Northeastern Naturalist Vol. 24, No. 1
L.R. Milbrath, A.S. Davis, and J. Biazzo
2017
37
2017 NORTHEASTERN NATURALIST 24(1):37–53
Demography of Invasive Black and Pale Swallow-wort
Populations in New York
Lindsey R. Milbrath1,*, Adam S. Davis2, and Jeromy Biazzo1
Abstract - Vincetoxicum nigrum (Black Swallow-wort) and Vincetoxicum rossicum (Pale
Swallow-wort) are perennial, twining vines introduced from Europe. Both species have become
invasive in northeastern North America in a variety of habitats. To develop parameters
for a population model for evaluating the control of swallow-worts, including biological
control, we collected data from 5 life stages on 20 different demographic rates involving
fecundity, germination, survival, and growth. We monitored 2 field and 2 forest populations
of Pale Swallow-wort, and 2 field populations of Black Swallow-wort in New York State using
a combination of marked individuals and sowing plots. Both species showed moderate
to high rates of seed germination and high survival of seedlings, with the primary exception
of a heavily shaded forest population. Survival generally continued to remain high postestablishment,
although transitions to different life stages varied by species, location, and
habitat. Black Swallow-wort became reproductive more quickly than Pale Swallow-wort.
These data add to the knowledge of swallow-wort demography and may offer insights into
the continued expansion and control of these invasive plants.
Introduction
Vincetoxicum nigrum (L.) Moench (= Cynanchum louiseae Kartesz &
Gandhi; Black Swallow-wort) and Vincetoxicum rossicum (Kleopow) Barb.
(= Cynanchum rossicum (Kleopow) Borhidi; Pale Swallow-wort) are herbaceous,
long-lived perennial plants in the Apocynaceae (subfamily Asclepiadoideae)
that were introduced into North America from Europe in the mid- to late-1800s
(DiTommaso et al. 2005b). Also known as “dog-strangling vines” in Ontario,
they have increased in abundance over the past 30–40 years in a variety of natural
and managed habitats in the northeastern US and southeastern Canada. They
are of particular concern in New York State, the southern part of New England,
and Ontario (L.R. Milbrath, pers. observ). Both species can establish in disturbed
and undisturbed habitats, and grow under a variety of soil pH and light levels,
although Pale Swallow-wort is particularly shade tolerant (Averill et al. 2010,
2011; Hotchkiss et al. 2008; Magidow et al. 2013; Smith et al. 2006). Besides
increasing control costs for land managers, the swallow-worts are a risk to plant
communities and associated fauna such as grassland birds; the alvar ecosystems
of the Lower Great Lakes region are one example of areas under threat (DiTommaso
et al. 2005b, Lawlor 2000).
1USDA-ARS Robert W. Holley Center for Agriculture and Health, 538 Tower Road, Ithaca,
NY 14853. 2USDA-ARS Global Change and Photosynthesis Research Unit, N-319 Turner
Hall, 1102 South Goodwin Avenue, Urbana, IL 61801. *Corresponding author - Lindsey.
Milbrath@ars.usda.gov.
Manuscript Editor: Sandy Smith
Northeastern Naturalist
38
L.R. Milbrath, A.S. Davis, and J. Biazzo
2017 Vol. 24, No. 1
Mechanical control of swallow-wort has been mostly ineffective. Herbicidal
control can be effective, but control costs and potential damage to other plant species
in natural areas are a concern (Averill et al. 2008, DiTommaso et al. 2013,
Mervosh and Gumbart 2015). In support of a biological-control program currently
being developed (e.g., Hazlehurst et al. 2012, Weed et al. 2011), we have been investigating
swallow-wort demography (i.e., germination, growth, and death rates
for populations of these 2 species). Matrix-population models can be a powerful
tool to identify key life-stage transitions that control the population growth of an
invasive plant, and therefore should be targeted for disruption (Caswell 2001).
Retrospective studies of other weed biological-control programs have indicated
that this approach has good potential for identifying effective biological-control
agents that should be prioritized for release (e.g., McEvoy and Coombs 1999, Shea
et al. 2005). The parameterization of such models requires collecting field data on
the survival, growth, and reproduction (vital rates) of different life stages of the
invasive species under consideration; including data on lower-level demographic
transitions (e.g., separating seed from seedling survival) can aid the identification
of more precise “target transitions” to further clarify the process of biocontrol-agent
selection and release. Demographic information for swallow-worts has increased
in recent years, but mainly for Pale Swallow-wort (e.g., Averill et al. 2010, 2011;
Hotchkiss et al. 2008), in contrast to Black Swallow-wort, for which field demography
is mostly unknown (but see Averill et al. 2011). Thus, additional field surveys
were needed to fill in the gaps in our understanding of swallow-wort population dynamics,
particularly regarding annual transitions among swallow-wort life stages.
We report here on the population density and vital-rate data obt ained from field
surveys of 6 populations of Black and Pale Swallow-wort that we conducted in support
of the population-modeling effort (model analyses to be reported elsewhere).
From these data, we addressed quantitative questions of demographic similarity or
differences among species, locations, and (for Pale Swallow-wort) habitats.
Field-Site Description
We chose field populations of Black and Pale Swallow-wort from representative
sites in New York State that had a history of swallow-wort infestation but where
the infestations were not being actively managed. We monitored 4 Pale Swallowwort
populations (2 locations, each with a forest and field population) and 2 Black
Swallow-wort populations (2 locations, each with only a field population because
Black Swallow-wort is uncommon in forested habitats) (Table 1). The Black Swallow-
wort site at Bear Mountain State Park is situated next to the Hudson River and
consists of a shallow soil composed primarily of fill. The Dutchess site (Cornell
University Cooperative Extension Office, Millbrook, NY) is an old field with deep
soils that had formerly been mown, but not during the 4 years previous to our study.
Both Black Swallow-wort sites are located in southeastern New York. The Pale
Swallow site at Great Gully Preserve (The Nature Conservancy, west-central New
York) consists of a deep-soil old field with an adjacent heavily shaded forest (1%
ambient light at canopy closure), whereas the site at Robert G. Wehle State Park
Northeastern Naturalist Vol. 24, No. 1
L.R. Milbrath, A.S. Davis, and J. Biazzo
2017
39
(northern New York, next to Lake Ontario) includes a shallow-soil alvar (limestone
barrens) field with an adjacent moderately shaded forest (9–30% ambient light;
Averill 2009) (Table 1).
Methods
Swallow-wort life history
Both Black and Pale Swallow-wort are long-lived, herbaceous perennials that
annually produce 1 to several twining stems from a subterranean, semi-woody rootstock.
The stems are typically 50–200 cm long, and the number of stems generally
increases over the years if growing conditions are suitable. Clusters of flowers are
produced in the leaf axils; flowers are dark purple in Black Swallow-wort and pink
to maroon in Pale Swallow-wort. Follicles (seed pods) dehisce from August into the
fall, and the seeds bear an apical tuft of hairs that enhance wind dispersal, similar
to the related Asclepias spp. (milkweeds). Most seedlings appear to emerge in the
spring, and the plants may remain in a vegetative state for a few to several years before
flowering (Averill et al. 2010; DiTommaso et al. 2005b; L.R. Milbrath, unpubl.
data). Based on the known life cycle, we identified 5 life stages to monitor vital rates
of survival, transitions to other life stages, and fecundity: seeds, seedlings, vegetative
juveniles (defined as being in at least their 2nd season of growth), small flowering
plants (defined as having 1–2 stems), and large flowering plants (3 or more stems).
For purposes of population modelling, we measured vital rates primarily on an annual
cycle from August to the following August (just prior to seed dispersal).
Seeds, seedlings, and small juvenile plants
We used seed-sown plots and seed bags to collect vital-rate data for seeds,
seedlings, and the earliest vegetative juvenile stages. We could not differentiate
seedlings from young juveniles in existing stands without greatly disturbing the
plants (destructive sampling is the only reliable method), and high densities of
plants at both of these stages (which can be greater than 1500 m-1; Smith et al. 2006)
Table 1. Site characteristics of the Black and Pale Swallow-wort study locations in New York State
during 2009–2012. Adapted from Averill (2009).
Black Swallow-wort Pale Swallow-wort
Great Gully Wehle
Bear Mt. Dutchess (field and forest) (field; forest)
County Rockland Dutchess Cayuga Jefferson
Lat., long. 41°18'N, 73°58'W 41°46'N, 73°44'W 42°48'N, 76°40'W 43°51'N, 76°17'W
Elevation (m) 5 120 190 80
Soil depth (cm) 0–25 >200 > 200 0–25; 20–60
Soil texture Gravelly sandy loam Gravelly loam Silt loam Silt loam; channery
silt loam
Drainage Well drained Somewhat excessively Well drained Excessively drained;
drained somewhat excessively
drained
Soil pH 7.2 5.4 7.0 6.7; 7.1
% organic matter 4.2 6.7 4.3 11; 26
Northeastern Naturalist
40
L.R. Milbrath, A.S. Davis, and J. Biazzo
2017 Vol. 24, No. 1
prevented us from reliably marking and tracking individuals within and between
seasons. We selected areas that were free of swallow-wort within or near existing
swallow-wort stands. We cut existing vegetation to 5 cm for plot establishment, but
did not cut it again during our study. Each year, we cut back or removed nearby
flowering swallow-wort plants to minimize natural seed rain (deposition of newlyproduced
seeds) into the plots. We established 1.8 m x 0.6 m plots during the early
summer of 2009 (2010 for Bear Mt. only). We divided the plots into three 0.6 m x
0.6 m subplots that we marked at each subplot corner with PVC pipes. Two subplots
were for sowing up to 2 cohorts of seeds, and 1 subplot served as an unsown
control for correcting observed seedling establishment in the sown subplots from
any preexisting seed bank. Subplots were randomly arranged within each plot. We
established 5 (Black Swallow-wort) or 10 (Pale Swallow-wort) plots for each of the
6 populations due to the size of the plant populations available.
We collected mature seeds from at least 100 plants at the same site and habitat
in which the seeds were to be sown. Filled seeds (likely to contain an embryo) were
counted into lots of 100 seeds. We determined initial viability by cold–wet stratifying
3 lots of 100 filled seeds at 4 °C for 3 months, germinating the seeds in an
incubator at 25:20 °C and a photoperiod of 14:10 h (L:D). We tested the remaining
non-germinated seeds for viability with a 1% solution of tetrazolium chloride. Initial
viability was 93–99%. We sowed and lightly pressed down seeds into subplots
in a 10 x 10 grid pattern on the soil surface. For field populations, 1 subplot was
sown in August and the 2nd in October, mainly in different years. We collected and
sowed early-sown (August) seeds on the same day to avoid altering seed dormancy.
Later-collected seeds show higher rates of dormancy (DiTommaso et al. 2005a). For
forest populations, mature seeds for sowing were only available later in the season.
We counted newly emerged seedlings monthly in September, October, June, July,
and August for up to 3 years. We corrected seedling-emergence rates for any background
seed bank germination by subtracting the number of seedlings counted in the
control subplots. Seedlings that emerged in September and October were defined as
fall-germinated. To prevent overestimating fall-emergence rates in the year when
seeds were sown, we corrected these rates by multiplying them by the proportion of
dehiscing pods (and thus seed rain) present at the time of sowing (13–60% of that season’s
seed rain, depending on the location). We tracked emergence for 4 time periods:
fall emergence from the current season’s seed rain (i.e., seeds that were sown), spring
and summer emergence from seeds <1 year old, fall emergence from seeds >1 year
old, and spring and summer emergence from seeds >1 year old. We marked up to 400
seedlings per population (20 per subplot) with a labeled, plastic ring anchored around
the base of the plant for further monitoring. We removed by hand seedlings that had
been counted but not marked. Overwintering survival of marked, fall-germinated
seedlings was assessed in June. We conducted an annual survey of marked plants in
August to determine survival, flowering status, and stem number, i.e., to track changes
in life stages. In order to gather data on all relevant vital rates, especially changes
in life stages, we conducted our surveys over a 3-year period.
Seedling-emergence rate equals germination rate multiplied by seed-survival
rate. We calculated values for the latter 2 parameters as follows. Seed survival was
Northeastern Naturalist Vol. 24, No. 1
L.R. Milbrath, A.S. Davis, and J. Biazzo
2017
41
first indirectly assessed for each population by placing 30 filled seeds into organza
bags that we placed inside wire-mesh cages to prevent predation by rodents. We
placed 2 seed bags per cage, with 4 cages deployed per population, for a total of 8
seed bags per population. The cages with seed bags were fixed to the ground next
to the sowing plots in September 2011. We recovered 1 bag in May 2012 after overwintering,
and the 2nd bag in August 2012. All seeds were immediately germinated
for 2 weeks at 25:20 °C and a photoperiod of 14:10 h (L:D), after which we scored
the remaining seeds for viability using tetrazolium chloride. We approximated the
seed-survival rate for each sown-subplot for each population by dividing the average
proportion of surviving viable seeds from the August seed bags (n = 4) by (1
– the cumulative proportion of seedling emergence from sown seeds observed the
first year). We estimated germination rates by dividing observed seedling-emergence
rates for each of the 4 time periods described earlier (adjusted as needed for
any previous year’s emergence) by the seed-survival rate. For each swallow-wort
population, we calculated an average value for each life-stage specific vital rate
using each subplot (usually 5–20) and year (1–2) combination from which relevant
observations were made.
Large juvenile and flowering plants
We collected vital-rate data from marked plants including larger vegetative
juveniles, small flowering plants, and large flowering plants. Five lower-density
areas, each approximately10 m x 10 m, were selected across each population. We
did not include high-density patches, which can have plant population densities
of ≥200 stems m-2 (L.R. Milbrath, unpubl. data), because of the difficulty of identifying
individual plants. Vital rates can differ between high- and lower-density
patches (e.g., Evans et al. 2012). We randomly selected up to 10 individuals of
each of the 3 life stages (if present) within each area and permanently marked
each with a flag and a labeled, plastic-coated wire ring anchored around the base
of the plant. Plants were also mapped to aid in relocation. We marked the plants in
2009 (2010 for Bear Mt. only) for a total of up to 50 individuals per life stage for
each population. Larger vegetative juveniles were 10–30 cm in height and usually
single-stemmed. While it is possible that some of the marked large flowering plants
consisted of more than 1 plant crown (see Averill et al. 2010), we considered them
an appropriate biological unit of interest that probably, but not always, originated
from a polyembryonic seed. Between 55% and 78% of Pale Swallow-wort and 22%
of Black Swallow-wort seeds, respectively, are polyembryonic (Averill et al. 2010;
Sheeley 1992; Smith et al. 2006; L.R. Milbrath, unpubl. data).
We censused marked plants annually in August (prior to seed dispersal) for survival,
flowering status, stem number, and pod number per plant. We collected up
to 4 years of pod data for most populations; pod numbers were estimated if vines
were excessively entangled with neighboring plants. We randomly sampled 50
pods from both small and large flowering plants, and counted filled seeds per pod
to estimate the number of viable seeds produced per plant. For vegetative juveniles,
we combined vital rates for small juveniles from the sown plots and large juveniles
from the sampled areas using a weighted average based on the percentages of small
Northeastern Naturalist
42
L.R. Milbrath, A.S. Davis, and J. Biazzo
2017 Vol. 24, No. 1
and large juveniles observed at a given location (see population structure subsection
below). We calculated an average value for each life-stage–specific vital rate
for each swallow-wort population using each area (usually 5), and year (2–4) in the
case of fecundity data, from which relevant observations were made.
For all vital rates (not including seedling-emergence data), we analyzed the data
for either the 4 field populations of Black and Pale Swallow-wort or the 2 field and 2
forest populations of Pale Swallow-wort. For the species comparisons, we conducted
analysis of variance (PROC MIXED, SAS 9.4, SAS Institute, Inc., Cary, NC) using
the fixed effects of species and location nested within species and the random effects
of subplot nested within plot nested within location nested within species (sowingplot–
derived data) or area nested within location nested within species (other data).
For the forest and field comparisons, we employed analysis of variance (PROC
MIXED) using the fixed effects of habitat, location, and their interaction, and the
random effects of subplot nested within plot nested within the habitat by location
interaction (sowing-plot data) or area nested within the habitat by location interaction
(other data). Proportional data were arcsine-square–root transformed for
analyses. No data were available from 1 forest population for 2 of the Pale Swallowwort
habitat comparisons; thus, we conducted a 1-way analysis of the 3 remaining
habitat–location combinations. We compared means for significant factors using the
least-significant difference test with Bonferroni correction. The results were used
to determine whether vital rates (model parameters) among all or some populations
should be combined for future population-model analyses.
Population structure
We assessed the population structure for each population in July 2010, prior to
the first annual census, except that we sampled seeds from the soil-seed bank in
2012. All stages except seeds were counted in quadrats. We subdivided each sampling
area into 4 zones and randomly placed a 1-m2 quadrat in each zone, avoiding
previously marked plants. We counted the total number of plants at each life stage
in the 1-m2 quadrats for the forest populations; seedlings and small vegetative juveniles
(less than 5 cm tall for Pale Swallow-wort; less than 10 cm tall for Black Swallow-wort)
were destructively sampled. Seedlings can be distinguished from what we classify
as juveniles by the lack of bracts, which are only found on the latter, and often an
attached seed coat. For the field populations, we placed 2 smaller quadrats (generally
0.0156– 0.0625 m2) in opposite corners of the 1-m2 quadrat. Due to the high
plant-density in these subplots, we dug plants out of the ground to assess numbers
of seedlings and small juveniles. We also destructively sampled all swallow-wort
plants in these subplots, regardless of life stage. We also counted larger-sized juveniles,
small flowering plants, and large flowering plants growing in the remaining
area of the 1-m2 quadrat. We calculated an average density per 1 m2 for each life
stage for each of the 5 areas within each population.
We conducted seed-bank sampling by taking 34 soil cores in a double-zigzag
pattern within each area using a 7.6-cm–diameter soil probe to a depth of 5 cm,
equivalent to sampling a surface area of 0.15 m2. Soil cores for each sampled area
were bulked and returned to the laboratory under refrigeration until processing.
Northeastern Naturalist Vol. 24, No. 1
L.R. Milbrath, A.S. Davis, and J. Biazzo
2017
43
We recovered seeds by elutriation, dried the elutriated sample, and counted filled
(potentially viable) and unfilled (non-viable) seeds. We determined the number of
viable seeds by cold–wet stratifying filled seeds at 4 °C for 3 months and then germinating
them in an incubator at 25:20 °C and a photoperiod of 14:10 h (L:D). We
further tested non-germinated seeds for viability with a 1% solution of tetrazolium
chloride. We calculated a grand-mean density per life stage for each population.
Results
Vital rates (survival, germination, other transitions between life stages, and fecundity)
often differed between swallow-wort species and habitats (field or forest)
and among locations (Tables 2, 3), indicating that we should not pool data from
the 6 populations in modeling analyses. Annual seedling-emergence was generally
highest the first year after sowing, ranging from 5% to 39% across swallow-wort
species, habitats, and the 2 sowings (Fig. 1). Fewer seedlings usually emerged in
the second year (0–11%) except for the first seed cohort sown at Wehle in both the
field and forest (20–25%; Fig. 1A). Only a few seedlings appeared in the third year
at any site (Fig. 1A). Fall germination and emergence (September to October) of
recently deposited Black and Pale Swallow-wort seeds was rare and only occurred
in 2 of the open-field populations (Bear Mt., Wehle). However, seedling recruitment
the following spring and summer was moderate to high at all locations, with
greater germination rates for field populations of Pale Swallow-wort than forest
populations (Tables 2, 4). The typical seasonal pattern we observed for both swallow-
wort species is that the majority of plants germinate in the late spring (May/
June), with continued but greatly decreasing emergence into the fall. After 1 year
in the seed bank, fall emergence (September, October) of the remaining seeds was
uncommon (Bear Mt., Great Gully) to moderate (Wehle) and occurred only in field
populations (Table 4). Thus, estimated fall germination rates were similar or higher
(0–31%) 1 year after seed rain than in the first few months following seed rain
(0–1%) (Table 4). Spring/summer germination rates in the second growing season
remained variable among Pale but not Black Swallow-wort populations, including
0% germination in the forest population at Great Gully (T ables 2, 4).
Estimated seed-survival rates were low to high (14–74%) depending on the
population, with the highest survival at the Great Gully site (Tables 2, 4). Survival
of all later stages was generally very high among all populations (averages of
72–100%; Table 4). One exception involved 33% survival of Pale Swallow-wort
seedlings at the low-light Great Gully forest. Spring-germinated and established
seedlings of Black Swallow-wort (Bear Mt. population) had lower survival than
other field populations (Tables 2, 4). Also at the Bear Mt. population, the few
juveniles that had originated from fall-germinated seedlings had a lower summersurvival
rate (50%) than similar juveniles of Pale Swallow-wort (Tables 2, 4).
Older Black Swallow-wort juveniles had higher survival than Pale Swallow-wort
juveniles in the open field, and Pale Swallow-wort juveniles at the Wehle site had
higher survival than at the Great Gully site (Tables 2, 4). We observed very little
mortality of flowering plants (Table 4).
Northeastern Naturalist
44
L.R. Milbrath, A.S. Davis, and J. Biazzo
2017 Vol. 24, No. 1
Figure 1. Percentage emergence of seedlings (mean ± SD, n = 5–10, 100 seeds sown per
subplot) for 6 Swallow-wort populations and 2 different cohorts of seeds monitored (A)
for 3 years, (except 2 years for Bear Mt) or (B) 2 years. Emergence for each year was from
September to the following August. Bear Mt and Dutchess were Black Swallow-wort field
sites; the remaining locations were Pale Swallow-wort.
Northeastern Naturalist Vol. 24, No. 1
L.R. Milbrath, A.S. Davis, and J. Biazzo
2017
45
Fecundity of small and large flowering plants was generally less at the Bear Mt.
site relative to other field populations of Black and Pale Swallow-wort. The forest
population at Great Gully generally produced the fewest seeds, and the field
population at Wehle had the most seeds among Pale Swallow-wort locations (Tables
3, 5). No plants matured to a flowering state within 1 year of growth (seedling or juvenile
[fall] to small flowering transitions, Table 5). We observed a limited proportion
(less than 6%) of marked vegetative juveniles of Pale Swallow-wort annually transitioning
to the small flowering stage, except at the Great Gully forest site, where we did not
record this phenomenon (Table 5). In contrast, at least one-quarter of Black Swallowwort
juveniles became small flowering plants the following year (Table 5). At the
Table 2. Analysis of variance results for vital rates of percentage germination and survival of different
life stages among either 4 open-field populations of Black and Pale Swallow-wort, or among 2
field and 2 forest populations of Pale Swallow-wort, in New York State. F statistics (numerator and
denominator degrees of freedom) and level of significance are given. *P < 0.05, **P < 0.01, and ***P
< 0.001. No statistics are given if there was no variation in values among locations or, for some Pale
Swallow-wort habitat comparisons, no data were available from forest populations (see Table 4). New
seeds were <1 year old, and old seeds were >1 year old.
Field population Pale Swallow-wort
comparisons field and forest comparisons
Location
Vital rate/Life stage Species (species) Habitat Location Habitat*location
% germination
Fall, new seeds 0.14 2.18 2.25 2.25 2.25
(1, 56) (2, 56) (1, 66) (1, 66) (1, 66)
Spring, new seeds 0.09 13.22 14.37 8.38 0.14
(1, 56) (2, 56)*** (1, 66)*** (1, 66)** (1, 66)
Fall, old seeds 5.01 5.66 10.98 5.70 5.70
(1, 56)* (2, 56)** (1, 66)** (1, 66)* (1, 66)*
Spring, old seeds 0.46 2.38 0.00 50.75 11.60
(1, 56) (2, 56) (1, 66) (1, 66)*** (1, 66)**
% survival
Seed 312.08 349.13 208.75 408.42 7.53
(1, 26)*** (2, 26)*** (1, 36)*** (1, 36)*** (1, 36)**
Fall seedling 0.17 1.89 - - -
(1, 17) (1, 17)
Spring seedling 7.19 5.15 51.06 56.10 40.75
(1, 50)** (2, 50)** (1, 64)*** (1, 64)*** (1, 64)***
Established seedling 10.85 27.92 23.84 19.12 17.09
(1, 50)** (2, 50)*** (1, 61)*** (1, 61)*** (1, 61)***
Juvenile (Fall) 8.19 0.76 - - -
(1, 16)* (1, 16)
Juvenile 8.14 3.05 1.75 10.64 0.07
(1, 16)* (2, 16) (1, 16) (1, 16)** (1, 16)
Small flowering 1.00 1.00 - - -
(1, 16) (2, 16)
Large flowering - - - - -
Northeastern Naturalist
46
L.R. Milbrath, A.S. Davis, and J. Biazzo
2017 Vol. 24, No. 1
Dutchess site only, some Black Swallow-wort plants became reproductive in their
third year of growth. Also at this site, some newly flowering individuals possessed 3
or more stems (all from the same root crown), and thus represented a juvenile-to-large
flowering transition. The transition from small to large flowering plants was variable
across locations. For Pale Swallow-wort, the field population at Great Gully showed
the most change in flowering size-classes compared with the other field and forest
populations. For Black Swallow-wort, we did not observe small flowering plants at
Bear Mt. that transitioned to a large size, whereas 50% of small flowering plants grew
into large flowering plants the following year at the Dutchess site (Table 5). A variable
proportion of large flowering plants of Pale Swallow-wort (that typically had 3 stems
the previous year), but not Black Swallow-wort, became small flowering plants (with
2 stems) the following year (Tables 3, 5).
Viable seed densities (seed bank) were very low at the 2 forest sites (Table 6). In
the lower-density patches utilized for this study, seedlings and small juveniles were
most abundant at the 2 Pale Swallow-wort field sites (Table 6). The density of large
flowering plants was low at all locations, and this stage was absent from sampled quadrats
in forest environments (Table 6), though a few such plants appeared in the Wehle
forest population during the course of the surveys. Large flowering plants typically
Table 3. Analysis of variance results for vital rates of fecundity and percentage of individuals transitioning
to other life stages among either 4 open-field populations of Black and Pale Swallow-wort, or
among 2 field and 2 forest populations of Pale Swallow-wort, in New York State. F statistics (numerator
and denominator degrees of freedom) and level of significance are given. *P < 0.05, **P < 0.01, and
***P < 0.001. No statistics are given if there was no variation in values among locations. For some Pale
Swallow-wort habitat comparisons, no data were available from 1 forest population, so we conducted
a 1-way analysis of the 3 remaining habitat–location combinations (see Table 5).
Field population Pale Swallow-wort
comparisons field and forest comparisons
Location
Vital rate/Life stage Species (species) Habitat Location Habitat*location
Fecundity
Small flowering 27.45 14.98 45.33 23.65 0.38
(1, 16)*** (2, 16)*** (1, 16)*** (1, 16)*** (1, 16)
Large flowering 2.78 36.26 - - 20.58
(1, 14) (2, 14)*** (2, 9)***
% life-stage transition
Seedling to small flowering - - - - -
Juvenile (fall) to small flowering - - - - -
Juvenile to small flowering 65.49 2.81 5.50 7.32 0.64
(1, 16)*** (2, 16) (1, 16)* (1, 16)* (1, 16)
Juvenile to large flowering 1.00 1.00 - - -
(1, 16) (2, 16)
Small to large flowering 6.93 38.71 5.00 1.91 6.82
(1, 16)* (2, 16)*** (1, 16)* (1, 16) (1, 16)*
Large to small flowering 11.83 0.43 - - 5.74
(1, 14)** (2, 14) (2, 9)*
Northeastern Naturalist Vol. 24, No. 1
L.R. Milbrath, A.S. Davis, and J. Biazzo
2017
47
Table 4. Vital rates (mean and range among subplots or sampling areas) of percentage germination and survival of different life stages among 6 Black
and Pale Swallow-wort populations in New York State observed over 2–3 years. New seeds were <1 year old, and old seeds were >1 year old. n = 10–20.
Survival was measured from August to the following August, except fall-germinated seedlings (overwinter survival), spring-germinated seedlings (June–
August survival), and juveniles originating as fall-germinated seedlings (June–August survival). n (subplot or sampled area and year combinations) = 5–10
(seed), 2–14 (fall seedling), 5–20 (spring and established seedling), 2–12 (juvenile-fall), 5 (juvenile and small flowering), 2–5 (large flowering). Although
large flowering plants were not initially marked in the 2 forest populations, some individuals did develop from small flowering plants at the Wehle location.
Black Swallow-wort Pale Swallow-wort
Bear Mt. Dutchess Great Gully Wehle
Vital rate/Life stage(s) Field Field Field Forest Field Forest
% germination
Fall, new seeds 0.7 (0–7.0) 0.0 (0.0–0.0) 0.0 (0.0–0.0) 0.0 (0.0–0.0) 1.1 (0.0–11.2) 0.0 (0.0–0.0)
Spring, new seeds 89.2 (8.3–100) 32.3 (0–100) 54.1 (24.7–83.8) 30.8 (0–65.3) 69.8 (0.0–100) 47.0 (30.2–69.6)
Fall, old seeds 6.3 (0–63.4) 0.0 (0.0–0.0) 2.3 (0.0–15.2) 0.0 (0.0–0.0) 31.2 (0.0–100) 0.0 (0.0–0.0)
Spring, old seeds 46.3 (0.0–100) 53.8 (0.0–100) 25.3 (0.0–100) 0.0 (0.0–0.0) 53.6 (0.0–100) 85.3 (37.9–100)
% survival
Seed 14.1 (12.0–16.4) 17.1 (16.1–19.7) 55.3 (48.5–63.9) 74.4 (63.7–84.6) 18.0 (15.6–20.2) 45.3 (39.8–60.3)
Fall seedling 100.0 (100–100) - 100.0 (100–100) - 90.0 (0.0–100) -
Spring seedling 72.1 (19.0–100) 96.3 (85.7–100) 95.1 (75.0–100) 33.6 (0.0–100) 96.9 (60.0–100) 95.7 (83.3–100)
Established seedling 38.9 (0.0–82.1) 99.0 (90.0–100) 88.8 (60.0–100) 32.7 (0.0–100) 90.4 (60.0–100) 85.2 (41.2–100)
Juvenile (Fall) 50.0 (0.0–100) - 100.0 (100–100) - 93.5 (50.0–100) -
Juvenile 96.0 (90.0–100) 100.0 (100–100) 86.7 (74.6–94.6) 83.3 (76.1–100) 93.9 (80.8–100) 93.7 (89.0–95.7)
Small flowering 98.0 (90.0–100) 100.0 (100–100) 100.0 (100–100) 100.0(100–100) 100.0 (100–100) 100.0 (100–100)
Large flowering 100.0 (100–100) 100.0 (100–100) 100.0 (100–100) - 100.0 (100–100) 100.0 (100–100)
Northeastern Naturalist
48
L.R. Milbrath, A.S. Davis, and J. Biazzo
2017 Vol. 24, No. 1
Table 5. Vital rates (mean and range among subplots or sampling areas) of fecundity and the percentage of individuals transitioning to other life stages
among 6 Black and Pale Swallow-wort populations in New York State. Fecundity = viable seeds per plant, collected over 2–4 years. n (sampled area and
year combinations) = 10–20 (small flowering) and 4–20 (large flowering). % life-stage transition = percentage of individuals of a given life stage transitioning
to a different life stage (excluding germination and seedling to juvenile survival, see Table 3). Measured from August to the following August, except
seedling-to-small flowering stage and juvenile (originating as fall-germinated seedling)-to-small flowering stage (June–August). n (subplot or sampled
area and year combinations) = 5–20 (seedling/small), 1–12 (juvenile-fall/small), 5 (juvenile/small, juvenile/large, small/large), 2-5 (large/small). Although
large flowering plants were not initially marked in the 2 forest populations, some individuals did develop from small flowering plants at the Wehle location.
Black Swallow-wort Pale Swallow-wort
Bear Mt. Dutchess Great Gully Wehle
Vital rate/Life stage(s) Field Field Field Forest Field Forest
Fecundity
Small flowering 17 (4–28) 98 (46–241) 95 (19–190) 4 (0–25) 177 (61–356) 67 (2–163)
Large flowering 48 (16–115) 963 (527–1462) 300 (78–624) -- 1051 (389–1841) 221 (164–344)c
% life-stage transition
Seedling to small flowering 0.0 (0.0–0.0) 0.0 (0.0–0.0) 0.0 (0.0–0.0) 0.0 (0.0–0.0) 0.0 (0.0–0.0) 0.0 (0.0–0.0)
Juvenile (Fall) to small flowering 0.0 (0.0–0.0) - 0.0 (0.0–0.0) - 0.0 (0.0–0.0) -
Juvenile to small flowering 26.0 (10.0–40.0) 33.9 (18.7–43.7) 1.1 (0.0–2.8) 0.0 (0.0–0.0)) 5.9 (0.0–9.8) 1.7 (0.0–5.8)
Juvenile to large flowering 0.0 (0.0–0.0) 0.4 (0.0–2.0) 0.0 (0.0–0.0) 0.0 (0.0–0.0) 0.0 (0.0–0.0) 0.0 (0.0–0.0)
Small to large flowering 0.0 (0.0–0.0) 50.0 (40.0–70.0) 16.0 (0.0–30.0) 0.0 (0.0–0.0) 2.0 (0.0-10.0) 4.0 (0–20.0)
Large to small flowering 0.0 (0.0–0.0) 0.0 (0.0–0.0) 8.0 (0.0–20.0) - 12.7 (0.0–33.0) 50.0 (50.0–50.0)
Northeastern Naturalist Vol. 24, No. 1
L.R. Milbrath, A.S. Davis, and J. Biazzo
2017
49
Table 6. Population densities (number per m2, mean ± SD, n = 5) of different life stages (including average percentage of small juveniles, n = 5) for 6 Black
and Pale Swallow-wort populations in New York State in 2010 (seed bank assessed in 2012). Black = Black Swallow-wort and Pale = Pale Swallow-wort.
Pale Swallow-wort: small juveniles were less than 5 cm tall, and large juveniles were ≥5 cm tall. Black Swallow-wort: small juveniles were less than 10 cm tall, and large
juveniles were ≥10 cm tall. Small flowering plants had 1–2 stems and large flowering plants had 3 or more stems.
Juveniles-small
Species Habitat Location Viable seeds Seedlings (% all juveniles) Juveniles-large Small flowering Large flowering
Black Field Bear Mt. 76.1 ± 25.6 32.0 ± 18.9 75.6 ± 73.3 (50.1) 43.5 ± 19.7 45.8 ± 16.7 0.2 ± 0.2
Black Field Dutchess 70.9 ± 43.3 46.8 ± 37.8 58.4 ± 82.1 (37.8) 56.2 ± 24.4 9.9 ± 4.3 1.4 ± 0.7
Pale Field Great Gully 6.4 ± 4.6 109.2 ± 104.8 104.0 ± 63.9 (72.1) 47.3 ± 60.0 5.7 ± 5.0 0.6 ± 0.5
Pale Forest Great Gully 0.0 ± 0.0 7.4 ± 8.0 15.6 ± 12.9 (23.9) 48.9 ± 12.9 9.7 ± 6.3 0.0 ± 0.0
Pale Field Wehle 130.3 ± 75.4 313.6 ± 325.9 201.2 ± 130.6 (67.2) 97.9 ± 77.3 17.3 ± 18.2 0.5 ± 0.6
Pale Forest Wehle 3.9 ± 5.8 48.3 ± 44.0 38.9 ± 29.7 (71.0) 18.0 ± 16.1 5.6 ± 4.9 0.0 ± 0.0
Northeastern Naturalist
50
L.R. Milbrath, A.S. Davis, and J. Biazzo
2017 Vol. 24, No. 1
possessed 3–5 stems. However, at the Dutchess site, where this life stage was more
abundant, we observed plants with up to 24 stems arising from a single root crown.
Discussion
New information on swallow-wort demography, particularly for Black Swallow-
wort, was generated from this survey of invasive populations in New York
State. The results indicate that some vital rates for the 2 species appear to be similar,
although we also detected distinct differences in life-stage transitions between
species, habitats, and locations. Vital rates may change as plant densities change,
i.e, they can exhibit density-dependence (e.g., Evans et al. 2012). Thus, our results
from lower-density patches do not necessarily represent the population dynamics
of higher-density patches.
The multiyear pattern of seedling recruitment that we typically observed
from a single sowing of seed for either species (most in year 1, less in year 2,
very few individuals in year 3; see also Averill et al. 2010, Ladd and Cappuccino
2005 for Pale Swallow-wort) suggests that the seed bank for Black and Pale
Swallow-wort lasts about 3 years. A short-lived seed bank is also known for
related species in the Apocynaceae (Burnside et al. 1981). The atypical pattern
at the Wehle field site, which involved much less seedling emergence in year 1
(2010) than year 2 (2011) for the first seed cohort (Fig. 1A), may have been due
to drought-type conditions observed in the spring and summer of 2010 followed
by adequate spring moisture in 2011. The role of dry conditions in delaying germination
for several months deserves further investigation, although it should
be noted that this delay did not increase seedling recruitment at the Wehle site in
the third year (2012).
Substantial variability occurred among populations in seedling emergence
within the first summer of growth. This pattern has also been reported from field
experiments with Pale Swallow-wort (3–58% emergence) and Black Swallowwort
(9–40%) (Averill et al. 2010, Ladd and Cappuccino 2005, Magidow et al.
2013). This variability in emergence is presumably due to desiccation of recently
germinated seedlings, prolonged saturated soils to which swallow-worts appear
sensitive, predation, plant competition, and other factors (Averill et al. 2010).
However, in general, swallow-wort emergence (and seedling survival) is much
higher than that reported for other species common in old fields, such as Dipsacus
sylvestris Huds. (Teasel), Phalaris arundinacea L. (Reed Canarygrass), and Solidago
altissima L. (Canada Goldenrod) (Lindig-Cisneros and Zedler 2002, Meyer
and Schmid 1999, Werner and Caswell 1977). The survival of new Pale Swallowwort
seedlings was greatly reduced only in the heavily-shaded forest at Great
Gully, but not in the moderately-shaded forest at Wehle or the open fields. Seedlings
and other life stages of Pale Swallow-wort typically perform better in fields
or forest light-gaps than the forest understory (Averill et al. 2011, Hotchkiss et
al. 2008, Smith et al. 2006). Besides low-light stress, we suspect that seedling
mortality in forests may also be due to predation by slugs and the smothering of
seedlings by leaf litter. However, this hypothesis requires further investigation.
Northeastern Naturalist Vol. 24, No. 1
L.R. Milbrath, A.S. Davis, and J. Biazzo
2017
51
Also, further research should be conducted to directly assess seed survival, including
determination of seed predation and use of germination trays in the field;
deriving seed survival and germination rates from seedling emergence data alone
might under- or overestimate these rates.
DiTommaso et al. (2005a) reported that 20–50% of Pale Swallow-wort seeds
produced in August can immediately germinate and that there may be both a late
summer/early fall and spring flush of seedlings, especially if seeds were buried.
However, we observed only a spring flush of surface-sown seeds at our study locations
for Black and Pale Swallow-wort, although germination did continue into the
fall. Averill et al. (2010) reported a similar pattern for experimental field populations
of Pale Swallow-wort. Fall emergence of seedlings from surface-sown seeds,
especially recently produced seeds, was rare in open fields and, in the case of Pale
Swallow-wort, never occurred in forested habitats. Most of the seed maturation and
pod dehiscence of Pale Swallow-wort in forests was often delayed up to 1 month
relative to neighboring field populations (L.R. Milbrath, pers. observ.), which, combined
with environmental effects on seed dormancy from shading (DiTommaso et
al. 2005a), likely prevents fall germination and establishment in forests. We did not
include infestations of Black Swallow-wort in forests in our study because of the
current scarcity of such populations. Reasons for this relative lack of recruitment
or persistence are unknown.
Once individual plants had transitioned to a vegetative juvenile stage, survival
rates remained generally high through all subsequent life stages of Black and Pale
Swallow-wort, regardless of the habitat. The average lifespan of individual plants
of these 2 long-lived perennial species is unknown, but high survival rates undoubtedly
contribute to sustaining high-density populations. The main effect of shading
on established Pale Swallow-wort growing in forests, besides reducing seed production,
appears to be limiting the annual proportion of juveniles transitioning to a
reproductive state. Shading also limited if not prevented growth of small flowering
plants (which we defined as having 1–2 stems) to a larger size. Averill et al. (2011)
had previously noted that an increase in stem number from one year to the next was
much less for Pale Swallow-wort plants in forests than in open fields.
We did not observe any instances of rapid maturation of seedlings to a reproductive
state; however, such quick development has been reported under some
field conditions, especially for Black Swallow-wort. For example, Averill (2009)
noted a single Pale Swallow-wort seedling flowering (with no seed pods) in the
first year and 6 plants (with seed pods and seeds) in the second year of growth
in a field experiment (see also Averill et al. 2010). Also, Magidow et al. (2013)
observed flowering and seed production within the first year of growth in a common-
garden experiment utilizing outdoor potted Pale Swallow-wort (0.1–0.5%
of seedlings) and Black Swallow-wort (13–17%) seedlings. Black Swallow-wort
appears to become reproductively mature sooner than Pale Swallow-wort, and at
1 location (Dutchess), it grew more rapidly to a larger reproductive size; these
were the 2 notable differences in life-stage transitions between the 2 species (see
Table 5). Black Swallow-wort also has a longer period of flowering in the field
and allocates more resources to aboveground tissues than Pale Swallow-wort
Northeastern Naturalist
52
L.R. Milbrath, A.S. Davis, and J. Biazzo
2017 Vol. 24, No. 1
(Milbrath 2008, Milbrath et al. 2016), which may promote its dispersal to new
areas. Although Pale Swallow-wort apparently has a longer juvenile phase, its
higher allocation to root growth beginning at the seedling stage (Averill et al.
2010, Milbrath 2008, Milbrath et al. 2016) may allow it to grow and survive under
a range of competitive environments.
These survey data add to the known natural life-history of these invasive species
in the northeastern US, and may assist naturalists and land managers in understanding
the ongoing invasion of natural and managed areas by these introduced plants.
Incorporation of these data into demographic population models will aid in the
development of control recommendations for Black and Pale Swallow-wort across
the various locations and habitats in which they are established.
Acknowledgments
We thank Jenna Antonino DiMare, Elizabeth Drake, Jamie Freeman, Michael Liu, Scott
Morris, Sarah Palmer, and Ariel Saffler (Cornell University) for their assistance in collecting
field data. Mention of trade names or commercial products in this article is solely
for the purpose of providing specific information and does not imply recommendation or
endorsement by the US Department of Agriculture. USDA is an equal-opportunity provider
and employer.
Literature Cited
Averill, K.M. 2009. Vincetoxicum spp. biology and ecology in New York State: Establishment
success, vegetative expansion, and physiology of two invasive perennial vines in
the Apocynaceae. M.Sc. Thesis. Cornell University, Ithaca, NY. 123 pp.
Averill, K.M., A. DiTommaso, and S.H. Morris. 2008. Response of Pale Swallowwort
(Vincetoxicum rossicum) to triclopyr application and clipping. Invasive Plant Science
and Management 1:196–206.
Averill, K.M., A. DiTommaso, C.L. Mohler, and L.R. Milbrath. 2010. Establishment of the
invasive perennial Vincetoxicum rossicum across a disturbance gradient in New York
State, USA. Plant Ecology 211:65–77.
Averill, K.M., A. DiTommaso, C.L. Mohler, and L.R. Milbrath. 2011. Survival, growth, and
fecundity of the invasive swallowworts (Vincetoxicum rossicum and V. nigrum) in New
York State. Invasive Plant Science and Management 4:198–206.
Burnside, O.C., C.R. Fenster, L.L. Evetts, and R.F. Mumm. 1981. Germination of exhumed
weed seed in Nebraska. Weed Science 29:577–586.
Caswell, H. 2001. Matrix Population Models: Construction, Analysis, and Interpretation.
Sinauer, Sunderland, MA. 722 pp.
DiTommaso, A., D.C. Brainard, and B.R. Webster. 2005a. Seed characteristics of the invasive
alien vine Vincetoxicum rossicum are affected by site, harvest date, and storage
duration. Canadian Journal of Botany 83:102–110.
DiTommaso, A., F.M. Lawlor, and S.J. Darbyshire. 2005b. The biology of invasive alien
plants in Canada.2. Cynanchum rossicum (Kleopow) Borhidi [= Vincetoxicum rossicum
(Kleopow) Barbar.] and Cynanchum louiseae (L.) Kartesz & Gandhi [= Vincetoxicum
nigrum (L.) Moench]. Canadian Journal of Plant Science 85:243–263.
DiTommaso, A., L.R. Milbrath, T. Bittner, and F.R. Wesley. 2013. Pale Swallowwort
(Vincetoxicum rossicum) response to cutting and herbicides. Invasive Plant Science and
Management 6:381–390.
Northeastern Naturalist Vol. 24, No. 1
L.R. Milbrath, A.S. Davis, and J. Biazzo
2017
53
Evans, J.A., A.S. Davis, S. Raghu, A. Ragavendran, D.A. Landis, and D.W. Schemske.
2012. The importance of space, time, and stochasticity to the demography and management
of Alliaria petiolata. Ecological Applications 22:1497–1511.
Hazlehurst, A.F., A.S. Weed, L. Tewksbury, and R.A. Casagrande. 2012. Host specificity
of Hypena opulenta: A potential biological control agent of Vincetoxicum in North
America. Environmental Entomology 41:841–848.
Hotchkiss, E.E., A. DiTommaso, D.C. Brainard, and C.L. Mohler. 2008. Survival and
performance of the invasive vine Vincetoxicum rossicum (Apocynaceae) from seeds of
different embryo number under two light environments. American Journal of Botany
95:447–453.
Ladd, D., and N. Cappuccino. 2005. A field study of seed dispersal and seedling performance
in the invasive exotic vine Vincetoxicum rossicum. Canadian Journal of Botany
83:1181–1188.
Lawlor, F.M. 2000. Herbicidal treatment of the invasive plant Cynanchum rossicum and
experimental post-control restoration of infested sites. M.Sc. Thesis. State University of
New York, Syracuse, NY. 78 pp.
Lindig-Cisneros, R., and J.B. Zedler. 2002. Phalaris arundinacea seedling establishment:
Effects of canopy complexity in fen, mesocosm, and restoration experiments. Canadian
Journal of Botany 80:617–624.
Magidow, L.C., A. DiTommaso, Q.M. Ketterings, C.L. Mohler, and L.R. Milbrath. 2013.
Emergence and performance of two invasive swallowworts (Vincetoxicum spp.) in contrasting
soil types and soil pH. Invasive Plant Science and Management 6:281–291.
McEvoy, P.B., and E.M. Coombs. 1999. Biological control of plant invaders: Regional
patterns, field experiments, and structured population-models. Ecological Applications
9:387–401.
Mervosh, T.L., and D. Gumbart. 2015. Cutting and herbicide treatments for control of Oriental
Bittersweet, Pale Swallow-wort and Morrow’s Honeysuckle. Natural Areas Journal
35:256–265.
Meyer, A.H., and B. Schmid. 1999. Seed dynamics and seedling establishment in the invading
perennial Solidago altissima under different experimental treatments. Journal of
Ecology 87:28–41.
Milbrath, L.R. 2008. Growth and reproduction of invasive Vincetoxicum rossicum and V. nigrum
under artificial defoliation and different light environments. Botany 86:1279–1290.
Milbrath, L.R., A. DiTommaso, J. Biazzo, and S.H. Morris. 2016. Tolerance of swallowworts
(Vincetoxicum spp.) to multiple years of artificial defoliation and clipping. Invasive
Plant Science and Management 9:1–11.
Shea, K., D. Kelly, A.W. Sheppard, and T.L. Woodburn. 2005. Context-dependent biological
control of an invasive thistle. Ecology 86:3174–3181.
Sheeley, S.E. 1992. Distribution and life history of Vincetoxicum rossicum (Asclepiadaceae):
An exotic plant in North America. M.Sc. Thesis. State University of New York,
Syracuse, NY. 126 pp.
Smith, L.L., A. DiTommaso, J. Lehmann, and S. Greipsson. 2006. Growth and reproductive
potential of the invasive exotic vine Vincetoxicum rossicum in northern New York State.
Canadian Journal of Botany 84:1771–1780.
Weed, A.S., A. Gassmann, A.M. Leroux, and R.A. Casagrande. 2011. Performance of
potential European biological control agents of Vincetoxicum spp. with notes on their
distribution. Journal of Applied Entomology 135:700–713.
Werner, P.A., and H. Caswell. 1977. Population growth rates and age versus stage-distribution
models for Teasel (Dipsacus sylvestris Huds.). Ecology 58:1103–1111.