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.