Seed Biology of the Weed Maryland Meadowbeauty (Rhexia mariana L.) in Blueberry (Vaccinium spp.)

ABSTRACT

Studies were conducted at three locations across the North Carolina coastal plain region to determine sexual reproductive potential, seedbank density, frequency of seed dormancy, and effect of temperature on Maryland meadowbeauty seed germination. Seed capsule density ranged from 500 per m² to 1124 per m² across locations. Each capsule contained an average of 74 seeds. Seeds germinated only in the 35/20 °C temperature regime, which represents the hottest month (August) of the growing season in southeastern North Carolina. Total number of germinated seeds differed across locations and ranged from 30% to 57%. The percent (average) of viable (66) and nonviable (26) meadowbeauty seeds was not different across locations. Relative germination and seed dormancy percentages were calculated based on the number of viable seeds. The percent relative germination and seed dormancy were different across locations and ranged from 47% to 86% and 14% to 55%, respectively. The highest number of germination resulted from 90 days of stratification. On average, 27 seeds germinated among soil samples, which is equivalent to 989 seeds per m². The data indicate that the populations of meadowbeauty in blueberry fields have the potential to sexually reproduce and contribute 5 × 10⁶ to 1.1 × 10⁷ seed capsules/ha and 3.7 × 10⁸ to 8.3 × 10⁸ seeds/ha of infestation. Freshly mature seeds can germinate and contribute 1.79 × 10⁸ to 7.14 × 10⁸ seedlings/ha in the year the seeds are produced and 5.18 × 10⁷ to 4.4 × 10⁸ seeds/ha can be dormant and incorporated into the seed bank on an annual basis. Approximately 9.89 × 10⁶ seeds/ha are dormant and viable in the soil and have the potential to germinate following adequate stratification.

KEYWORDS:

• Seed germination
• seed dormancy
• seed bank
• tetrazolium testing

Introduction

North Carolina is the sixth largest producer of blueberries in the United States with approximately 6400 acres harvested in 2014 (USDA, 2014). Within the state, blueberries are the most lucrative fruit commodity with a value of $72 million (USDA, 2014). Although blueberry production is scattered across the state, the coastal plain of North Carolina accounts for approximately 55% of the state’s growers and 96% of the state’s total acreage (USDA, 2007). Within the coastal plain region of North Carolina, the southern coastal plain dominates production, accounting for 45% of the state’s growers and 93% of the state’s total acreage (USDA, 2007).

Weed communities observed in blueberry fields in the coastal plain of North Carolina appear to be dominated by perennial species. Pine (Pinus spp.), holly (Ilex spp.), greenbriar (Smilax spp.), small deciduous trees, and various species of herbaceous perennials consistently present weed control problems for growers. Maryland meadowbeauty (hereafter called meadowbeauty) is a herbaceous perennial weed in blueberries in North Carolina that reproduces both sexually and vegetatively (Coneybeer-Roberts et al., 2016; Craine, 2002). After a mature plant flowers in late summer, shoots are produced from shallow rhizomes and remain small until growth resumes in the spring (James, 1956). This ability enables the species to grow in high densities and spread aggressively in some conditions (Craine, 2002). Kral and Bostick (1969) found that large populations of meadowbeauty may be composed of only a few genetically distinct individuals.

The natural habitat of meadowbeauty is similar to that of blueberries and can be found in wet, open areas from Massachusetts to Florida, and as far west as southern Indiana, Missouri, and Texas (Gleason and Cronquist, 1991). In the Cape Cod area, meadowbeauty and the highbush blueberry (Vaccinium corymbosum L.) are commonly associated in natural communities that occur in acidic, nutrient-poor and wet soils. Meadowbeauty also depends on full sunlight for survival. In North Carolina, meadowbeauty is common in highly disturbed areas like roadside ditches, agricultural fields, and forest edges. Observations of meadowbeauty becoming aggressive and invasive have been noted in both disturbed and non-disturbed areas in Florida (Craine, 2002).

The impact of sexual reproduction on the meadowbeauty population in blueberry fields is not fully understood. The species produces seeds that are released from a capsule in an urn-shaped persistent hypanthium that is 9–11 mm long at maturity (Craine, 2002). Each seed is between 0.3 and 0.7 mm long and cochleate with papillae occurring in longitudinal rows (James, 1956). A seed bank study performed by Keddy and Reznicek (1982) found that the closely related Rhexia virginica L. contributed an estimated 900 seeds per m² to a lake shore seed bank.

Studies were conducted to determine the sexual reproductive potential, seedbank density, frequency of seed dormancy, and the effect of temperature on germination of meadowbeauty at three North Carolina blueberry farms.

Materials and methods

Sexual reproductive potential

Sexual reproduction potential of meadowbeauty was considered in these studies to consist of capsules and the seeds they contain. To determine the potential of meadowbeauty to produce capsules, meadowbeauty plants were collected from six 1-m² quadrats (replicates) each on three locations, which were blueberry farms (bushes in fields >3 years old and bearing) in the communities of Harrells (34° 45ʹ 59″ N, 78° 08ʹ 49″ W), Rowan (34° 37ʹ 33″ N, 78° 19ʹ 41″ W), and Burgaw (34° 36ʹ 17″ N, 77° 51ʹ 34″ W), North Carolina on 13 Aug. 2007. All plants with stems occurring in the quadrat were harvested and each quadrat stored individually. Seed capsules were removed from plants and counted. The capsules collected from these six 1-m² quadrants at each location were saved for use in the studies on temperature effects on seed germination, and relative seed germination and dormancy. To determine the mean number of seeds produced in each seed capsule, meadowbeauty seed capsules were collected from the same three locations as above on 13 Aug. 2007. Sampling was replicated three times at each location. A sampling replicate included 10 randomly selected intact seed capsules, which were collected and stored individually. Each seed capsule was crushed, and fully developed seeds were counted using a dissecting microscope. For the seed and capsule counts, a one-way analysis of variance (ANOVA), with replicates in locations as error, was carried out to compare locations using Proc Glm (SAS 9.1.3, SAS Institute Inc., Cary, NC, USA). Pairwise comparisons of location means were performed using Fisher’s protected Least Significant Difference (LSD, p = 0.05).

Temperature effects on seed germination

Seeds from capsules collected at the three blueberry farms (locations) on 13 Aug. 2007, as described previously, were used in studies to determine the temperature effects on seed germination.

Seeds were extracted by crushing the desiccated hypanthium and then separated from debris using sieves (#40 425-micrometer sieve and #50 300-micrometer sieve, Hoffman Manufacturing Inc., Jefferson, OR, USA). Sifted seeds were cleaned using forced air (South Dakota seed blower, Seedburo Equipment Co., Chicago, IL, USA) with sufficient air velocity to separate debris and empty seeds from intact seeds. Germination tests were performed at the North Carolina Department of Agriculture and Consumer Services Plant Industry Division Seed Lab in Raleigh, NC. Incubators were set at four temperature regimes that approximate maximum and minimum temperatures [20/20 °C (March), 25/15 °C (April/May), 30/20 °C (June/July), 35/20 °C (August)] during the growing season in eastern North Carolina. Maximum and minimum temperature regimes were 12 h each. Seeds were germinated in full light in plastic boxes on two sheets of filter paper (Baskin et al., 1999) saturated with either distilled water, or 0.2% KNO₃ [2 g KNO₃ in 1000 ml distilled water (Sigma American Chemical Society grade, Sigma-Aldrich Corporation, St. Louis, MO, USA) to reduce or overcome dormancy present in the seeds]. Thus, a complete replicate, or run, consisted of four incubators, one per temperature regime, with each incubator containing four boxes of 100 seeds for each location. The four boxes (two water, two KNO₃) for the three locations at a given temperature regime were arranged randomly within the incubator. Seedling counts were taken beginning 2 weeks after seeding and continued every 7 days until counts were suspended following two counts without new germination. Only those seedlings that had grown to the cotyledon stage and fully shed the seed coat were considered germinated. Germinated seedlings were removed from the boxes upon counting. Germination tests were replicated over time in triplicate (three runs), and duplicate boxes for water and KNO₃ treatments for a given site and temperature regime were treated as subsamples and averaged to simplify the ANOVA. Germination percentages at each date conformed to a split plot design with temperature as the main plot factor, location and seed treatment (water or KNO₃) as sub-plot factors, and runs as blocks. However, germination was negligible in three of the four temperature regimes and ANOVA was carried out on data for only one regime (separately for each date and total germination) with runs as blocks, and location, seed treatment (water or KNO₃), and location by water treatment interaction as fixed effects. Where interaction was not significant, comparisons of main effect means were performed using Fisher’s protected LSD (p = 0.05).

Relative seed germination and dormancy

Seeds from capsules collected at the three blueberry farms on 13 Aug. 2007, as described previously, were used in studies to determine the relative germination and dormancy of meadowbeauty. Seeds were extracted by crushing the desiccated hypanthium and then separated from debris using sieves (#40 425-micrometer sieve and #50 300-micrometer sieve (Hoffman Manufacturing Inc., Jefferson, OR, USA). Sifted seeds were cleaned using forced air (South Dakota seed blower, Seedburo Equipment Co., Chicago, IL, USA) with sufficient air velocity to separate debris and empty seeds from intact seeds.

Seed viability and relative germination and dormancy were estimated via testing with a tetrazolium chloride (TZ) staining technique. Tetrazolium testing of meadowbeauty seeds was performed at the North Carolina Department of Agriculture and Consumer Services Plant Industry Division Seed Lab in Raleigh, NC according to a protocol adapted from the Tetrazolium Testing Handbook from the Association of Official Seed Analysts (Peters, 2000). Seeds were imbibed in the dark for 18 h inside a water-saturated paper towel in an incubator set at 20/30 °C (9 h each temperature). Following imbibation, the funiculi-like structures that held the seeds to the ovary were removed from 100 seeds using a scalpel and a dissecting microscope. The seeds were placed in a 1% TZ solution for 18 h in a 45 °C incubator and then drained and placed in a clearing solution of glycerol. Seeds were separated based on color change under a dissecting scope into viable, dead, and empty seed. Tetrazolium testing was replicated over time in triplicate for each location. Two-way ANOVA, with factors replicate and location, was carried out on the percentages of viable, dead, and empty seed, followed by comparison of location means by Fisher’s protected LSD (p = 0.05; SAS 9.1.3, SAS Institute Inc., Cary, NC, USA). Relative percent germination was calculated using total germination from the 20/35 temperature regime in the seed germination study and seed viability, appropriately averaged, via the following equation (1):

(1)

Percent dormancy was calculated by subtracting the relative percent germination from 100. Relative germination data were subjected to the same analysis used for total germination.

Seed density in the seed bank

Soil was collected from the Rowan location, which has a history of severe meadowbeauty infestation. The top 2.5 cm of soil was collected from 50 randomly selected 1 m² quadrats. The soil collected from all quadrats was homogenized, sifted through a screen to remove large debris, and water was added to achieve consistent soil moisture prior to weighing. Soil was separated into 200-g samples and placed in plastic trays (Com-Packs bedding plant containers #601, 7” × 5-1/4”, Hummert International, St. Louis, MO, USA). A 200-g sample equated to 602 cm³ of soil with a surface area of 273 cm² in the tray. Two sheets of filter paper were placed in each tray prior to being filled with soil. The 200-g soil samples above were subject to treatments of no stratification, or stratification at 3 to 4 °C for 15, 30, 60, or 90 days (DOS), or stratification at 20 °C (±3.5) for 90 days in dark conditions in a refrigerated cooler. All stratification treatments were placed in the refrigerator and then removed when they reached the desired length of time for the treatment. Treatments were replicated six times. All 200-g samples were then placed according to a completely randomized design in a reach-in chamber (Saravitz et al., 2008) set at 12-h photoperiod and a corresponding 12-h temperature shift (35/20 °C). During germination procedures in the chambers, samples were watered twice a day to achieve soil saturation. The study was conducted twice (giving two runs over time) at the NC State University Phytotron. Samples were stirred every 3 weeks and seedling counts were taken every 7 days for 12 weeks following placement in the chamber. Seedlings were only counted if they reached the cotyledon stage and shed the seed coat. Counted seedlings were removed from the trays. Counts of seedlings for the six replicates per treatment at each week in each run were averaged to simplify the analysis. A repeated measures ANOVA was carried out on resulting averages with stratification treatment, week and interaction as fixed effects and run, run by treatments and run by week as random effects. Due to strong treatment by week interaction and variance heterogeneity by week, comparisons of treatment means were carried out in separate analyses by week, using Fisher’s protected LSD (p = 0.05).

Results and discussion

Sexual reproductive potential

Seed capsule density differed among locations (p = 0.015), ranging from 500 per m² at Burgaw to 1124 per m² at Rowan. The Harrells location had an intermediate density of 697 capsules per m². Meadowbeauty seeds per capsule averaged 74 seeds per capsule and did not differ across locations. Meadowbeauty demonstrated a projected sexual reproductive potential of 5 × 10⁶ to 1.1 × 10⁷ seed capsules per ha and 3.7 × 10⁸ to 8.3 × 10⁸ seeds per ha of infestation.

The Rowan location had the highest visible density of meadowbeauty compared to the other locations. Lower seed capsule density at the Rowan location may have been attributed to greater density of shoots and competition for associated resources (Begon et al., 2006). If resources are limited due to intra-specific competition among meadowbeauty in the Rowan population, those resources are likely unavailable for reproduction.

Temperature effects on seed germination

Meadowbeauty seeds germinated only in the 35/20 °C temperature regime, which represented the hottest month (August) of the growing season in southeastern North Carolina coastal plain. Thus, germination data is only presented for this regime. Germination occurred for three consecutive counts beginning 7 days after seeding (DAS) and counts were suspended at 35 days following 14 days without germination. Seed counts at 7, 14, and 21 DAS differed among locations (Table 1), while KNO₃ treatment did not have a significant effect on germination at any other time or location (data not shown). The total number of germinated seeds differed among locations (Table 1). Harrells and Burgaw had similar total germination at 30% and 36%, respectively. Rowan had the highest level of total germination at 57%.

Table 1. Termination (%) and relative germination (%) of meadowbeauty seeds from Harrells, Rowan, and Burgaw, North Carolina 7, 14, and 21 DAS^z and total germination exposed to 35/20 °C.

	Germination (DAS)				Relative germination
Location	7	14	21	Total	Total
	%				%
Harrells	15	13	1	29	45
Rowan	13	21	23	57	86
Burgaw	25	10	1	36	55
LSD (0.05)	9	NS^z	4	10	16

Note. ^zAbbreviations: DAS—days after seeding. LSD—least significant difference. NS—F test for location effect not significant.

The observation that germination only occurred in the temperature regime of 35/20 °C supports previous observations on the closely related variety R. mariana var. interior, in which highest levels of germination occurred under the same temperature regime (Baskin et al., 1999). However, the germination rates of freshly matured meadowbeauty seeds from North Carolina are higher than those observed in R. mariana var. interior collected from Tennessee (Baskin et al., 1999). The difference in percent germination between North Carolina and Tennessee seeds may be the result of ecotypic or climatic adaptations of meadowbeauty. Winter temperatures in the southeastern coastal plain of North Carolina are generally milder than those temperatures in Tennessee where the seeds were collected. Meadowbeauty seeds from the North Carolina coastal plain would not experience the same intensity and duration of temperatures required for stratification and may have adapted for less dormancy in freshly mature seeds.

Relative germination and dormancy

The percent of viable meadowbeauty seeds was not different among locations (averaged 66; data not shown). Likewise, nonviable (averaged 26) was not different across locations. Information on viable seeds allowed for the calculation of relative germination and dormancy of meadowbeauty seeds. Percent germination was previously based on 100 mature, cleaned seeds. If only 66 of those seeds were viable on average, then germination percentages must be calculated based on the number of viable seeds.

The percent relative seed germination and dormancy differed among locations (Table 1). The Rowan location had 86% relative germination and 14% relative dormancy compared to 45% and 55% relative germination and 55% and 45% relative dormancy at Harrells and Burgaw, respectively. Germination may be higher at Rowan due to a combination of genetic and environmental effects that may be specific to this meadowbeauty population and location.

Seed density in the seed bank

Meadowbeauty seeds germinated in all treatments. Total mean germination of meadowbeauty seeds was highest in the 90 DOS treatment during the second week of counts (Table 2). The data indicate that 90 DOS resulted in the highest amount of germination, suggesting that the meadowbeauty seeds from NC blueberry farms experience some level of dormancy that is overcome by a period of moisture and exposure to cold temperatures of greater than 60 d but less than 90 d. Those seeds that experience dormancy may remain viable in the seed bank and contribute to future infestations of meadowbeauty.

Table 2. Mean germination of soil germinated meadowbeauty seeds at each weekly count separated by stratification treatment.

	Meadowbeauty germination
	Week
	1	2	3	4	5	6	7	8	9	10	11	12
Treatment^z	Mean^y
No stratification	0	1	1	2	6	0	2	1	1	1	2	2
15 DOS^x	1	1	0	0	0	1	1	0	1	3	1	2
30 DOS	8	0	0	2	1	1	0	0	1	8	0	0
60 DOS	1	2	0	0	4	2	5	0	1	7	1	3
90 DOS	4	17	11	11	0	0	15	0	0	0	0	2
90 d RT	0	9	0	3	0	1	6	0	0	0	1	4
LSD (0.05)	NS^x	9	3	4	6	NS	3	NA	NS	4	NS	NS

Note. ^zTreatments: No stratification, germinate immediately; 15 DOS, germinate following 15 DOS at 3 to 4 °C; 30 DOS, germinate following 30 DOS at 3 to 4 °C; 60 DOS, germinate following 60 DOS at 3 to 4 °C; 90 DOS, germinate following 90 DOS at 3 to 4 °C; 90 days RT, germinate following 90 days at room temperature (20 °C).

^yMean: mean germination of meadowbeauty seeds averaged across runs and replications within runs for each weekly count.

^xAbbreviations: DOS—days of stratification; NS—F test for treatment is not significant; NA—Anova not appropriate because most counts were 0.

Those counts that followed stirring of the soil samples (counts 4, 7, and 10) had the next highest levels of germination, providing further support that meadowbeauty seeds require exposure to light for germination (Baskin et al., 1999). Counts were different between treatments during weeks 2, 3, 4, 5, 7, 8, and 10 (Table 2). A mean of 27 seedlings germinated in each sample of soil, which is equivalent to 989 seeds per m². The number of meadowbeauty seeds that germinated from the soil collected in North Carolina appears to be similar to the number of seeds of Rhexia virginica that germinated in a lake shore seed bank study (Keddy and Reznicek, 1982).

Overall, the data indicate that the populations of meadowbeauty in blueberry fields contributed 5 × 10⁶ to 1.1 × 10⁷ seed capsule per /ha and 3.7 × 10⁸ to 8.3 × 10⁸ seeds per ha of infestation. Relative germination ranged from 47% to 86%, indicating that freshly mature seeds have the potential to germinate and contribute 1.79 × 10⁸ to 7.14 × 10⁸ seedlings per ha in the year the seeds are produced. Dormancy ranged from 14% to 53%, indicating that 5.18 × 10⁷ to 4.4 × 10⁸ seeds per ha can be incorporated into the seed bank on an annual basis to await natural stratification and future germination. The presence of 989 seeds per m² in the soil seed bank indicates that approximately 9.89 × 10⁶ seeds per ha are dormant and viable in the soil and have the potential to germinate following adequate stratification.

The ability of meadowbeauty to contribute large numbers of seeds capable of immediate germination and viable dormant seeds to the soil seed bank creates a challenge for growers attempting to control the species. Not only must weed management strategies target the herbaceous perennial mother plants, but these strategies must also target the seed bank that may contribute indefinitely to future re-infestations of meadowbeauty in grower fields. Weed management measures could include mowing or application of herbicides to mother plants to prevent flowering or formation of seeds. Management of the soil seed bank could include the use of PRE herbicides to prevent seedling establishment as well as shallow cultivation or mulch to bury the small meadowbeauty seeds, which cannot germinate in darkness.

Acknowledgments

The authors wish to thank the North Carolina blueberry growers for their assistance and Dr. Cavell Brownie, Statistician and Emeritus Professor (1982–2007), NC State University Statistics Department, Raleigh, NC, for her assistance.

Funding

This research was supported in part by a grant from the North Carolina Blueberry Council, Inc.

Additional information

Funding

This research was supported in part by a grant from the North Carolina Blueberry Council, Inc.