Content uploaded by Jared W Westbrook
Author content
All content in this area was uploaded by Jared W Westbrook on Sep 26, 2019
Content may be subject to copyright.
Resistance to Phytophthora cinnamomi in American Chestnut (Castanea dentata)
Backcross Populations that Descended from Two Chinese Chestnut
(Castanea mollissima) Sources of Resistance
Jared W. Westbrook,
1,†
Joseph B. James,
2
Paul H. Sisco,
1
John Frampton,
3
Sunny Lucas,
4
and Steven N. Jeffers
5,†
1
The American Chestnut Foundation, Asheville, NC 28804
2
The American Chestnut Foundation and Chestnut Return Farms, Seneca, SC 29672
3
Department of Forestry and Environmental Resources, North Carolina State University, Raleigh, NC 27695
4
United States Department of Agriculture Forest Service Resistance Screening Center, Asheville, NC 28806
5
Department of Plant and Environmental Sciences, Clemson University, Clemson, SC 29634
Abstract
Restoration of American chestnut (Castanea dentata) depends on combin-
ing resistance to both the chestnut blight fungus (Cryphonectria parasitica)
and Phytop hthora cinnamomi, which causes Phytophthora root rot, in a di-
verse population of C. dentata. Over a 14-year period (2004 to 2017), sur-
vival and root health of American chestnut backcross seedlings after
inoculation with P. cinnamomi were compared among 28 BC
3
,66BC
4
,
and 389 BC
3
F
3
families that descended from two BC
1
trees (Clapper and
Graves) with different Chinese chestnut grandparents. The 5% most resis-
tant Graves BC
3
F
3
families survived P. cinnamomi infection at rates of 75
to 100% but had mean root health scores that were intermediate between
resistant Chinese chestnut and susceptible American chestnut families.
Within Graves BC
3
F
3
families, seedling survival was greater than survival
of Graves BC
3
and BC
4
families and was not genetically correlated with
chestnut blight canker severity. Only low to intermediate resistance to
P. cinnamomi was detected among backcross descendants fromthe Clapper
tree. Results suggest that major-effect resistance alleles were inherited by
descendants from the Graves tree, that intercrossing backcross trees enhan-
ces progeny resistance to P. cinnamomi, and that alleles for resistance to
P. cinnamomi and C. parasitica are not linked. To combine resistance to both
C. parasitica and P. cinnamomi, a diverse Graves backcross population will
be screened for resistance to P. cinnamomi, survivors bred with trees se-
lected for resistance to C. parasitica, and progeny selected for resistance
to both pathogens will be intercrossed.
Keywords: backcross breeding, chestnut blight, Cryphonectria parasitica,
heritability, host resistance, ink disease, Phytophthora root rot.
Although the introduction of the fungus that causes chestnut blight
(Cryphonectria parasitica (Murr.) Barr) in the early 20th century
eventually caused the functional extinction of American chestnut trees
(Castanea dentata (Marsh.) Borkh.) in the eastern United States,
Phytophthora root rot, caused by the soilborne oomycete Phytophthora
cinnamomi Rands (Crandall et al. 1945; Tucker 1933), is thought to
have extirpated American chestnut trees from lower-elevation forests
in the southeastern United States prior to the introduction of Crypho-
nectria parasitica (Anagnostakis 2012; Crandall et al. 1945; Freinkel
2007; Zentmyer 1980). A disease suspected to be Phytophthora root
rot (also known as ink disease) was first reported in 1824 killing
Allegheny chinquapin shrubs (Castanea pumila Mill.) in Georgia
(Anagnostakis 2001, 2012), and American chestnut is also highly
susceptible to P. cinnamomi (Crandall et al. 1945; Jeffers et al.
2009). Therefore, extensive mortality of American chestnut trees in
the southeastern United States in the early to mid-1800s has been spec-
ulated to be from Phytophthora root rot (Crandall et al. 1945; Zentmyer
1980). Today, Phytophthora root rot has the potential to limit reintro-
duction of American chestnut trees with improved resistance to
Cryphonectria parasitica across large areas of the native range of
Castanea dentata (Anagnostakis 2001; Jacobs 2007). Whereas
American chestnut resprouts from the roots after stems are killed
by chestnut blight, the whole plant is killed by Phytophthora root rot.
Until recently, Phytophthora root rot on American chestnut was
reported to be caused only by P. cinnamomi (Crandall et al. 1945;
Tucker 1933). The source, location, and timing of the earliest intro-
duction of P. cinnamomi to the United States is not known but
P. cinnamomi is speculated to have been imported into the southeast-
ern United States on exotic plants from southeastern Asia, where
P. cinnamomi is thought to have originated (Crandall and Gravatt
1967; Crandall et al. 1945; Zentmyer 1980, 1988). Currently, P. cinna-
momi is the most common and virulent species of Phytophthora on
American chestnut but P. cambivora (most likely P. ×cambivora)(Jung
et al. 2017), P. cryptogea, and P. heveae recently were isolated from
American and backcross hybrid chestnut seedlings and have been
found to cause lesions on roots of American chestnut seedlings after
artificial inoculation (Sharpe 2017).
The chestnut blight fungus produces propagules that are both wind-
borne and splash dispersed, and this pathogen is ubiquitous throughout
the native range of American chestnut (Anagnostakis 2012; Freinkel
2007). By contrast, P. cinnamomi is a soilborne pathogen with a ran-
dom distribution in time and space (Crandall et al. 1945; Meadows
and Jeffers 2011). Soil moisture is essential for the production of spo-
rangia and release of zoospores by P. cinnamomi (Zentmyer 1980).
Phytophthora root rot on American chestnut seedlings was most dam-
aging in moist soils in controlled experiments (Rhoades et al. 2003).
P. cinnamomi spreads to host plants in water or soil containing infec-
tive propagules (i.e., motile zoospores or chlamydospores), by growth
of plant roots into infested soil, or by root-to-root contact (Ristaino
and Gumpertz 2000; Zentmyer 1980). The current geographic range
of P. cinnamomi in the eastern United States overlaps with the range
of American chestnut from Alabama to southern Pennsylvania. P. cin-
namomi is not prevalent north of 40° N latitude at the present time, pre-
sumably due to the adverse effects of prolonged cold temperatures
†
Corresponding authors: J. W. Westbrook; E-mail: jared.westbrook@acf.org;
and S. N. Jeffers; E-mail: sjffrs@clemson.edu
Funding: This material is based on research that was supported by The Amer-
ican Chestnut Foundation and by the United States Department of Agriculture
National Institute of Food and Agriculture, under project numbers SC-
1700309, SC-1700445, SC-1700481, and SC-1700534 at Clemson Univer-
sity. Technical Contribution No. 6715 of the Clemson University Experiment
Station.
The author(s) declare no conflict of interest.
Accepted for publication 1 February 2019.
©2019 The American Phytopathological Society
Plant Disease / July 2019 1631
Plant Disease •2019 •103:1631-1641 •https://doi.org/10.1094/PDIS-11-18-1976-RE
during winter (Balci et al. 2007; Benson 1982). However, as the cli-
mate warms, the range of P. cinnamomi is predicted to expand into
the northeastern region of the United States by 2080 (Burgess et al.
2017).
Recent efforts to restore the American chestnut to its native range
have focused on improving resistance to Cryphonectria parasitica in
American chestnut populations (Anagnostakis 2012). Since 1983,
The American Chestnut Foundation (TACF) has pursued backcross
breeding to generate hybrids that combine resistance to C. parasitica
from Chinese chestnut (Castanea mollissima Blume) with the
timber-type form of American chestnut (Burnham et al. 1986). In
a TACF breeding program, two sources of resistance to C. parasitica,
which descended from two distinct C. mollissima trees, have been ad-
vanced to the third backcross generation (BC
3
) with C. dentata. Two
generations of intercrossing between backcross trees selected for re-
sistance to C. parasitica have been performed to enhance this resis-
tance in the BC
3
F
2
and BC
3
F
3
generations (Steiner et al. 2017).
Chinese chestnut also is resistant to P. cinnamomi (Crandall et al.
1945; Jeffers et al. 2009, 2012). Thus, American chestnut backcross
descendants selected for resistance to C. parasitica potentially
inherited alleles for P. cinnamomi resistance from C. mollissima.
However, TACF has not selected for resistance to P. cinnamomi in
backcross generations. Several coauthors of this study (i.e., S. N.
Jeffers, J. B. James, and P. H. Sisco) began screening seedlings of
American chestnut backcross families for resistance to P. cinnamomi
in 2004 after P. cinnamomi had killed over 80% of BC
2
American
chestnut seedlings growing in a planting in Oconee County, SC be-
tween 2001 and 2003 (James 2011a,b; Laurie 2014). In annual trials
conducted from 2004 to 2011, these collaborators evaluated seedling
resistance to P. cinnamomi among American chestnut BC
3
and BC
4
progeny of BC
2
and BC
3
trees selected for resistance to C. parasitica
(Jeffers et al. 2009, 2012). In 2012, this group began screening seed-
lings in BC
3
F
3
families that were derived from open pollination
among BC
3
F
2
trees. TACF collaborated with J. Frampton (another
coauthor) at North Carolina State University (NCSU) to evaluate ad-
ditional BC
3
F
3
families from 2014 to 2016. Beginning in 2017, eval-
uation of seedlings in BC
3
F
3
families was conducted by TACF in
collaboration with personnel at the United States Department of Ag-
riculture (USDA) Forest Service Resistance Screening Center in
Asheville, NC.
The overall objective of this project was to evaluate first-year hy-
brid American chestnut seedlings for resistance to P. cinnamomi un-
der standardized conditions. If resistance was identified, a secondary
objective was to identify the source or sources of the resistance. We
present results of evaluating seedlings of American chestnut, Chinese
chestnut, and American chestnut backcross progeny for resistance to
P. cinnamomi over 14 growing seasons (2004 to 2017). Genetic var-
iation in resistance to P. cinnamomi was estimated among three gen-
erations of American chestnut backcross families (BC
3
,BC
4
, and
BC
3
F
3
) that descended from two BC
1
trees. This dataset enabled
us to test for loss of P. cinnamomi resistance with successive back-
crosses to C. dentata and for increases in resistance through inter-
crossing backcross trees. Among BC
3
F
3
families, P. cinnamomi
resistance was compared for a subset of families screened at multiple
locations using different isolates of P. cinnamomi and different pro-
cedures for disease phenotyping. A genetic correlation between resis-
tance to P. cinnamomi and C. parasitica was estimated among
American chestnut BC
3
F
3
families that were screened for resistance
to both pathogens. Implications for breeding to combine resistance
to P. cinnamomi and C. parasitica are discussed.
Materials and Methods
Sources of seed and research trial locations. The American
chestnut backcross seedlings evaluated for resistance to P. cinna-
momi were descendants of the Clapper and Graves sources of resis-
tance to C. parasitica. Clapper and Graves are two BC
1
trees ((C.
mollissima ×C. dentata) ×C. dentata) with different C. mollis-
sima grandparents (Anagnostakis 2007; Clapper 1952; Hebard
2006). A subset of BC
2
and BC
3
descendants of Clapper and
Graves, which were selected for resistance to C. parasitica at
TACF Meadowview Research Farms, were crossed with suscep-
tible C. dentata parents to generate BC
3
and BC
4
progeny that
were evaluated for resistance to P. cinnamomi. In total, progeny
of 10 of 41 Clapper BC
2
selections, 20 of 83 Clapper BC
3
selec-
tions,18of41GravesBC
2
selections, and 46 of 71 Graves BC
3
selections were evaluated for resistance to P. cinnamomi.
The BC
3
F
3
seed were obtained from TACF Meadowview Re-
search Farms through open pollination of 124 Clapper and 265
Graves BC
3
F
2
mothers by neighboring BC
3
F
2
trees. Similarly, the
BC
3
F
2
parents were generated through open pollination among
BC
3
trees. To put the number of BC
3
F
3
families that were evaluated
for P. cinnamomi resistance into context, approximately 36,000
BC
3
F
2
progeny of 83 Clapper BC
3
trees and 27,000 BC
3
F
2
progeny
of 71 Graves BC
3
trees have been planted in Meadowview seed
orchards since 2002. These approximately 63,000 BC
3
F
2
trees were
artificially inoculated with C. parasitica, and the most susceptible
trees were culled based on severity of blight canker development
(Steiner et al. 2017). At the time of the writing of this article, approx-
imately 5,000 Clapper BC
3
F
2
trees and 4,250 Graves BC
3
F
2
trees
remained in Meadowview seed orchards. The final objective is to
cull all but 1% of trees that are most resistant to C. parasitica or
P. cinnamomi (i.e., retain approximately 600 trees). Although we have
evaluated a small subset of the total number of potential BC
3
F
3
fam-
ilies for P. cinnamomi resistance, we have likely screened BC
3
F
3
grand-progeny from a majority of the BC
3
trees selected for blight
resistance. The BC
3
F
2
mothers whose progeny were screened for
resistance to P. cinnamomi were themselves the progeny of 41 of
71 Graves BC
3
selections and 34 of 83 Clapper BC
3
selections. It
is likely that grand-progeny of additional BC
3
selections have been
represented in the screening through the unknown paternal contribu-
tion in the BC
3
F
2
and BC
3
F
3
generations. In this article, seedlings de-
rived from open pollination of a given mother tree or from a specific
controlled pollination are referred to as a family. Seed were collected
in the fall and stratified in moist peat in plastic bags over the winter at
4 to 8°C.
Chestnut seedlings were evaluated for resistance to P. cinnamomi
at three locations: Clemson University personnel conducted trials in
collaboration with colleagues at Chestnut Return Farms (Oconee
County, SC); NCSU personnel conducted trials on the main campus
in Raleigh, NC; and TACF personnel conducted trials in collabora-
tion with colleagues at the USDA Forest Service Resistance Screen-
ing Center (Asheville, NC). At each location, different inoculation
and phenotyping methods were used (detailed below) to insure that
results were not artifacts of a single location, inoculation procedure,
or evaluation method. To compare family rankings for P. cinnamomi
resistance between pairs of locations and methods, subsets of 53
BC
3
F
3
families were evaluated by Clemson University and NCSU
in 2014, 2015, and 2016. Another 17 BC
3
F
3
families were evaluated
by Clemson University and TACF with the USDA Forest Service in
2017. In each year and at each location, resistant Chinese chestnut
and susceptible American chestnut seedlings were included as
controls.
Trials conducted by Clemson University at Chestnut Return
Farms: 2004 to 2017. The 14 annual trials conducted at Chestnut
Return Farms have been summarized, including dates of planting, in-
oculation, and evaluation of Phytophthora root rot phenotypes;
lengths of time between when seedlings were planted, inoculated,
and evaluated; and numbers of seedlings and families evaluated
(Table 1). Among 505 American chestnut backcross families evalu-
ated during this time, 62 were descendants of Clapper and 166 were
descendants of Graves. Detailed analyses of resistance to P. cinna-
momi within hybrid and backcross families that descended from ad-
ditional Chinese chestnut and Japanese chestnut trees, which were
included in these trials, are beyond the scope of this article. Prelim-
inary results of P. cinnamomi resistance in hybrid and backcross
American chestnut seedlings, including those that descended from
sources of resistance other than Clapper or Graves, have been
reported (Jeffers et al. 2009, 2012).
Seed were sowed outdoors in early to mid-April (plus late March
2008) in 568-liter tubs (Structural Foam Stock Tank FG424500BLA:
1632 Plant Di sease / Vol. 103 No. 7
147 cm long by 99 cm wide by 64 cm tall; Rubbermaid Commercial
Products, High Point, NC) when radicles had emerged from approx-
imately 50% of the seed. The tubs contained a soilless growing mix
composed primarily of peat and bark (Fafard 3B Mix; currently pro-
duced by Sun Gro Horticulture, Agawam, MA). A subset of seed
from each family was planted in 3 to 10 tubs. Between 1 and 25 seeds
from each family (mean = 6) were planted in each tub, depending on
the year. Seed from individual families were planted together in
a row, with approximately 1 to 3 cm between seeds and at a depth
of 2 to 4 cm. The position of families within tubs was randomized
among tubs. Tubs were watered as needed throughout the growing
season. Once seedlings had emerged, slow-release fertilizer (Osmo-
cote; The Scotts Company, Marysville, OH) was added to each tub in
late April or early May.
Seedlings were inoculated with a combination of two isolates of
P. cinnamomi (JJ-1 and JJ-52, both mating type A2) that were isolated
in 2003 from backcross chestnut seedlings with Phytophthora root
rot growing at Chestnut Return Farms. These isolates are stored in
a permanent collection that is maintained by S. N. Jeffers at Clemson
University. The two isolates were grown independently on sterile rice
grains (Holmes and Benson 1994; Meadows et al. 2011) in 2004 to
2007 and then on sterile vermiculite moistened with V8 juice broth
(Roiger and Jeffers 1991) from 2007 to 2017. Both types of inoculum
were used in 2007. Equal volumes of inoculum containing each iso-
late were thoroughly mixed before inoculation. Seedlings were inoc-
ulated 12 to 16 weeks after planting, usually in early to mid-July
(Table 1). To inoculate, furrows (2 to 3 cm deep and 95 to 97 cm
long) were made between every other row of seedlings, and inoculum
(i.e., 10 ml of rice or 50 ml of V8-vermiculite) was sprinkled in the
furrow, evenly distributed, and covered. After inoculation, seedlings
in all tubs were watered to incorporate the inoculum and prevent it
from desiccating. Tubs were flooded each year for a 6- to 8-h period
one to three times in July or August to enhance disease development.
At 20 to 29 weeks (mean = 24 weeks) after inoculation, dormant
seedlings were evaluated for disease severity (Table 1). Each seed-
ling was removed carefully from a tub to keep the roots intact and
washed in tap water. On each seedling, the roots and the lower stem
were inspected for typical symptoms of Phytophthora root rot. Dis-
ease severity was scored on a 0-to-3 scale, where 0 = no lesions on
roots, plant healthy; 1 = lesions on some lateral roots or plant stunted;
2 = lesions on the tap root or severe root rot on lateral roots; and 3 =
100% root rot, plant dead (Jeffers et al. 2009).
Each year, one to three diseased seedlings and approximately 1 li-
ter of container mix from each tub were collected and taken to the lab
to confirm the presence of P. cinnamomi. Root systems were washed
again under running tap water and blotted dry. Tissue pieces from
advancing lesions were embedded in PARPH-V8 selective me-
dium, and isolation plates were placed at 20°C for 5 to 7 days
(Ferguson and Jeffers 1999; Jeffers 2015). Subsamples of container
mix from each tub were assayed with a baiting bioassay using camellia
and rhododendron leaf pieces as baits (Ferguson and Jeffers 1999;
Meadows et al. 2011).
Trials conducted by NCSU: 2014 to 2016. During April and May
of each year, 20 stratified seeds from each American chestnut, Chi-
nese chestnut, and American chestnut backcross family were sowed
individually into D40 Deepots (650 ml, 25 × 6 cm; Stuewe & Sons,
Tangent, OR) containing a 1:1 peat-perlite medium (Table 2). After
initial growth at TACF Meadowview Research Farms in Virginia,
seedlings were transferred to the Horticultural Field Lab at NCSU
in Raleigh, NC on 24 June 2014, 22 May 2015, and 10 June 2016.
Seedlings were grown outdoors under 40% shade cloth in a com-
pletely randomized design and were inoculated 7 to 10 weeks after
planting by inserting three infested rice grains about 2.5 cm deep into
separate holes in the medium of each Deepot and covering the grains.
The rice grains had been colonized by a single isolate of P. cinna-
momi (23ss04) obtained from a Fraser fir tree (Abies fraseri (Pursh)
Poir.) in North Carolina with Phytophthora root rot (Frampton et al.
2013). Eight weeks after inoculation, surviving seedlings were rein-
oculated with the same isolate of P. cinnamomi to ensure that they
had been challenged sufficiently by the pathogen. To promote dis-
ease development, seedlings were irrigated from above three times
daily for 20 min (approximately 1.5 cm/day). Seedlings were fertil-
ized periodically with 15-16-17 (N-P-K) Peters Peat-lite Special con-
taining micronutrients (The Scotts Company) in an aqueous solution
Table 1. Summary of 14 consecutive annual trials to evaluate seedlings of American, Chinese, and backcross chestnut families for resistance to Phytophthora
cinnamomi at Chestnut Return Farm in Oconee County, SS based on the percentage of seedlings surviving at the end of each trial
Dates on which seedlings were
a
Durations
(weeks)
b
Number of
c
Number of families
d
One-way ANOVA:
P>Fvalues
e
Year Planted Inoculated Evaluated P-I I-E P-E Tubs Seeds BC Am Ch Total Am vs. Ch
f
BC families
g
2004 14 Apr 04 08 Jul 04 12 Jan 05 12 27 39 2 508 6 1 2 9 NC NC
2005 05 Apr 05 28 Jun 05 20 Dec 05 12 25 37 6 1,250 26 1 1 28 <0.001 <0.001
2006 05 Apr 06 14 Jul 06 13 Dec 06 14 22 36 6 1,031 23 1 3 27 NC <0.001
2007 04 Apr 07 19 Jul 07 18 Dec 07 15 22 37 7 1,545 42 5 1 48 <0.001 <0.001
2008 27 Mar 08 14 Jul 08 17 Dec 08 16 22 38 7 1,279 23 4 1 28 <0.001 <0.001
2009 08 Apr 09 09 Jul 09 28 Jan 10 13 29 42 7 1,020 32 3 2 37 <0.001 <0.001
2010 09 Apr 10 22 Jul 10 07 Dec 10 15 20 35 9 1,502 54 1 1 56 <0.001 <0.001
2011 05 Apr 11 11 Jul 11 13 Dec 11 14 22 36 10 1,633 48 1 1 50 <0.001 <0.001
2012 02 Apr 12 12 Jul 12 09 Jan 13 14 26 40 13 1,914 76 3 1 80 <0.001 <0.001
2013 17 Apr 13 30 Jul 13 18 Dec 13 15 20 35 11 1,480 71 1 1 73 <0.001 <0.001
2014 02 Apr 14 10 Jul 14 09 Dec 14 14 22 36 13 1,869 32 1 1 34 <0.001 <0.001
2015 08 Apr 15 07 Jul 15 13 Jan 16 13 27 40 13 1,639 29 1 1 31 <0.001 <0.001
2016 06 Apr 16 06 Jul 16 12 Jan 17 13 27 40 11 1,210 23 1 1 25 <0.001 <0.001
2017 13 Apr 17 06 Jul 17 23 Jan 18 12 29 41 5 645 20 1 1 22 <0.001 <0.001
Total …………………18,525 505 25 18 548 ……
a
Abbreviations: April (Apr), July (Jul), January (Jan), Jun (June), December (Dec), and March (Mar) in years 2004 (04) through 2018 (18).
b
Lengths of time in weeks between activities involving seedlings: P = planted, I = inoculated, and E = evaluated.
c
Number of 568-liter tubs planted with seedlings and numbers of seeds planted each year.
d
Numbers of different families planted each year. A family is a specific chestnut genotype: Am = open-pollinated American chestnut, Ch = open-pollinated
Chinese chestnut, and BC = open-pollinated and specific crosses of (American chestnut × Chinese chestnut) × American chestnut backcross hybrids.
e
Results from statistical analyses by one-way analysis of variance (ANOVA): P>Fis the probability of a greater Fstatistic occurring. Analyses were not
conducted for some data (NC) because of insufficient replication.
f
Results showing the significant difference in survival between the American and Chinese chestnut families each year based on a single-degree-of-freedom
linear contrast.
g
Results showing the significant difference in survival among only the hybrid chestnut families evaluated each year based on a one-way ANOVA of a subset
of the data from this trial.
Plant Disease / July 2019 1633
prepared to deliver N at 200 ppm. Symptom development was visu-
ally assessed 8 and 16 weeks after inoculation in 2014 and biweekly
for 16 weeks in 2015 and 2016. Foliage health was assessed by esti-
mating percent shoot (leaf and stem) necrosis on a 0-to-100 scale in
increments of 10 but also including 1, 2, and 95% levels. Seedlings
were considered dead if 100% of the shoot was necrotic and alive
(i.e., surviving) if otherwise. Each year, a control group of seedlings
from each Chinese chestnut, American chestnut, and BC
3
F
3
family
was not inoculated with P. cinnamomi and served as controls to es-
timate mortality rates in the absence of Phytophthora root rot.
Trial conducted by TACF and the USDA Forest Service in
2017. Forty seeds from each American chestnut, Chinese chestnut,
and American chestnut backcross family were sowed in Fafard 3B
Mix in D40 Deepots in a greenhouse at the USDA Forest Service
Resistance Screening Center in February (Table 2). The pots were
arranged in 20-cell trays in a randomized complete block design.
The blocks consisted of 10 trays placed in flood tables that measured
1.2 by 1.8 by 0.3 m. Once seedlings emerged, they were fertilized
with time-release Osmocote fertilizer and irrigated overhead three
times per week for 13 weeks prior to inoculation. After inoculation,
seedlings were subirrigated three times each week for the duration of
the trial. Subirrigation completely saturated the container mix in each
Deepot and provided conducive conditions for disease development.
Irrigation water was pumped from the flood tables to holding tanks
and treated with bleach (NaOCl) after use to prevent contaminating
the site with P. cinnamomi. Seedlings were inoculated once with
the same isolate of P. cinnamomi (23ss04) used in the NCSU trials,
and the inoculum was prepared as colonized V8-vermiculite, as in the
Clemson University trials. A 5-ml aliquot of V8-vermiculite inocu-
lum was placed in each pot, the inoculum was covered immediately
with sand, and the pots were watered to prevent desiccation of the in-
oculum. Survival was assessed by visual observation of foliage
symptoms and removing seedlings with wilted crowns at 4, 8, 16,
and 24 weeks after inoculation. Seedlings with wilted crowns were
scored as dead, and roots on these seedlings were inspected for
typical symptoms of Phytophthora root rot.
Evaluation of resistance to C. parasitica.In total, 103 Graves
BC
3
F
3
families that were evaluated for resistance to P. cinnamomi
were also evaluated for resistance to C. parasitica at TACF Meadow-
view Research Farms. These 103 families were a subset of 544 fam-
ilies that had been evaluated for resistance to C. parasitica from 2011
to 2016 in separate experiments (Steiner et al. 2017). Thirty BC
3
F
3
seeds per family were planted in the field at Meadowview Research
Farms at 1-by-2.3-m spacing in a randomized complete block design
(2011 to 2013) or resolvable incomplete design (2014 to 2016). Prog-
eny from each family was inoculated in the third growing season with
two isolates of C. parasitica: SG2-3 with low virulence and Ep155
with high virulence. Cankers lengths and subjective canker severity
ratings (1 = no canker expansion, 2 = some expansion beyond initial
lesion, and 3 = large, sunken canker with sporulation) were assessed
6 months after inoculation. Canker severity was estimated from the
sum of the normalized canker length and canker severity rating for
the two isolates of C. parasitica (Steiner et al. 2017).
Data analysis. The original root rot severity scores of 0 to 3 used
in the Clemson University trials were reversed to root health scores of
3 to 0, so that families could be ranked for resistance to P. cinna-
momi. Seedlings scored as 3, 2, and 1 were alive while those scored
as 0 were dead. An analysis of variance to detect differences in sur-
vival among families in each year and at each location was performed
with the R package ‘easyanova’(Arnhold 2013). Tubs were treated
as complete blocks in analysis of survival in trials conducted by
Clemson University and by TACF and the USDA Forest Service,
whereas results from the trials conducted by NCSU were analyzed
as a completely randomized design. Family means were judged sig-
nificantly different if Pvalues from Ftests of family effects were
less than 0.05.
Mixed-model analyses of Phytophthora root rot phenotypes were
performed in ASReml-R (Butler et al. 2009). Phenotypes of back-
cross descendants of the Clapper and Graves trees were analyzed sep-
arately. Root health rating (3-to-0 scale) or survival (0,1) of seedlings
in the Clemson University trials were analyzed using univariate
models that included the population mean and the random effects
of inoculation year, tub, family, and error. Crown health (0 to 100) or
survival (0,1) of seedlings in the NCSU trials were analyzed employ-
ing univariate models that included the population mean and the ran-
dom effects of inoculation year, family, and error. Seedling survival
in the TACF-USDA Forest Service trial was analyzed in a model that
included the population mean and the random effects of block, fam-
ily, and error.
The heritability of Phytophthora root rot phenotypes was esti-
mated for individual BC
3
F
3
trees with the formula:
h2=4s2
familys2
phenotype (1)
where s
2
phenotype
is the sum of the variance components associated
with random effects in the model (i.e., year, family, block, and error).
Table 2. Summary of four annual trials to evaluate seedlings of American, Chinese, and BC
3
F
3
families for resistance to Phytophthora cinnamomi at North
Carolina State University (NCSU) and the USDA Forest Service Resistance Screening Center (RSC) in Asheville, NC based on the percentage of seedlings sur-
viving at the end of each trial
Dates on which seedlings were
a
Duration
(weeks)
b
Number of families
c
Number of seedlings
One-way ANOVA:
P>Fvalues
d
Trial
e
Year Planted Inoculated Evaluated P-I I-E P-E BC Am Ch Total BC Am Ch Total Am vs. Ch
f
BC
3
F
3
g
NCSU 2014 15 May 14 23 Jul 14 12 Nov 14 10 16 26 95 0 0 95 2,259 0 0 2,259 NC <0.001
NCSU 2015 01 Apr 15 29 May 15 18 Sep 15 8 16 24 97 2 2 101 1,892 40 32 1,964 <0.001 <0.001
NCSU 2016 11 May 16 01 Jul 16 21 Oct 16 7 16 23 74 3 1 78 1,482 40 20 1,542 <0.001 <0.001
RSC 2017 15 Feb 17 23 May 17 07 Nov 17 14 24 38 37 1 1 39 1,036 27 39 1,102 <0.001 <0.001
Total …… … …………287 6 4 297
h
6,669 107 91 6,897 ……
a
Abbreviations: July (Jul), November (Nov), April (Apr), September (Sep), October (Oct), and February (Feb) in years 2014 (14) through 2017 (17).
b
Lengths of time in weeks between activities involving seedlings: P = planted, I = inoculated, and E = evaluated.
c
Numbers of different families planted each year. A family is a specific chestnut genotype: Am = open-pollinated American chestnut, Ch = open-pollinated Chi-
nese chestnut, and BC = specific BC
3
F
3
generation of American chestnut backcross hybrids.
d
Results from statistical analyses by one-way analysis of variance (ANOVA): P>Fis the probability of a greater Fstatistic occurring. Analyses were not con-
ducted for some data (NC) because of insufficient replication.
e
NCSU = North Carolina State University and RSC = The American Chestnut Foundation–United States Department of Agriculture Forest Service Resis-
tance Screening Center.
f
Results showing the significant difference in survival between the American and Chinese chestnut families each year based on a single-degree-of-freedom
linear contrast.
g
Results showing the significant difference in survival among only the American chestnut BC
3
F
3
families evaluated each year based on a one-way ANOVA
of a subset of the data from this trial.
h
Note that 16 families were each evaluated in two separate years; therefore, the sum of the number of families evaluated in all years (313) is greater than the
total number of families evaluated (297).
1634 Plant Di sease / Vol. 103 No. 7
This heritability estimate assumes random mating among BC
3
F
2
trees and that BC
3
F
3
progeny were half-siblings (Lynch and Walsh
1998). The family variance component and best linear unbiased pre-
dictions (BLUPs) of Phytophthora root rot phenotypes for each family
were estimated by incorporating the inverse of the pedigree-derived
additive genetic relationship matrix into the model (Butler et al.
2009). The BLUPs were scaled from 0 (American chestnut mean) to
100 (Chinese chestnut mean) with the formula:
Scaled BLUP = 100½1−ðBLUP −mChin ÞmAm −mChinÞ=(2)
where BLUP is in the original scale and m
Chin
and m
Am
are the aver-
age BLUPs for Chinese chestnut and American chestnut families,
respectively.
Bivariate analyses were performed to estimate genetic correlations
between different Phytophthora root rot phenotypes from the same
trees, survival of the same families inoculated with P. cinnamomi
at different screening locations, and between Phytophthora root rot
survival and chestnut blight canker severity. Covariance was esti-
mated among all random effects (i.e., family, tub, and error) for anal-
yses involving different phenotypes from the same trees. For
bivariate analyses of Phytophthora root rot survival at different sites
or between Phytophthora root rot survival and chestnut blight canker
severity, experimental design effects and residuals were assumed to
be uncorrelated while genetic covariance was estimated among fam-
ilies. Genetic correlations were estimated with the formula:
Covtrait1;trait2 ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
Vartrait1
p+ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
Vartrait2
p(3)
where Cov and Var are family covariance and variance components,
respectively. Standard errors for genetic correlations and heritabil-
ities were estimated with the Dmethod (Lynch and Walsh 1998). Ge-
netic correlations were considered significantly different than zero if
the Pvalue from the likelihood ratio test of the correlation was less
than 0.05 (Butler et al. 2009). Figures were prepared using the R
package ‘ggplot2’(Wickham 2009).
Results
Seedling survival within families after inoculation with
P. cinnamomi.Significant differences in seedling survival (Ftest,
P< 0.001) were observed between resistant Chinese chestnut and
susceptible American chestnut controls and among backcross fami-
lies in all years and at all screening locations (Tables 1 and 2). In
all, 11% of American chestnut seedlings inoculated with P. cinna-
momi survived to the end of assessment periods—which, on average,
was 38 weeks at Chestnut Return Farms, 24 weeks at the Resistance
Screening Center, and 16 weeks at NCSU. In contrast, 97% of the
Chinese chestnut seedlings survived over the same time period. In
111 Graves BC
3
F
3
families evaluated in the Clemson University trials,
an average of 32% of the seedlings survived, whereas 75 to 100%
of the seedlings survived in the 5% most resistant families. Similar
rates of survival were observed for seedlings in 156 Graves BC
3
F
3
families evaluated in NCSU trials, where seedling survival averaged
24% overall, and 75 to 95% of the seedlings survived in the 5% most
resistant families. Average survival of seedlings in Clapper BC
3
F
3
families was higher in NCSU trials (99 families, mean survival =
24%) as compared with Clemson University trials (30 families, mean
survival = 7%) (Ftest, P< 0.001).
High rates of survival were observed for control plants in the
NCSU trials that were not inoculated with P. cinnamomi. Survival
was 91% for American chestnuts (20 of 22 seedlings), 100% for Chi-
nese chestnuts (20 of 20 seedlings), and 90% for BC
3
F
3
trees (258 of
287 seedlings). In Clemson University trials, P. cinnamomi was iso-
lated consistently from the roots of diseased plants and container mix
samples collected annually from individual tubs. These results con-
firm that the majority of the mortality of inoculated seedlings could
be attributed to Phytophthora root rot and not to other causes.
Heritability of Phytophthora root rot phenotypes among
BC
3
F
3
families. Phenotypes associated with resistance to P. cinna-
momi (i.e., survival, root health, and crown health) were highly her-
itable (h
2
= 0.85 ± 0.09 to 1.0 ± 0.11) among seedlings in Graves
BC
3
F
3
families over 6 years of evaluation at the three locations
(Tables 3 and 4). For Graves BC
3
F
3
families included in the 2015
and 2016 NCSU trials, the heritability of survival of seedlings in-
creased rapidly following inoculation and approached a maximum
of 1 at 10 to 12 weeks after the first of two inoculations (Fig. 1). In
the TACF-USDA Forest Service trial, where seedlings were inocu-
lated once with P. cinnamomi, heritability of seedling survival in
Graves BC
3
F
3
families increased more gradually to 0.8 ± 0.17 at
24 weeks (Fig. 1). The Phytophthora root rot phenotypes were less
heritable among seedlings in the Clapper BC
3
F
3
families (h
2
=
0.08 ± 0.07 to 0.32 ± 0.08). Survival of seedlings in Clapper
BC
3
F
3
families was more strongly heritable in trials at NCSU
(h
2
= 0.27 ± 0.09) than in Clemson University trials (h
2
= 0.11 ±
0.08) (Tables 3 and 4).
Resistance to P. cinnamomi inherited across backcross gener-
ations in Graves families. In Clemson University trials, seedlings in
families from three backcross generations (BC
3
,BC
4
, and BC
3
F
3
)
were evaluated, which enabled an intergenerational comparison of
P. cinnamomi resistance. Among backcross descendants of the
Graves BC
1
tree, variation in survival of seedlings in BC
3
F
3
and
BC
3
families was positively skewed and spanned the range of seed-
ling survival in resistant Chinese chestnut and susceptible American
chestnut control seedlings (Fig. 2A). By contrast, when root health
ratings were used as the phenotype to assess P. cinnamomi resistance,
the most resistant Graves BC
3
F
3
families had resistance intermediate
between Chinese chestnut and American chestnut (Fig. 2B). Al-
though resistance of seedlings estimated from root health was less
than that of seedlings estimated from survival, root health and seed-
ling survival were strongly genetically correlated among Graves
BC
3
F
3
families (r
genetic
= 0.98 ± 0.01, likelihood ratio test P< 0.001).
Average survival of seedlings in BC
3
F
3
families was greater than
average survival of seedlings in BC
3
families, BC
4
families, and
American chestnut families (Tukey test, P< 0.05). Survival BLUPs
for multiple BC
3
F
3
families were regressed on those of single BC
4
Table 3. Narrow-sense heritability (h
2
) of survival and root health ratings at 35 to 42 weeks after inoculation with Phytophthora cinnamomi among seedlings
of American chestnut BC
3
F
3
families that descended from the Clapper (C) or Graves (G) BC
1
trees in six Clemson University trials (2012 to 2017)
Survival (h
2
6SE)
a
Root health (h
2
6SE)
b
Number of
families
Median number seedlings
per family (maximum,
minimum)
Year C G C G C G C G
2012 …0.64 ± 0.19 …0.75 ± 0.20 …39 …14 (5,17)
2013 0.02 ± 0.12 1.12 ± 0.26 0 1.18 ± 0.25 19 35 13 (7,17) 14 (5,22)
2014 …0.91 ± 0.28 …0.72 ± 0.25 …19 …26 (6,30)
2015 0.14 ± 0.10 0.87 ± 0.38 0.12 ± 0.09 0.68 ± 0.32 13 9 34 (23,54) 35 (26,37)
2016 …0.76 ± 0.22 …0.84 ± 0.23 …21 …41 (34,47)
2017 …0.88 ± 0.26 …0.62 ± 0.22 …17 …27 (30,12)
All years 0.11 ± 0.08 0.90 ± 0.11 0.08 ± 0.07 0.87 ± 0.11 32 122 14 (7,54) 16 (5,54)
a
Survival of seedlings 20 to 29 weeks after inoculation. SE = standard error.
b
Root health ratings 20 to 29 weeks after inoculation: 3 = no lesions on roots, 2 = lesions on some lateral roots; 1 = lesions on the tap root or severe root rot
on lateral roots; and 0 = severe root rot, plant dead.
Plant Disease / July 2019 1635
families that descended from a common BC
3
relative, where the BC
3
tree was a grandparent of the BC
3
F
3
trees and a parent of the BC
4
trees (Fig. 3). The slope of the regressionline was positive—indicating
that inheritance of resistance to P. cinnamomi from the BC
3
rela-
tive was detectable after two generations of intercrossing by open
pollination to generate the BC
3
F
3
trees. The intercept of the regres-
sion line was greater than zero, indicating that BC
3
F
3
families
were, on average, more resistant than their BC
4
relatives. There
was a large variation in seedling survival BLUPs among BC
3
F
3
families that descended from a common BC
3
grandparent.
Minimal P. cinnamomi resistance among Clapper families in
Clemson University trials. Survival BLUPs for seedlings in Clap-
per BC
3
,BC
4
,andBC
3
F
3
families were, on average, not sig-
nificantly different from those of susceptible American chestnut
families (Fig. 4). Root health BLUPs for Clapper BC
3
families
were, on average, marginally greater than those for American chest-
nut (Tukey test, P= 0.027).
Resistance to P. cinnamomi in NCSU and TACF-USDA Forest
Service trials. Similar to results observed in Clemson University tri-
als, genetic variation in survival of seedlings in Graves BC
3
F
3
fam-
ilies at 16 weeks spanned the range of variation in survival of
seedlings between Chinese chestnut and American chestnut controls
in NCSU and TACF-USDA Forest Service trials (Fig. 5A). Genetic
variation in crown health was reduced relative to variation in seedling
survival (Fig. 5B) but family rankings of seedling survival and crown
health were strongly correlated (r
genetic
= 0.84 ± 0.03). Seedlings in
Graves BC
3
F
3
families had significantly higher average BLUPs for
survival and crown health than those in Clapper BC
3
F
3
families
and American chestnut controls (Tukey test, P< 0.05). Seedlings
in Clapper BC
3
F
3
families with the greatest resistance in NCSU trials
had BLUPs for survival and crown health that were intermediate be-
tween the American chestnut and Chinese chestnut controls (Fig. 5).
This result contrasts with those from Clemson University trials,
where no resistance to P. cinnamomi was detected among seedlings
in Clapper BC
3
F
3
families.
Correlations in survival of seedlings between trials. A strong
genetic correlation was observed between survival of seedlings in tri-
als conducted by Clemson University and those in trials conducted
by NCSU (r
genetic
= 0.94 ± 0.03, likelihood ratio test P< 0.001)
for 53 BC
3
F
3
families (9 Clapper and 44 Graves) screened by both
groups. To test whether mean seedling survival among all families
differed between Clemson University and NCSU trials, a regression
was performed between the BLUPs for survival of seedling in BC
3
F
3
families in the NCSU and Clemson University trials. The slope of the
regression line was greater than one (ttest, P< 0.001), indicating that
BLUP values for seedling survival in NCSU trials were greater, on
average, than those for seedlings in Clemson University trials (Fig.
6A). For seedlings in 17 Graves BC
3
F
3
families that were inoculated
with P. cinnamomi in trials conducted by Clemson University and
TACF-USDA Forest Service in 2017, the genetic correlation in seed-
ling survival between the two trial locations was 0.81 ± 0.13 (likeli-
hood ratio test P= 0.002), and the slope of the regression line was not
significantly different than 1 (ttest, P>0.05) (Fig. 6B).
Resistance to C. parasitica and P. cinnamomi was
not correlated. The genetic correlation between chestnut blight can-
ker severity and seedling survival after inoculation with P. cinna-
momi was not significantly different than zero (r
genetic
=−0.01 ±
0.13, likelihood ratio test P= 0.1) among 103 Graves BC
3
F
3
famil-
ies (Fig. 7).
Discussion
Since 1983, TACF has pursued backcross breeding to introduce
Chinese chestnut resistance to the chestnut blight fungus (C. para-
sitica) into hybrids that are expected to have inherited 94% of their
genome from American chestnut parents (Burnham et al. 1986; Dis-
kin et al. 2006; Hebard 2006; Steiner et al. 2017). This program orig-
inally did not include evaluating backcross families for resistance to
P. cinnamomi, which causes Phytophthora root rot. Fortunately, Chi-
nese chestnut is also resistant to P. cinnamomi (Crandall et al. 1945;
Jeffers et al. 2009, 2012). Therefore, personnel from Clemson Uni-
versity, Chestnut Return Farms, and TACF began evaluating seed-
ling progeny from a variety of BC
2
and BC
3
American chestnut
backcross trees that had been selected for resistance to C. parasitica
to determine whether any of these backcross trees had also inherited
resistance to P. cinnamomi. Significant differences in seedling sur-
vival among backcross families inoculated with P. cinnamomi were
observed each year in trials conducted from 2004 to 2011. Based on
these results, it was determined that some BC
2
and BC
3
American
chestnut backcross parents inherited alleles for resistance to P. cinna-
momi. Consequently, open-pollinated BC
3
F
3
progeny of BC
3
F
2
trees
that descended from the BC
2
and BC
3
trees evaluated in earlier trials
were evaluated for resistance to P. cinnamomi from 2012 to 2017 in
trials conducted by personnel at Clemson University, NCSU, and
TACF (with the USDA Forest Service) to determine which BC
3
F
2
parents had inherited resistance to P. cinnamomi.
Over the 14-year duration of this project, 18 annual trials were
conducted at three locations in two states. Significant differences
in seedling survival at 16 to 29 weeks after inoculation with P. cin-
namomi occurred consistently among American chestnut backcross
families in all 18 trials. In these trials, seedling survival was deter-
mined at the end of one growing season and was based on a combina-
tion of three root rot severity scores that ranged from no symptoms to
severe root rot symptoms. We know from years of experience (J. B.
James and S. N. Jeffers, unpublished data) that many of the seedlings
with severe root rot symptoms do not survive a second growing sea-
son after transplanting into pots or in the field; however, survival af-
ter one growing season has proven to be the best parameter for
identifying differences in resistance to P. cinnamomi among seed-
lings in backcross families under the experimental conditions used
in these trials.
Differences in survival and average root heath ratings among
backcross families were highly heritable, suggesting that the root
health and survival phenotypes of individual trees accurately reflects
their underlying genetic resistance to P. cinnamomi. Correlated sur-
vival within backcross families that were inoculated with different
isolates of P. cinnamomi at multiple screening locations in different
Table 4. Narrow-sense heritability (h
2
) of survival and crown health ratings at 16 weeks after inoculation with Phytophthora cinnamomi among seedlings
of American chestnut BC
3
F
3
families that descended from the Clapper (C) or Graves (G) BC
1
trees in three North Carolina State University trials (2014
to 2016) and one trial conducted by The American Chestnut Foundation and the United States Department of Agriculture Forest Service (2017)
Survival (h
2
6SE)
a
Crown health (h
2
6SE)
b
Number
families
Median number seedlings
per family (maximum,
minimum)
Year C G C G C G C G
2014 0.33 ± 0.08 0.66 ± 0.24 0.38 ± 0.08 0.72 ± 0.25 78 17 25 (10,31) 23 (18,28)
2015 0.20 ± 0.11 0.92 ± 0.15 0.28 ± 0.13 1.0 ± 0.16 28 69 20 (14,21) 20 (10,22)
2016 …1.13 ± 0.16 …1.13 ± 0.16 …75 …20 (18,22)
2017 …0.74 ± 0.17 ………37 …30 (7,40)
All years 0.27 ± 0.09 0.85 ± 0.09 0.32 ± 0.08 1.0 ± 0.11 99 189 24 (10,49) 20 (7,57)
a
Survival of stems 16 weeks after inoculation. SE = standard error.
b
Crown health is percentage of crown that had wilted 16 weeks after inoculation.
1636 Plant Di sease / Vol. 103 No. 7
years further confirmed that our trials accurately measured the ge-
netic component of P. cinnamomi resistance despite differences in
experimental conditions.
A number of observations supported the hypothesis that backcross
descendants of the Graves BC
1
tree inherited major-effect alleles for
P. cinnamomi resistance from Mahogany, the C. mollissima grand-
parent of Graves. Chinese chestnut control seedlings survived at
a high level (97%) whereas American chestnut controls survived at
a low level (11%), which confirmed that P. cinnamomi resistance
in backcross trees was conferred by alleles inherited from C. mollis-
sima. The Graves BC
3
and BC
3
F
3
families that were most resistant to
P. cinnamomi had seedling survival percentages that approached
those of resistant Chinese chestnut seedlings. Third-backcross trees
are expected to only have inherited 1/16th of their genome from
C. mollissima (Steiner et al. 2017); therefore, major-effect alleles for
P. cinnamomi resistance were expected to be located on small seg-
ments of backcross genomes inherited from C. mollissima. It is re-
markable that alleles for resistance to P. cinnamomi that were
inherited from C. mollissima had not been bred out through back-
crossing to C. dentata because TACF did not select for resistance
to P. cinnamomi.
We showed in this study that chestnut blight canker severity and
Phytophthora root rot survival were not genetically correlated among
Graves BC
3
F
3
families; therefore, major-effect alleles for P. cinna-
momi resistance are expected to be unlinked with alleles for blight
resistance. Quantitative trait locus (QTL) mapping supports the
hypothesis that alleles that confer resistance to C. parasitica and
P. cinnamomi are not linked. Among C. mollissima ×C. dentata F
2
progeny of the Mahogany C. mollissima tree, QTLs for blight resis-
tance were detected on linkage groups B, F, and G (Kubisiak et al.
1997, 2013). Among BC
1
descendants of another C. mollissima tree,
a QTL for resistance to P. cinnamomi was first detected on link-
age group E by Kubisiak (2010). By reanalyzing Kubisiak’sdata,
Olukolu et al. (2012) found an additional locus on linkage group E
plus a locus on linkage group A. Using much larger BC
1
fami-
lies from two C. mollissima trees (Mahogany and the one used by
Kubisiak), T. Zhebentyayeva (personal communication) confirmed
the two resistance loci on linkage group E and found an additional locus
for resistance on linkage group K. QTLs for resistance to P. cinnamomi
also were discovered on linkage groups E and K among C. sativa ×
C. crenata F
1
hybrids (Santos et al. 2017). These results suggest that
Japanese chestnut (C. crenata) and Chinese chestnut share common
loci that confer resistance to P. cinnamomi.
Our results suggest that alleles for P. cinnamomi resistance have
dominant effects. Average and maximum survival and root heath
scores among seedlings in BC
3
families, which are at best heterozy-
gous for alleles that confer resistance to P. cinnamomi, had margin-
ally lower survival and root health scores as compared with seedlings
in BC
3
F
3
families. The BC
3
F
3
progenies were generated from two
generations of intercrossing among BC
3
and BC
3
F
2
trees. Thus,
BC
3
F
3
trees potentially inherited P. cinnamomi resistance alleles
from both parents in a homozygous state. The observation that sur-
vival of the most resistant heterozygous BC
3
families was similar
to that of the most resistant and potentially homozygous BC
3
F
3
fam-
ilies suggests that random intercrossing among BC
3
and BC
3
F
2
trees
selected for resistance to C. parasitica rarely generates progeny that
are homozygous for P. cinnamomi resistance alleles at one or more
loci. Regardless, the greater average resistance and wide segregation
for P. cinnamomi resistance among seedlings in BC
3
F
3
families as
compared with BC
4
progeny that descended from common BC
3
trees
suggests that intercrossing backcross trees selected for resistance to
P. cinnamomi has the potential to enhance P. cinnamomi resistance
in the subsequent generation.
The observation that seedlings in the most resistant Graves BC
3
F
3
families had root health that was intermediate between Chinese and
American chestnut suggests that Graves BC
3
F
2
and BC
3
F
3
progeny
did not inherit the full complement of P. cinnamomi resistance alleles
from the C. mollissima Mahogany tree in a homozygous state. Resis-
tance to P. cinnamomi may be enhanced to levels similar to C. mol-
lissima trees through controlled pollinations between the most
resistant backcross trees and additional selection for P. cinnamomi
resistance among their progenies.
Our results were inconclusive as to whether descendants of Clap-
per inherited little to no resistance to P. cinnamomi, as suggested by
Clemson University trials, or intermediate levels of resistance, as
suggested by NCSU trials. Backcross descendants of Clapper may
have inherited fewer or smaller-effect resistance alleles as compared
with descendants of Graves. Alternatively, BC
3
F
2
progeny of Clap-
per BC
3
mothers may have inherited resistance to P. cinnamomi from
neighboring Graves BC
3
trees, more distant C. dentata backcross
Fig. 1. Heritability (h
2
±1 standard error) of survival of chestnut seedlings in Graves
BC
3
F
3
families over time after inoculation with Phytophthora cinnamomi. Seedlings
were inoculated twice in the North Carolina State University trials in 2015 (circles)
and 2016 (triangles) at 0 and 8 weeks, whereas seedlings were inoculated once in
The American Chestnut Foundation–United States Department of Agriculture Forest
Service trial in 2017 (squares).
Fig. 2. Genetic variation in A, survival and B, root health among backcross
descendants of the Graves chestnut tree 35 to 42 weeks after inoculation with
Phytophthora cinnamomi in Clemson University trials. Box-plots were created from
variation in best linear unbiased predictions (BLUPs) of survival and root health
among 18 BC
3
families, 46 BC
4
families, 122 BC
3
F
3
families, 13 Chinese chestnut
families, and 13 American chestnut families. The BLUPs were scaled from
0 (American chestnut mean) to 100 (Chinese chestnut mean).
Plant Disease / July 2019 1637
descendants of additional C. mollissima trees, or pure C. mollissima
trees that were planted on the same farm as the Clapper BC
3
trees.
Large numbers of BC
3
F
2
progeny were needed to generate a subset
of progeny that were homozygous for multiple blight resistance
alleles; therefore, open pollination among BC
3
trees was used to gen-
erate BC
3
F
2
trees (Steiner et al. 2017). Individual Clapper BC
3
selec-
tions were planted within 2.1 to 475 m (average distance = 154 m) of
other Clapper BC
3
selections. Maximum pollination occurs among
trees that are less than 30 m apart and the likelihood of pollination
decreases exponentially as distance between trees increases (Rutter
1990). Consequently, it is likely that most of the Clapper BC
3
F
2
trees were the progeny of Clapper BC
3
×BC
3
crosses. However,
with open pollination, it is possible that a subset of the trees referred
to as Clapper BC
3
F
2
s were the progeny of Clapper BC
3
mothers
pollinated by trees other than Clapper BC
3
fathers. The Clapper
BC
3
selections were planted 15 to 548 m (average distance =
234 m) from Graves BC
3
selections and a minimum of 140 m from
Chinese chestnuts on TACF Glenn C. Price Farm in Meadowview,
VA. These pollen sources may have contributed alleles for resis-
tance to P. cinnamomi to Clapper BC
3
F
2
and BC
3
F
3
seedlings. A
larger number of Clapper BC
3
F
3
families were screened in NCSU
trials (99 families) as compared with Clemson trials (32 families).
Therefore, it is more likely that NCSU trials detected partial resis-
tance in Clapper or pollen contamination. Genotyping of Clapper
and Graves BC
3
and BC
3
F
2
trees as well as pure Chinese and Amer-
ican chestnut trees is ongoing (J. Westbrook and J. Holliday, un-
published). With these data, we will be able to determine whether
Clapper and Graves BC
3
trees intercrossed to generate Clapper ×
Graves BC
3
F
2
trees and whether any of the BC
3
F
2
trees were the
progeny of outcrosses to Chinese chestnuts.
Selection for resistance to P. cinnamomi from within a diverse
backcross breeding population is the first step toward breeding to
combine resistance to C. parasitica and P. cinnamomi. Additional
evaluation for resistance to P. cinnamomi will focus primarily on
descendants of the Graves tree due to the greater potential for gains
in resistance as compared with descendants of the Clapper tree.
Within TACF’s national breeding program, in total, 181 backcross
lines that descended from the Graves tree have been generated by
breeding BC
2
and BC
3
selections from the Meadowview farm with
wild-type American chestnut trees with locations ranging from
Maine to Georgia (Westbrook 2018a). To date, only BC
3
F
3
trees
from the Meadowview farm breeding program, which are descend-
ants of American chestnuts from southwest Virginia, have been
screened for resistance to P. cinnamomi. Starting in 2018, TACF be-
gan to increase the genetic diversity of backcross trees selected for
resistance to P. cinnamomi by screening Graves BC
3
F
2
and BC
4
F
2
descendants of American chestnuts from Maine, Vermont, New
Hampshire, Massachusetts, Pennsylvania, northern Virginia, Ken-
tucky, North Carolina, South Carolina, and Georgia. Forty progeny
per BC
3
or BC
4
parent were inoculated with P. cinnamomi. Progeny
that survive inoculation with P. cinnamomi will be planted at orchard
sites where P. cinnamomi is already present so that seedlings will
Fig. 3. In trials conducted by Clemson University, survival of chestnut seedlings after
inoculation with Phytophthora cinnamomi in multiple Graves BC
3
F
3
families versus
single Graves BC
4
families that shared a common BC
3
ancestor. The BC
3
ancestor
is a grandparent of the BC
3
F
3
trees and a parent of the BC
4
trees. Best linear
unbiased predictions (BLUPs) of survival were scaled from 0 (American chestnut
mean) to 100 (Chinese chestnut mean). The regression line is the least squares
regression of BC
3
F
3
versus BC
4
survival in BLUPs.
Fig. 5. Genetic variation in A, survival and B, crown health among chestnut seedlings
in BC
3
F
3
families that descended from the Clapper and Graves trees. Box-plots were
created from variation in best linear unbiased predictions (BLUPs) of survival and
crown health 16 weeks after inoculation with P. cinnamomi; 99 Clapper BC
3
F
3
families, 189 Graves BC
3
F
3
families, 4 American chestnut families, and 4 Chinese
chestnut families included in three North Carolina State University trials and one
The American Chestnut Foundation–United States Department of Agriculture Forest
Service trial. The BLUPs were scaled from 0 (American chestnut mean) to 100
(Chinese chestnut mean).
Fig. 4. Genetic variation in A, survival and B, root health among backcross
descendants from the Clapper chestnut tree at 35 to 42 weeks after inoculation
with Phytophthora cinnamomi in Clemson University trials. Box-plots were created
from variation in best linear unbiased predictions (BLUPs) of survival and root
health among 10 BC
3
families, 20 BC
4
families, 32 BC
3
F
3
families, 13 Chinese
chestnut families, and 13 American chestnut families. The BLUPs were scaled from
0 (American chestnut mean) to 100 (Chinese chestnut mean).
1638 Plant Di sease / Vol. 103 No. 7
also be exposed to the pathogen in natural settings. The final objec-
tive is to obtain at least one flowering BC
3
F
2
or BC
4
F
2
seedling
from each BC
3
or BC
4
mother tree with improved resistance to
P. cinnamomi.
Considering that resistance to C. parasitica and resistance to
P. cinnamomi are not correlated, additional breeding and selection will
be required to combine the resistances to these two pathogens. It is
not feasible to select trees in the current BC
3
F
2
generation that
inherited high levels of resistance to both C. parasitica and P. cinna-
momi. In screening for resistance to C. parasitica in the BC
3
F
2
gen-
eration, TACF selects 1 of 150 trees (Steiner et al. 2017). Even with
this high selection intensity, the selected trees are predicted to have
C. parasitica resistance that is intermediate between Chinese chest-
nut and American chestnut. Additional selection for P. cinnamomi
resistance would be required, leaving too few selections for further
intercrossing.
One potential breeding strategy to combine resistance to both
pathogens is to cross BC
3
F
2
(or BC
4
F
2
) trees selected for resistance
to P. cinnamomi with BC
3
F
2
trees that have been selected for resis-
tance to C. parasitica. The BC
3
F
3
progeny would be evaluated for
resistance to P. cinnamomi first; then, survivors would be evaluated
for resistance to C. parasitica. Another generation of intercrossing
and selection for resistance to both pathogens would be required to
obtain selections that are homozygous for alleles that confer resis-
tance to both pathogens. A challenge to this strategy is that it requires
large numbers of progeny from each cross due to multigenic control
of the resistances to both pathogens, and it is difficult to obtain large
numbers of seed through controlled pollinations. We estimate that at
least 200 BC
3
F
3
progeny per cross would need to be evaluated to ob-
tain one progeny with high levels of resistance to both pathogens.
A second breeding strategy is to outcross descendants of transgenic
C. parasitica-resistant American chestnut trees with P. cinnamomi-
resistant backcross trees. Collaborators at the State University of
New York in Syracuse have developed a transgenic American
chestnut tree that is resistant to C. parasitica by overexpressing the
oxalate oxidase (OxO) gene from wheat (Zhang et al. 2013). When
the transgenic trees are outcrossed with wild-type American chestnut
trees, approximately 50% of the progeny that inherited the OxO gene
have been demonstrated to have high levels of resistance after artifi-
cial inoculation with C. parasitica (Newhouse et al. 2014). By sim-
plifying the genetics of C. parasitica resistance to a single transgene
with a dominant effect, as few as 40 to 80 progeny per cross between
a transgenic tree and a P. cinnamomi-resistant backcross tree may be
sufficient to generate selections that inherited the OxO gene and
major-effect P. cinnamomi resistance alleles. The inheritance of the
OxO gene can be detected without inoculation with C. parasitica
by using an enzymatic assay to detect OxO activity in plant tissue
(Zhang et al. 2013). Progeny that inherit the OxO gene will then
be screened for resistance to P. cinnamomi. Surviving plants will
be planted at orchard locations where P. cinnamomi already is pres-
ent in the soil. Selections will be intercrossed to generate progeny that
inherited both the OxO gene and P. cinnamomi resistance alleles in
a homozygous state.
Currently, one transgenic insertion of the OxO gene into one
American chestnut tree from New York is being considered for
deregulated status by the U.S. federal government (W. Powell, per-
sonal communication). If regulatory approval is obtained, three gen-
erations of outcrossing the transgenic founder clone to wild-type
American chestnuts and backcross trees will be required to expand
the effective population size to >500 (Westbrook 2018b). Resistance
to C. parasitica and P. cinnamomi will be combined in the final gen-
eration of outcrossing so that selection for OxO activity and P. cin-
namomi resistance will be required only after the final outcross and
first intercross generations (Westbrook 2018a). It is estimated that
it will take 15 to 35 years from the time of federal regulatory approval
to release transgenic trees to complete five additional generations of
breeding (i.e., three outcross generations and two intercross genera-
tions) to combine resistance to these two pathogens within a geneti-
cally diverse population that is homozygous for resistance to both
pathogens (Westbrook 2018b). The intermediate objective of com-
bining resistance to these pathogens in a less genetically diverse pop-
ulation may be accomplished within 10 years of the release of
transgenic trees. This objective will require one generation of breed-
ing between transgenic trees resistant to C. parasitica and backcross
trees resistant to P. cinnamomi, selection among the progeny, and
intercrossing among the selections to generate large quantities of
seed.
In conclusion, restoration of American chestnut trees depends on
combining resistance to C. parasitica and P. cinnamomi, two patho-
gens that have effectively eliminated this tree species from forests in
the eastern United States. Resistance must be combined within breed-
ing populations that are sufficiently diverse so that the species may
continue to evolve within its native range. Our results demonstrate
that American chestnut backcross descendants of one of TACF’s
sources of resistance to C. parasitica also have inherited major-effect
Fig. 6. Genetic correlation of chestnut seedling survival within BC
3
F
3
families
inoculated with Phytophthora cinnamomi at different geographic locations: A, 44
Graves families (triangles) and 9 Clapper families (circles) were evaluated in trials
conducted simultaneously by Clemson University and North Carolina State
University (NCSU); B, 17 Graves families were evaluated in trials conducted
simultaneously by Clemson University and The American Chestnut Foundation–United
States Department of Agriculture Forest Service Resistance Screening Center
(RSC). The least squares regression line ±95% confidence interval (solid line) is
compared with the 1:1 line (dashed line).
Fig. 7. Genetic correlation between survival after inoculation with Phytophthora
cinnamomi and chestnut blight canker severity (best linear unbiased predictions
[BLUP] values) among chestnut seedlings from 103 Graves BC
3
F
3
families.
Plant Disease / July 2019 1639
alleles for resistance to P. cinnamomi. Future selection for P. cinna-
momi resistance in backcross populations and breeding these trees
with C. parasitica-resistant backcross and transgenic trees will be
used to produce trees with resistance to both pathogens.
Acknowledgments
We thank the members and volunteers of and the donors to The American
Chestnut Foundation who have provided long-term financial support, moral sup-
port, and labor for the mission to restore the American chestnut to its native range;
the many people who assisted with the Clemson University trials that were con-
ducted in collaboration with Chestnut Return Farms over 14 years, including
R. Baker, S. Barilovitz, S. Barilovitz, Jr., P. Bowers, R. Cain, E. R. Camacho,
C. Colburn, C. Collier, D. Dlugos, D. Drechsler, L. Georgi, A. Gitto, B. Glenn,
M. Greene, B. Harrelson, F. Hebard, T. Hodge, J. Hodges, J. Hwang, J. Jeffers,
M. Jeffers, C. Kennedy, P. Laurie, L. Luszcz, F. McLaughlin, I. Meadows, S.
Oak, B. Paris, J. Payne, T. Perkins, E. Perry, C. Santos, S. Schreier, E. Schwartzman,
C. Scott, J. Sevic, S. Sharpe, J. Vissage, Y. Wamishe, T. Watson, W. White, T.
Zhebentyayeva, and D. Zwart; R. Blaedow, K. Frick, B. Jarrett, B. Minton, E.
Schwartzman, and B. Williams at the USDA Forest Service Resistance Screening
Center, who assisted with Phytophthora root rot evaluation; A. M. Braham at
NCSU for supervising and conducting the inoculation procedure; and T. Allison,
M. Escanfarla, W. Kohlway, Y. Kurt, J. Pearson, R. Porter, and C. Williams for
assisting with inoculum preparation, inoculation, and phenotypic assessments
Literature Cited
Anagnostakis, S. L. 2001. The effect of multiple importations of pests and
pathogens on a native tree. Biol. Invas. 3:245-254.
Anagnostakis, S. L. 2007. The chestnut plantation at Sleeping Giant: The lega cy of
Arthur Harmount Graves. J. Am. Chestnut Found. 21:34-39.
Anagnostakis, S. L. 2012. Chestnut breeding in the United States for disease and
insect resistance. Plant Dis. 96:1392-1403.
Arnhold, E. 2013. Pacote em ambiente R para an´
alise de variˆ
ancia e an´
alises
complementares [Package in the R environment for analysis of variance and
complementary analyses]. Braz. J. Vet. Res. Anim. Sci. 50:488-492.
Balci, Y., Balci, S., Eggers, J., MacDonald, W. L., Juzwik, J., Long, R. P., and
Gottschalk, K. W. 2007. Phytophthora spp. associated with forest soils in
eastern and north-central US oak ecosystems. Plant Dis. 91:705-710.
Benson, D. M. 1982. Cold inactivation of Phytophthora cinnamomi.
Phytopathology 72:560-563.
Burgess, T. I., Scott, J. K., Mcdougall, K. L., Stukely, M. J. C., Crane, C., Dunstan,
W. A., Brigg, F., Andjic, V., White, D., Rudman, T., Arentz, F., Ota, N., and
Hardy, G. E. 2017. Current and projected global distribution of Phytophthora
cinnamomi, one of the world’s worst plant pathogens. Glob. Change Biol.
23:1661-1674.
Burnham, C. R., Rutter, P. A., and French, D. W. 1986. Breeding blight-resistant
chestnuts. Plant Breed. Rev. 4:346-397.
Butler, D. G., Cullis, B. R., Gilmour, A. R., and Gogel, B. J. 2009. ASReml-R
Reference Manual: Mixed Models for S Language Environments. The State
of Queensland Department of Primary Industries and Fisheries, Brisbane,
QLD, Australia.
Clapper, R. B. 1952. Relative blight resistance of some chestnut species and
hybrids. J. For. 50:453-455.
Crandall, B. S., and Gravatt, G. F. 1967. The distribution of Phytophthora
cinnamomi. Part II—Geographic distribution. Ceiba 13:57-78.
Crandall, B. S., Gravatt, G. F., and Ryan, M. M. 1945. Root disease of Castanea
species and some coniferous broadleaf nursery stocks, caused by Phytophthora
cinnamomi. Phytopathology 35:162-180.
Diskin, M., Steiner, K. C., and Hebard, F. V. 2006. Recovery of American chestnut
characteristics following hybridization and backcross breeding to restore blight-
ravaged Castanea dentata. For. Ecol. Manage. 223:439-447.
Ferguson, A. J., and Jeffers, S. N. 1999. Detecting multiple species of
Phytophthora in container mixes from ornamental crop nurseries. Plant Dis.
83:1129-1136.
Frampton, J., Isik, F., and Benson, D. M. 2013. Genetic variation in resistance to
Phytoph thora cin namomi in seedlings of two Turkish Abies species. Tree Genet.
Genomes 9:53-63.
Freinkel, S. 2007. American Chestnut. The Life, Death, and Rebirth of a Perfect
Tree. University of California Press, Berkeley, CA, U.S.A.
Hebard, F. 2006. The backcross breeding program of The American Chestnut
Foundation. Pages 61-77 in: Restoration of American Chestnut to Forest
Lands—Proc. Conf. Workshop, The North Carolina Arboretum. K. C. Steiner
and J. E. Carlson, eds. Natural Resources Report NPS/NCR/CUE/NRR–2006/
001. National Park Service, Washington, DC.
Holmes, K. A., and Benson, D. M. 1994. Evaluation of Phytophthora parasitica
var. nicotianae for biocontrol of Phytophthora parasitica on Catharanthus
roseus. Plant Dis. 78:193-199.
Jacobs, D. F. 2007. Toward development of silvical strategies for forest restoration
of American chestnut (Castanea dentata) using blight-resistant hybrids. Biol.
Conserv. 137:497-506.
James, J. B. 2011a. Phytophthora: The stealthy killer. J. Am. Chestnut Found. 25:
9-11.
James, J. B. 2011b. Phytophthora: The stealthy killer Part 2. J. Am. Chestnut
Found. 25:14-17.
Jeffers, S. N. 2015. Protocol 07-04.1: PARP(H)-V8A. In: Laboratory Protocols for
Phytophthora species. K. Ivors, ed. American Phytopathological Society,
St. Paul, MN, U.S.A.
Jeffers, S. N., James, J. B., and Sisco, P. H. 2009. Screening for resistance to
Phytophthora cinnamomi in hybrid seedlings of American chestnut. Pages
188-194 in: Proc. Fourth Meet. Int. Union For. Res. Organ. (IUFRO)
Working Party S07.02.09: Phytophthoras in Forests & Natural Ecosystems. E. M.
Goheen and S. J., Frankel, tech. coords. Gen. Tech. Rep. PSW-GTR-221. U.S.
Dep. Agric. For. Serv., Pacific Southwest Research Station, Albany, CA, U.S.A.
Jeffers, S. N., Meadows, I. M., James, J. B., and Sisco, P. H. 2012. Resistance to
Phytophthora cinnamomi among seedlings from backcross families of hybrid
American chestnut. Pages 194-195 in: Proc. Fourth Int. Workshop Genet.
Host–Parasite Interactions in Forestry: Disease and Insect Resistance in
Forest Trees. R. A. Sniezko A. D. Yanchuk, J. T. Kliejunas, K. M. Palmieri,
J. M. Alexander, and S. J. Frankel, tech. coords. Gen. Tech. Rep. PSW-GTR-
240. U.S. Dep. Agric. For. Serv., Pacific Southwest Research Station,
Albany, CA, U.S.A.
Jung, T., Jung, M. H., Scanu, B., Seress, D., Kov´
acs, G. M., Maia, C., P´
erez-Sierra,
A., Chang, T. T., Chandelier, A., Heungens, K., van Poucke, K., Abad-Campos,
P., L´
eon, M., Cacciola, S. O., and Bakonyi, J. 2017. Six new Phytophthora
species from ITS Clade 7a including two sexually functional heterothallic
hybrid species detected in natural ecosystems in Taiwan. Persoonia 38:100-135.
Kubisiak, T. 2010. NE-1333 Technical Committee Meeting Minutes. https://
ecosystems.psu.edu/research/chestnut/meetings/crees-ne-projects/minutes-pdfs/
2010-research-meeting-minutes
Kubisiak, T., Hebard, F., Nelson, C., Jhang, J., Bernatzky, R., Huang, H.,
Anagnostakis, S., and Doudrick, R. 1997. Molecular mapping of resistance to
blight in an interspecific cross in the genus Castanea. Phytopathology 87:
751-759.
Kubisiak, T. L., Nelson, C. D., Staton, M. E., Zhebentyayeva, T., Smith, C.,
Olukolu, B. A., Fang, G.-C., Hebard, F. V., Anagnostakis, S., Wheeler, N.,
Sisco, P. H., Abbott, A. G., and Sederoff, R. R. 2013. A transcriptome-based
genetic map of Chinese chestnut (Castanea mollissima) and identification of
regions of segmental homology with peach (Prunus persica). Tree Genet.
Genomes 9:557-571.
Laurie, P. 2014. The once and future perfect tree. S. C. Wildl. 10:4-9.
Lynch, M., and Walsh, B. 1998. Genetics and Analysis of Quantitative Traits.
Sinauer Associates, Sunderland, MA, U.S.A.
Meadows, I. M., and Jeffers, S. N. 2011. Distribution and recovery of
Phytophthora cinnamomi in soils of mixed hardwood-pine forests of the
south-eastern USA. N. Z. J. For. Sci. 41S:S39-S47.
Meadows, I. M., Zwart, D. C., Jeffers, S. N., Waldrop, T. A., and Bridges, W. C.
2011. Effects of fuel reduction treatments on incidence of Phytophthora species
in soil of a southern Appalachian Mountain forest. Plant Dis. 95:811-820.
Newhouse, A. E., McGuigan, L. D., Baier, K. A., Valletta, K. E., Rottmann, W. H.,
Tschaplinski, T. J., Maynard, C. A., and Powell, W. A. 2014. Transgenic
American chestnuts show enhanced blight resistance and transmit the trait to
T
1
progeny. Plant Sci. 228:88-97.
Olukolu, B. A., Nelson, C. D., and Abbott, A. G. 2012. Mapping resistance to
Phytophthora cinnamomi in chestnut (Castanea sp.). Page 177 in: Proc.
Fourth Int. Workshop Genet. Host–Parasite Interactions in Forestry: Disease
and Insect Resistance in Forest Trees. R. A. Sniezko, A. D. Yanchuk, J. T.
Kliejunas, K. M. Palmieri, J. M. Alexander, and S. J. Frankel, tech. coords.
Gen. Tech. Rep. PSW-GTR-240. U.S. Dep. Agric. For. Serv., Pacific
Southwest Research Station, Albany, CA, U.S.A.
Rhoades, C. C., Brosi, S. L., Dattilo, A. J., and Vincelli, P. 2003. Effect of soil
compaction and moisture on incidence of Phytophthora root rot on American
chestnut (Castanea dentata) seedlings. For. Ecol. Manage. 184:47-54.
Ristaino, J. B., and Gumpertz, M. L. 2000. New frontiers in the study of dispersal
and spatial analysis of epidemics caused by species in the genus Phytophthora.
Annu. Rev. Phytopathol. 38:541-576.
Roiger, D. J., and Jeffers, S. N. 1991. Evaluation of Trichoderma spp. for
biological control of Phytophthora crown and root rot of apple seedlings.
Phytopathology 81:910-917.
Rutter, P. A. 1990. Chestnut Pollinator’s Guide. Badgersett Research Corporation
Bulletins. http://www.badgersett.com/sites/default/files/info/publications/
Bulletin1v1_0.pdf
Santos, C., Nelson, C. D., Zhebentyayeva, T., Machado, H., Gomes-Laranio, J.,
and Costa, R. L. 2017. First interspecific genetic linkage map for Castanea
sativa ×Castanea crenata revealed QTLs for resistance to Phytophthora
cinnamomi. PLoS One 12:e0184381.
Sharpe, S. 2017. Phytophthora species Associated with American, Chinese, and
Backcross Hybrid Chestnut Seedlings in Field Sites in the Southeastern
United States. M.S. thesis, Clemson University, Clemson, SC, U.S.A.
Steiner, K. C., Westbrook, J. W., Hebard, F. V., Georgi, L. L., Powell, W. A., and
Fitzsimmons, S. F. 2017. Rescue of American chestnut with extraspecific genes
following its destruction by a naturalized pathogen. New For. 48:317-336.
Tucker, C. M. 1933. The Distribution of the Genus Phytophthora. Res. Bull.
Univ. Mo. Agric. Exp. Stn. no. 0184. University of Missouri Agricultural
Experimental Station, Columbia, MO. https://mospace.umsystem.edu/xmlui/
handle/10355/53375
1640 Plant Di sease / Vol. 103 No. 7
Westbrook, J. W. 2018a. Merging backcross breeding and transgenic blight
resistance to accelerate restoration of the American chestnut: The American
Chestnut Foundation’s breeding and selection plan 2015-2025. The American
Chestnut Foundation. https://www.acf.org/wp-content/uploads/2018/03/TACF_
2015-2025_BreedingSciencePlan_updated2_9_18.pdf
Westbrook, J. W. 2018b. A plan to diversify transgenic blight-tolerant American
chestnut populations. J. Am. Chestnut Found. 32:31-36.
Wickham, H. 2009. ggplot2: Elegant Graphics for Data Analysis. Springer-Verlag,
New York, NY, U.S.A.
Zentmyer, G. A. 1980. Phytophthora cinnamomi and the Diseases it Causes.
Monogr. No. 10. The American Phytopathological Society, St. Paul, MN, U.S.A.
Zentmyer, G. A. 1988. Origin and distribution of four species of Phytophthora.
Trans. Br. Mycol. Soc. 91:367-378.
Zhang, B., Oakes, A. D., Newhouse, A. E., Baier, K. A., Maynard, C. A., and
Powell, W. A. 2013. A threshold level of oxalate oxidase transgene
expression reduces Cryphonectria parasitica-induced necrosis in a transgenic
American chestnut (Castanea dentata) leaf bioassay. Transgenic Res. 22:
973-982.
Plant Disease / July 2019 1641
