequal variances was conducted, testing the null hypothesis that the mean difference
in herbivory equaled zero. Results In total, ten chestnut sprouts and ten oak sprouts were sampled, which gave a final sample size of 200 leaves from 20 saplings. Chestnut (Fig. 2) sprouts
experienced significantly less herbivore damage than red oak (Fig. 3) sprouts. The mean percent herbivory on chestnut saplings was 1.54% (± 0.43% standard error) (Fig. 1). Oak saplings
experienced 5.80% herbivore damage (± 1.60% standard error) (Fig. 1). A two sample t-test showed that the chestnut group experienced significantly less herbivore damage than the oak
group (P = 0.027). Discussion The findings of this survey demonstrate that the insect attackers of a plant species which experienced a near-extinction over the past century might have
become less common or absent. Specifically, this study found that American chestnut sprouts experience significantly less herbivore damage than saplings of their close relative, red
oak. This dynamic is in some ways similar to how exotic species escape their natural herbivores and pathogens by invading a new location. Native species co-evolve with specialized herbivores
and pathogens; thus, the near
disappearance of a plant, if it spans an extended period of time may trigger the decline or disappearance of those pests. Previous studies which tested the enemy release hypothesis found
that highly invasive exotic plants experience less herbivory than their less invasive counter parts (Carpenter and Cappuccino, 2005). Thus, the ability of some exotic species to escape
from pest herbivores is one factor that helps lead to the success of those species. Perhaps a similar escape from herbivory by American chestnut has contributed to abundance of chestnut
sprouts which persist in the understory at the current time. A second hypothesis which is important to consider in the context of this study is the Janzen-Connell effect; the idea that
it is advantageous for a seedling to establish away from adult trees of its own species to escape pathogens and pests which the parent tree is suffering from (Hyatt et al 2003). This
hypothesis also presents the idea that it is advantageous for a seedlings to grow in low densities, because this helps them avoid pests and pathogens (Hyatt et al 2003). In this survey,
oak saplings were likely to be growing in the same area as adult oak trees simply because oak is an abundant component of the canopy, where as chestnut sprouts were growing in the same
forest devoid of adult chestnuts. Possibly, oaks suffer more attacks from insect herbivores than chestnut because they grow in proximity to adult oaks which harbor populations of oak
pests. In addition, chestnut sprouts may be experiencing less herbivory because they are rarer than oak sprouts. This study did not measure sprout density. The results of this study
do suggest that there may be some benefit to being rare, whether it is due to the fact that having a population of extremely low numbers for a prolonged period of time might cause a
parallel decline or even extinction in co-evolved pests, or simply because abundant plants are easier for herbivores and pests to find. The present survey was fairly limited in scope,
sampling 20 trees trees from a single conservation area in western Massachusetts. It is unclear from this study whether the trends we are seeing can be explained by the Janzen-Connell
effect (that chestnut had less herbivory simply
because it is rarer than oak) or the enemy release hypothesis (that chestnut had less herbivory because there are actually fewer herbivores attacking it). To experimentally differentiate
between these two effects, further studies could be done which would sample different locations which have low and high densities of chestnuts. If the enemy release hypothesis is what
we are seeing in the case of chestnut, then the density of saplings should not matter. On the other hand, if locales with different densities of chestnut are found to experience different
rates of herbivory, then the pattern could be explained by the Janzen-Connell effect. It would be interesting to see if similar patterns would be found if this survey was expanded to
include larger sample sizes and multiple locations throughout southern New England. The findings of this survey are particularly exciting as they could provide a starting point for future
studies exploring herbivory rates on chestnut and other tree species, for example, Elm, in eastern North America which have experienced near-extinction. This study suggests that the
decline of a foundation species may have a significant impact on the interspecific interactions occuring between itself and its herbivores. These findings have broad implications in
today’s world of unparalleled ecological change (Ellison et al 2005). Chestnut is not the only foundation tree species in North America to experience a dramatic decline. Eastern hemlock
(Tsuga canadensis) has also experienced significant loss since the 1980s due to the introduced aphidlike hemlock woolly adelgid (Adelges tsugae) (Ellison et al 2005; Orwig and Foster
1998). The declines of foundation tree species such as chestnut and hemlock have large impacts on how their ecosystems function (Ellison et al 2005). The species compositions of eastern
North American forests have already been altered dramatically, and it is likely that we will have to continue to confront these changes in the future.
Figure 1. Mean percent leaf tissue loss by species, from a sample of 10 chestnut saplings and 10 oak saplings at Fitzgerald Lake Conservation Area in Northampton, MA on Oct. 23, 2009.
American chestnut experienced a mean percent herbivory of 1.54% ± 0.43%. Mean percent herbivory on red oak
was 5.80% ± 1.60%.
References Carpenter, D. and Cappuccino, N. (2005). “Herbivory, time since introduction and the invasiveness of exotic plants”. Journal of Ecology, 93: 315-321. Ellison, A., Bank, M.,
Clinton, B., Colburn, E., Elliot, K., Ford, C., Foster, D., Kloeppel, B., Knoepp, J., Lovett, G., Mohan, J., Orwig, D., Rodenhouse, N., Sobczak, W., Stinson, K., Stone, J., Swan, C.,
Thompson J., Holle, B. and Webster, J. (2005). “Loss of foundation species: consequences for the structure and dynamics of forested ecosystems”. Frontiers in Ecology and the Environment,
3: 9, 479-486. Hyatt, L., Rosenberg, M., Howard, T., Bole, G., Fang, W., Anastasia, J., Brown, K., Grella, R., Hinman, K., Kurdziel, J., Gurevitch, J. (2003). “The distance dependence
prediction of the Janzen-Connell hypothesis: a meta-analysis”. OIKOS 103: 590-602. Lewis, K., Bazzaz, F., Liao, Q. and Orians, C. (2006). “Geographic patterns of herbivory and resource
allocation to defense, growth and reproduction in an invasive biennial, Alliaria petiolata”. Ecophysiology, 148: 384-395. Orwig, D. and Foster, D. (1998). “Forest response to the introduced
hemlock woolly adelgid in southern New England, USA”. Journal of the Torrey Botanical Society, 125: 1, 60-73. Paillet, F. (1988). “Character and distribution of American chestnut sprouts
in southern New England woodlands. Bulletin of the Torrey Botanical Club, 115:1, 32-44. Paillet, F. (1984). “Growth-form and ecology of American chestnut sprout clones in northeastern
Massachusetts. Bulletin of the Torrey Botanical Club, 111: 3, 316-328.