You are on page 1of 7

Biol Invasions (2013) 15:1887–1893

DOI 10.1007/s10530-013-0430-2

INVASION NOTE

Assessment of plant community restoration following


Tree-of-Heaven (Ailanthus altissima) control by Verticillium
albo-atrum
Patrick T. Harris • Gabrielle H. Cannon •

Nathan E. Smith • Norris Z. Muth

Received: 3 September 2012 / Accepted: 2 February 2013 / Published online: 14 February 2013
Ó Springer Science+Business Media Dordrecht 2013

Abstract Tree-of-Heaven (TOH) is a highly inva- Introduction


sive woody species incurring substantial investment in
control efforts across its extensive adventive range. A A recently discovered strain of Verticillium albo-
recently isolated strain of the fungus Verticillium albo- atrum found in southern Pennsylvania by Don Davis
atrum has been found to cause near 100 % mortality of and colleagues has been found to cause near complete
TOH in laboratory and field tests. We assessed plant mortality of Tree-of-Heaven (Ailanthus altissima,
communities in experimentally infected TOH stands hereafter TOH) in artificially infected stands (Schall
5–6 years post treatment and compared them to and Davis 2009a, b). TOH is considered a noxious
uninfected control stands. We found no statistically weed in the US, Australia and in many Mediterranean
significant differences in introduced or native species regions of Europe (Kowarik and Säumel 2007;
cover between treated and control stands. Healthy PLANTS 2012; Weeds Australia 2012) and has been
TOH stands often harbor substantial populations of introduced from its native Asian range to every
native and introduced invasive species, and on continent except Antarctica (Kowarik and Säumel
average, successful control of overstory TOH by 2007). Where it has been introduced, TOH has been
V. albo-atrum did not alter vegetation cover in these found to significantly impact species richness and
communities. V. albo-atrum appears to be a promising diversity (Gaertner et al. 2009; Motard et al. 2011) and
tool in targeted control of TOH that carries a relatively its potential allelopathic effects raise concerns about
low risk of opportunistic weed replacement. longer term ecological legacies (Ding et al. 2006;
Kowarik and Säumel 2007). In part due to these and
Keywords Ailanthus altissima  Tree of Heaven  similar concerns, extensive TOH control efforts have
Verticillium  Biocontrol  Restoration ecology been implemented throughout its adventive range
(Kowarik and Säumel 2007). Current mechanical (e.g.
sapling removal, cutting) and chemical (e.g. foliar
sprays, basal bark application, hack-and-squirt and
P. T. Harris
cut-stump methods) control techniques have met
Department of Plant Biology, Southern Illinois
University, 1125 Lincoln Drive, Life Science II, with limited success and are costly, time-consuming,
Room 420, Carbondale, IL 62901-6509, USA and can pose a risk of serious non-target effects
(Swearingen and Pannill 2009). The introduction of an
G. H. Cannon  N. E. Smith  N. Z. Muth (&)
effective and host-restricted biological control agent
Department of Biology, Juniata College, 1700 Moore St.,
Huntingdon, PA 16652, USA (Schall and Davis 2009b) into the tool box of control
e-mail: muth@juniata.edu measures for TOH could be of great utility to natural

123
1888 P. T. Harris et al.

areas managers across the United States, and perhaps should be found in the smallest, most irregularly
beyond its borders. shaped sites. Furthermore, as we expect substantial
As with any introduced species control effort, to light competition in recovering sites, cover by intro-
more fully understand the efficacy of V. albo-atrum in duced species is predicted to be negatively associated
TOH site restoration, it is necessary to not only with cover by native species.
examine its effects and host-specificity with respect to
TOH, but also to document the fate of treated sites
with respect to restoration of native plant communities Methods
or desired ecosystem functions (Jäger et al. 2009;
Pearson and Ortega 2009; Flory and Clay 2009; Study species and sites
DeMeester and deB Richter 2010). Assessing site
restoration takes additional time and money and has Introduced to the United States in 1784 as an ornamental
historically been overlooked in the majority of intro- (Heisey 1990), TOH, Ailanthus altissima (Sapindales:
duced species control efforts (D’Antonio and Meyerson Simaroubaceae), is a dioecious broad-leaved tree native
2002; Thomas and Reid 2007; Reid et al. 2009). This is to eastern Asia. By means of vigorous clonal root
particularly troubling as disturbances inevitably cre- propagation, TOH can mold itself to the size of the open
ated by weed management are often colonized by space it encounters, whether a single individual in a
other opportunistic introduced species (Luken et al. sidewalk crack, or a large many-stemmed clone in a
1997; Turner et al. 2008; DeMeester and deB Richter forest gap. TOH is also capable of producing copious
2010). This reinvasion phenomenon, likened to an windborne seeds (Kowarik and Säumel 2007) and this
‘‘invasive species treadmill’’ by Thomas and Reid mixed method of population growth allows it to rapidly
(2007), is well documented and it belies the usefulness colonize disturbed sites by seed recruitment and the
of many introduced weed management projects subsequent rapid vegetative growth fills in the sur-
(McEnvoy and Coombs 2000; Buckley et al. 2007). rounding open area. To date, natural enemies of TOH
In this study we assess the vascular plant commu- found in the eastern US seem to be limited in their
nities found in former TOH stands 5–6 years post effects and, as noted above, most control efforts have
infection by V. albo-atrum. With the current study we focused on mechanical and chemical methods.
seek to address the following specific objectives and Verticillium albo-atrum (Ascomycota) is a phyto-
questions. (1) We will quantitatively characterize the pathogenic fungus that affects a broad diversity of
plant communities found in former TOH stands that plant species and is relatively common throughout its
have been experimentally infected with V. albo-atrum. wide range (including Europe, New Zealand and parts
Based on the growth habit of TOH (forming a canopy of North America) (European and Mediterranean Plant
monoculture), the high mortality caused by V. albo-atrum, Protection Organization 2007). Some strains of
and the abundance of introduced colonizing species in V. albo-atrum are significant agricultural pathogens
the study areas, we predict that infected sites will noted for developing host specificity to alfalfa,
primarily be occupied by introduced grasses, forbs, eggplant, cantaloupe, cotton, potato and other species
and shrubs. (2) We will compare the plant communi- (Correl et al. 1988). Of the numerous described strains
ties found in these infected TOH stands to untreated of V. albo-atrum, most are oligophagous systemic
control sites. As many introduced invasive grasses, xylem-infecting phytopathogens. The strain of
forbs, and shrubs co-occur with healthy TOH as dense V. albo-atrum that is the focus of the current study
substrata, we predict that TOH mortality will not result (PSU Isolate 140) was cultured by Don Davis and
in significantly different plant communities compared colleagues from TOH stands in central Pennsylvania
to healthy control stands. (3) Finally, we will deter- that exhibited severe TOH wilt and mortality. Schall
mine if variation in site size, site shape, or aspects of and Davis (2009a, b) further demonstrated that TOH is
the vegetative community influence site restoration by extremely susceptible to artificial inoculation in the
native species. Specifically, we predict the greatest lab and field. Non-target effects in the field were
native forest tree regeneration will occur in sites where largely limited to artificially inoculated striped maples
TOH mortality opens up the least amount of canopy. (Acer pennsylvanicum). Due to the clonal nature of
Therefore, the highest cover by young native trees TOH, the infection of one or a few stems in a stand

123
Assessment of plant community restoration 1889

with V. albo-atrum can result in the death of entire throughout each site from a random starting point. We
clones (Schall and Davis 2009a, b). Persistence of the oriented transects to be perpendicular to the long axis
pathogen in the soil and ability to infect secondary of the site with distances between parallel transects
TOH recruits is currently being examined (Matt varying with the size of each site. We recorded cover
Kasson, personal communication). data in 5 m linear segments along each transect and
Our treated study sites were all located in central each subsequent 5 m segment served as a separate
Pennsylvania (Blair, Franklin, Huntingdon and Mifflin observational unit. The total number of observational
counties) and consisted of 18 sites used by Schall and units per site varied proportional to its size (ultimately
Davis for experimental inoculations of TOH stands the number of observational units per site ranged from
with cultured V. albo-atrum (treatment details in 5 to 20).
Schall and Davis 2009a, b). At the time of our surveys To test for treatment effect (infected versus control
(June and July 2011), treatment sites were between 5 plots) we averaged percent cover data for each site and
and 6 years post inoculation. Cover by mature canopy performed a t test for those dependent variables that
TOH stems was effectively reduced to zero in all sites did not depart from normality (invasive species cover,
used in the current study (cover by TOH recruitment native species cover, and total vegetation cover). In
varied by site and was quantified). As vegetative addition to being bounded, the percent cover data for
response was not part of the initial study plan there are each vegetation subcategory included a fair number of
no pre-inoculation vegetation data for any treatment zero values; subsequently no transformations yielded
sites. Instead, to help us contextualize the plant an appropriate distribution of residuals for these
communities found in the treated sites, we selected variables. We therefore tested the treatment effect in
five un-treated control sites in close proximity to these cases using the nonparametric Wilcoxon signed-
treatment sites (Franklin, Huntingdon and Mifflin rank test. Given the 13 dependent variables analyzed,
counties). the Bonferroni corrected significant p value to main-
All TOH stands were located in disturbed patches tain an overall Type I error rate of 5 % was p B 0.004.
of otherwise oak dominated mixed hardwood forests. In an attempt to assess whether the degree of site
TOH stands were characterized by multiple mature restoration was affected by the size or shape of the site
TOH stems (approximately 25 cm or greater DBH) (which could dramatically affect light availability), we
and abundant smaller TOH stems (live stems in assessed correlations between the percent cover of
control sites and dead stems in treatment sites). We native non-canopy trees (log transformed) and site
limited our vegetation analyses to areas under contig- area and irregularity of site shape. Here, site shape was
uous TOH cover (or, in treated stands, contiguous treated as the length of site perimeter (in km) per site
standing dead TOH stems). Because of the history of area (in km2), where more irregularly shaped sites had
these sites as canopy gaps that were colonized by larger values and more regularly shaped sites had
TOH, there was almost no native canopy cover smaller values. The site shape variable was arcsine
present. Typically, TOH stands delineated in this square-root transformed to meet assumptions of
manner corresponded to the footprint of one or a few normality. Finally, all additional pairwise correlations
mature TOH clones with many tens to hundreds of between different vegetation types (combinations of
stems. The perimeter and area of each site were nativity and form) were also examined in an effort to
estimated using a GPS unit and the averages of three detect possible cases of facilitation or inhibition.
independent estimates of these variables were used in
subsequent statistical analyses.
Results
Vegetation surveys and statistical analyses
Five to six years after infection with V. albo-atrum,
Surveys were conducted to estimate percent cover of with the exception of mature TOH cover (effectively
native and introduced seed plants for each of the zero in treated stands), plant communities in recover-
following vegetation forms: grasses, forbs, shrubs, ing TOH stands were statistically indistinguishable
vines, and tree recruits. Five parallel line-transects of from control sites (Fig. 1). None of the differences
up to 25 m in length were systematically spaced between infected and control sites were statistically

123
1890 P. T. Harris et al.

significant following Bonferroni correction (Wilcoxon with cover by introduced species overall (r = 0.18,
signed-rank tests). Only the native forb category was p = 0.46; r = -0.08, p = 0.76 for site size and shape
significant prior to correction (p = 0.048), where respectively). The relationship between overall cover
infected sites tended to have greater native forb cover by native species and overall cover by introduced
compared to treated sites (Fig. 1). As with control species, though weakly negative, could not be distin-
TOH stands, experimentally infected TOH stands had guished from a non-zero relationship at a 95 %
appreciable cover of most vegetation types, including confidence level (r = -0.32, p = 0.30). Similarly,
a mix of introduced and native species (Fig. 1). when broken down by vegetation form, there were no
Across our 18 treatment sites we had a substantial statistically significant associations between native
range of site sizes (5.0 9 10-3–20.6 9 10-2 ha, and introduced species cover.
mean = 8.71 9 10-2 ha) and shapes (0.6–10.8 km/ha,
mean = 1.83 km/ha, shape calculated as perimeter/
area as described in methods). However, correlations Discussion
between native tree cover and site area (r = -0.25,
p = 0.31) and site shape (r = 0.21, p = 0.42) were not In the short-term response, Verticillium-treated TOH
statistically different from zero. Variation in site size stands do not necessarily become dominated by
and shape also had no significant association with cover introduced species, or by any particular vegetation
by native species overall (r = -0.31, p = 0.21; r = type (e.g. grass, forbs, shrubs, or vines; Fig. 1). While
0.05, p = 0.84 for site size and shape respectively) or there certainly exists an opportunity for treated TOH

Fig. 1 Mean percent cover of native and introduced vegetation statistically differ for any of the categories. See sections
by form in Verticillium-infected and control TOH stands. Bars ‘‘Methods’’ and ‘‘Results’’ for details on statistical analysis
represent standard error. Control and infected sites did not and significance

123
Assessment of plant community restoration 1891

stands to be replaced by other undesirable introduced (2003) found that native species often came to
weeds, our data suggest that most healthy TOH stands dominate TOH stands treated by various herbicide
already have relatively high cover of various intro- applications without supplemental sowing (it should
duced species and removing TOH canopies has be noted that, as with our study, they did not include a
surprisingly little impact on these understory commu- supplemental sowing treatment for comparison). Due
nities. Where there is already an understory of to differences in the methods of quantifying cover, the
honeysuckle or Microstegium vimineum, these species lack of statistical analysis of vegetation response in the
seem likely to remain, but removing the TOH canopy Burch and Zedaker (2003) study, and the lack of pre-
doesn’t necessarily lead to their expansion. Of course treatment data in our current study, it is difficult to
if these invasive species are very near to, but not assess the degree to which the findings of these two
currently a component of potential treatment sites, studies are consistent. It is worth pointing out that
treatment activities could tip the balance in favor of neither study detected a clear trend towards coloniza-
their colonization. We also note that the data from this tion by introduced species. However, the degree to
initial assessment of vegetation response (less than a which longer-term responses or treatment in more
decade since treatment) are preliminary and a longer- disturbed habitats produces similar results is as yet
term analysis will be necessary to determine the unknown as both of these studies assessed only the
responses of woody vegetation and the degree of short-term response of TOH control in largely undis-
successful site restoration. turbed contexts. While the average response in our
With respect to site size and shape, despite large study demonstrated no statistically significant differ-
differences in these attributes, we did not detect any ences in colonization by native or introduced species,
significant relationships between these factors and any we did note that several sites became dominated by
proxy for site restoration success (e.g. no observed introduced species cover. As such, it would be prudent
increase in native species cover or decrease in to perform a careful evaluation of nearby seed or
introduced species cover). Furthermore, our data parent-material sources (including equipment as a
suggest that, in the average treated site, native species potential vector) before large-scale treatment of TOH,
cover was independent of introduced species cover. regardless of the methods employed. Variation in the
Verticillium albo-atrum is extremely efficient at diversity and abundance of herbivores and seed
causing TOH mortality (Schall and Davis 2009a, b), predators could also impact vegetation response and
but appears to vary substantially in its rate of spread. should be considered in specific restoration efforts
Presumably this variation is due in large part to (e.g. Davidson 1993; Côté et al. 2004; Pedersen and
whether spread occurs systemically, through con- Wallis 2004). Likewise, a pre-treatment assessment of
nected vascular tissues, or has to jump to unconnected native tree recruitment or advanced regeneration could
stems through the soil or by intermediary vectors. be helpful in determining whether treatment of
From a management perspective, this restricted rate of V. albo-atrum should be accompanied by supplemen-
spread may be beneficial in cases where specifically tal seeding or planting of native species.
targeted treatments are desirable as it may allow for As is the case with many introduced species, the
the inoculation of isolated stands that have limited effects of TOH on wildlife and non-vegetative taxa are
cover of invasive species in the treatment area, while poorly understood and rarely rigorously documented.
leaving stands in more degraded areas temporarily With respect to the current study, the removal of TOH
intact. On the other hand, if management objectives led to the formation of a canopy gap that may have been
are intended to more broadly reduce TOH as much as similar to that which allowed for colonization by TOH
possible, V. albo-atrum may need to be applied at a in the first instance. As such, these gaps would likely
relatively high spatial frequency to increase the affect wildlife in many of the same ways as other forest
chances of infecting most of the separate vascular gaps and, more generally, early successional habitats
networks present in the treatment area. (Thompson and DeGraaf 2001; Felix et al. 2004).
Whether or not supplemental sowing of native Whether or not the habitats that immediately arise from
species would improve the restoration of these or TOH treatment are valuable themselves, or are only a
similarly treated TOH stands is largely unknown. temporary means to establishing a more mature forest,
Using a visual estimation of cover, Burch and Zedaker depends on the specific management goals. The

123
1892 P. T. Harris et al.

response of treated TOH stands in otherwise open or European and Mediterranean Plant Protection Organization
edge habitat may differ from the responses observed (2007) Verticillium albo-atrum and V. dahliae on hop.
EPPO Bulletin 37:3. http://www.eppo.int/QUARANTINE/
here, and the resulting habitats would likely bear fungi/Verticillium_alboatrum/VERTSP_ds.pdf. Accessed
greater resemblance to those characterized by frequent 1 August 2012
disturbances like roadsides and right-of-ways. Felix AB, Campa H, Millenbah KF, Winterstein SR, Moritz WE
In summary, over the short-term it appears that on (2004) Development of landscape-scale habitat-potential
models for forest wildlife planning and management. Wildl
average, treating TOH stands with V. albo-atrum Soc Bull 32(3):795–806
effectively controls mature TOH stems. Furthermore, Flory SL, Clay K (2009) Invasive plant removal method
the persistence of Verticillium in the soil suggests that determines native plant community responses. J Appl Ecol
newly establishing TOH from the seedbank should 4:434–442. doi:10.1111/j.1365-2664.2009.01610.x
Gaertner M, Breeyan AD, Hui C, Richardson DM (2009)
also be vulnerable. Future studies should focus on the Impacts of alien plant invasions on species richness in
longer-term response of TOH and native vegetation in Mediterranean-type ecosystems: a meta-analysis. Prog
these sites, as well as the response of TOH sites that Phys Geogr 33:319–338. doi:10.1177/0309133309341607
occur in more open and disturbed areas. Heisey RM (1990) Evidence for allelopathy by tree-of-heaven
(Ailanthus altissima). J Chem Ecol 16:2039–2055. doi:
10.1007/BF01020515
Acknowledgments We are grateful to Don Davis and Mark Jäger H, Kowarik I, Tye A (2009) Destruction without extinc-
Kasson for advice and assistance throughout the course of this tion: long-term impacts of an invasive tree species on
project. Thanks also to Gene Odato and Jeff Krause for field site Galapagos highland vegetation. J Ecol 97:1252–1263. doi:
access. Thanks to Rebecca Coleman for assistance with 10.1111/j.1365-2745.2009.01578.x
fieldwork. Funding for this research was graciously provided Kowarik I, Säumel I (2007) Biological flora of Central Europe:
by the William J. von Liebig Summer Research Fellowship and Ailanthus altissima (Mill.) Swingle. Perspect Plant Ecol
the D.C. Goodman Summer Research Award. 8:207–237. doi:10.1016/j.ppees.2007.03.002
Luken JO, Kuddes LM, Tholemeier TC (1997) Response of
understory species to gap formation and soil disturbance in
References Lonicera maackii thickets. Restor Ecol 5:229–235. doi:
10.1046/j.1526-100X.1997.09727.x
Buckley YM, Bolker BM, Rees M (2007) Disturbance, invasion McEnvoy PB, Coombs EM (2000) Why things bite back:
and re-invasion: managing the weed-shaped hole in dis- unintended consequences of biological weed control. In:
turbed ecosystems. Ecol Lett 10:809–817. doi:10.1111/ Follett PA, Duan JJ (ed) Nontarget effects of biological
j.1461-0248.2007.01067.x control. Springer, pp 167–194
Burch PL, Zedaker SM (2003) Removing the invasive tree Motard E, Muratet A, Clair-Maczulajtys D, Machon N (2011)
Ailanthus altissima and restoring natural cover. J Arboric Does the invasive species Ailanthus altissima threaten
29(1):18–24 floristic diversity of temperate peri-urban forests? C R Biol
Correl JC, Gordon TR, McCrain AH (1988) Vegetative com- 334(12):872–879. doi:10.1016/j.crvi.2011.06.003
patibility and pathogenicity of Verticillium albo-atrum. Pearson DE, Ortega YK (2009) Managing invasive species in
Phytopathology 78:1017–1021. doi:10.1094/Phyto-78- natural areas: moving beyond control. In: Kingely RV (ed)
1017 Weeds: management. Economic Impacts and Biology,
Côté SD, Rooney TP, Tremblay JP, Dussault C, Waller DM Hauppauge, pp 1–21
(2004) Ecological impacts of deer overabundance. Annu Pedersen BS, Wallis AM (2004) Effects of white-tailed deer
Rev Ecol Syst 113–147 herbivory on forest gap dynamics in a wildlife preserve,
D’Antonio C, Meyerson LA (2002) Exotic plant species as Pennsylvania, USA. Nat Areas J 24(2):82–94
problems and solutions in ecological restoration: a syn- PLANTS (2012) Ailanthus altissima. USDA Natural Resources
thesis. Restor Ecol 10:703–713. doi:10.1046/j.1526-100X. Conservation Service. http://plants.usda.gov/java/profile?
2002.01051.x symbol=aial. Accessed 30 July 2012
Davidson DW (1993) The Effects of herbivory and granivory on Reid AM, Morin L, Downey PO, French K, Virtue JG (2009)
terrestrial plant succession. Oikos 68:23–35 Does invasive plant management aid the restoration of
DeMeester JE, deB Richter D (2010) Restoring restoration: natural ecosystems? Biol Conserv 142:2342–2349. doi:
removal of the invasive plant Microstegium vimineum from 10.1016/j.biocon.2009.05.011
a North Carolina wetland. Biol Invasions 12:781–793. doi: Schall MJ, Davis DD (2009a) Ailanthus altissima wilt and
10.1007/s10530-009-9481-9 mortality: etiology. Plant Dis 93:747–751. doi:10.1094/
Ding J, Wu Y, Zheng H, Fu W, Reardon R, Liu MM (2006) PDIS-93-7-0747
Assessing potential biological control of the invasive plant, Schall MJ, Davis DD (2009b) Verticillium wilt of Ailanthus
tree-of-heaven, Ailanthus altissima. Biocontrol Sci Techn altissima: susceptibility of associated tree species. Plant
16:547–566. doi:10.1080/09583150500531909 Dis 93:1158–1162. doi:10.1094/PDIS-93-11-1158

123
Assessment of plant community restoration 1893

Swearingen J, Pannill P (2009) Least wanted: Ailanthus altiss- Turner PJ, Scott JK, Spafford H (2008) The ecological barriers
ima. Plant Conservation Alliance. http://www.nps.gov/ to the recovery of bridal creeper (Asparagus asparagoides
plants/alien/fact/aial1.htm. Accessed 28 July 2012 (L.) Druce) infested sites: impacts on vegetation and the
Thomas MB, Reid AM (2007) Are exotic natural enemies an potential increase in other exotic species. Austral Ecol
effective way of controlling invasive plants? Trends Ecol 33:713–722. doi:10.1111/j.1442-9993.2008.01839.x
Evol 22:447–453. doi:10.1016/j.tree.2007.03.003 Weeds Australia (2012) Noxious weeds list for Australian states
Thompson FR III, DeGraaf RM (2001) Conservation approa- and territories. Australian Weeds Committee. http://www.
ches for woody, early successional communities in the weeds.org.au/docs/weednet6.pdf. Accessed 30 July 2012
eastern United States. Wildl Soc Bull 29(2):483–494

123

You might also like