Printer Friendly

Pollination biology and seed production of a federally endangered perennial, Echinacea laevigata (asteraceae:heliantheae).

INTRODUCTION

The Piedmont Physiographic Region of the southeastern United States compared with the Coastal Plain Region (Fenneman, 1938), supports relatively few rare plant species (Walker, 1993; Estill and Cruzan, 2001; LeBlond, 2001; Sortie and Weakley, 2001, 2006) and, consequently, has received less attention from conservation botanists. Echinacea laevigata (Boyton and Beadle) Blake (coneflower) is a federal rare species of the Piedmont (Murdock, 1992) and is included in the Center for Plant Conservation's National Collection of Endangered Plants (Urbatsch et al., 2006). Most coneflower populations are small (<50 plants) and over half are declining in numbers (Murdock, 1995).

Managing rare plant populations is facilitated by knowledge of species' reproductive biology (Ratsirarson and Silander, 1996; Kaye, 1999; Timmerman-Erskine and Boyd, 1999; Anderson et al., 2001; Marten and M. Quesada, 2001). Management of small populations, in particular, must consider potential threats to population survival including relatively low seed production and seed germination rates and small seedlings. These reduced reproductive parameters can be caused by sub-optimal plant--pollinator relationships (Sih and Baltus, 1987; Jennersten, 1988; Steffan-Dewenter and Tscharntke, 1999; Suzuki, 2000; Fausto et al., 2001), limiting habitat resources (Horvitz and Schemske, 1988; Lamont et al., 1993; Bosch and Waser, 1999; Morgan, 1999; Cunningham, 2000; Muotoz and Arroyo, 2006; Ne'eman et al., 2006; Farji-Brener and Ghermandi, 2008; Tsaliki and Diekmann, 2009), and genetic mechanisms including local loss of S alleles (Byers and Meagher, 1992; Shore, 1993; Reinartz and Les, 1994) and inbreeding (Menges, 1991, 1995; Young et al., 1996; Fischer and Matthies, 1998; Bruna, 2002; Donaldson et al., 2002; Hooftman et al., 2003; Severns, 2003; Hensen and Oberpreiler, 2005; Kolb, 2005; Ward and Johnson, 2005).

Previous studies of Echinacea laevigata (Edwards and Madsen, 1993; Apsit and Dixon, 2001; Philippi et al., 2001; Alley and Affolter, 2004; Alley et al., 2005; Peters et al., 2009) contribute to the conservation of this species. The study reported here of pollination and reproductive success in both large and small populations will further support coneflower conservation by addressing a primary threat to this rare species, small population size. The goals of the current coneflower study are to (1) describe flowering phenology and diversity of flower visitors in small and large populations and estimate the importance of flower visitor groups as pollinators, (2) compare pollen viability among populations, test for pollen limitation, and compare seed production in outcross, selfed, and autogeitonogamy pollen treatments, (3) determine whether soil cations limit seed production, and (4) compare seed germination rate and seedling size in a small and large population.

METHODS

SPECIES AND STUDY POPULATIONS

Echinacea laevigata is an herbaceous perennial with short rhizomes that provide limited vegetative reproduction (North Carolina Natural Heritage Program, 2001). Most plants produce a single flowering head on a peduncle that can be up to 40 cm long (Murdock, 1995; Urbatsch et al., 2006). Disc flowers develop in a series of concentric rings, one ring opening daily beginning with the outermost and progressing inward (Wagenius, 2004). The stigma of a disc flower becomes receptive one day after the anthers dehisce. It takes 1012 days for all disc flowers of most individual heads to open. Various bees, beetles and butterflies have been reported visiting coneflower in South Carolina (Edwards and Madsen, 1993), but the importance of these insects as pollinators has, apparently, not been documented.

Echinacea laevigata likely occurred in Piedmont prairies which formerly were common in the southeastern United States but have become nearly extirpated due to fire suppression and anthropogenic habitat conversion (Murdock, 1995; Juras, 1997). In North Carolina, coneflower occurs only in Durham, Granville, Mecklenburg and Rockingham Counties (Buchanan and Finnegan, 2010; Weakley, 2010). This study was conducted during 2004 and 2005 and included a large population in a managed powerline fight-of-way in Granville County, Picture Creek (PC), and five small populations in Granville and Durham Counties, North Carolina (Table 1). Each population was separated from its nearest neighbor population by at least 0.4 km in a matrix of forest habitat and suburban development. The small Knap of Reeds (KOR) and Snow Hill Road (SHR) populations were lost to the study when land managers burned them and the surrounding habitat during spring, 2005. Three other small populations, Freudenberg (FB), Briardale (BD), and Lakeside (LS), were added to the study during 2005. PC included more than 1000 times the number of flowering plants and [greater than or equal to]9 times the daily flowering stem density than the small populations (Table 1).

FLOWERING PHENOLOGY AND FLOWER VISITORS

Flowering phenology.--To assess the likelihood that plant species were competing with E. laevigata for pollinators or facilitating its pollination at PC, we monitored community flowering phenology during 2004. On seven days beginning the third week of May until the fourth week of Jul., approximately weekly, we counted flowering individuals of all non-grass, sedge, or rush species in ten randomly distributed 2 x 10 m transects. We also recorded the blooming period for E. laevigata in each small population. Plant nomenclature followed the PLANTS Database (United States Department of Agriculture, Natural Resources Conservation Service, 2006).

Flower visitors.--Because bees have been reported as common flower visitors to other species of Echinacea (Edwards and Madsen, 1993; Leuszler et al., 1996) and because many bees are generalist flower visitors (Mitchell, 1962), we determined which plant species bees visited at PC during 2004. On seven days beginning the third week of May until the fourth week of Jul., approximately weekly, we observed bee visits during slow walks throughout the length of each 2 x 10 m transect during 900-1100, the peak of bee activity.

To determine how important different groups of insect flower visitors were as coneflower pollinators (Wiggam and Ferguson, 2005), we observed frequencies of visits to coneflower heads and determined quantitative features of pollen loads on individual visitors (Herrera, 1987). During the second--fourth weeks of Jun. of each study year at PC, we counted flower visits and captured visitors in eight randomly distributed 2 X 2 m plots. We counted visits during 41 15-min periods in 2004 and 36 periods in 2005. We limited to [less than or equal to] 8 the number of observation periods per day to minimize undue influence of data from a single day. Fifty-seven observation periods were between 0900 and 1300 and 20 were between 1300 and 1600. Before each observation period in each plot, we counted the blooming coneflower heads so we could express visits on a per head basis. We tested for a pollinator group effect on number of visits per head per 15 min with 1-way ANOVA using SAS PROC GLM (SAS Institute Inc., 2008). We square root transformed data and pooled it over observation plots, days, and years. Differences among pollinator groups were determined with Tukey's studentized range test at an alpha level of 0.05.

Based on observations during the first two weeks of Jun. 2004, the most common flower visitors were Bombus spp. (Apidae; bumblebees), Megachile spp. (Megachilidae, leaf cutter bees), Xylocopa virginica L.(Apidae; carpenter bee), Lygaeus kalmii Stal (Lygaeidae, seed bug), Hesperiidae (skippers), and Nymphalidae (brush-footed butterflies). On 17-18 Jun. 2004 and 8-16 Jun. 2005 we captured visitors in the 2 x 2 m plots. We collected 13 individuals each of Bombus and Megachile and five individuals each of Xylocopa virginica, Lygaeus, Hesperiidae (skippers) and Nymphalidae and determined the distribution of pollen on the body of each at 10-20 X. We assigned a score of 0-4 to each individual based on the location of pollen on its body which reflected the probability of the pollen being deposited on stigmas (Table 3). We analyzed pollen distribution data with the Kruskal-Wallis test for ranked data (Sokal and Rohlf, 1969). Critical [chi square] value was determined in Rohlf and Sokal (1969).

We estimated the number of Echinacea laevigata pollen grains and its proportion in the total pollen load for each captured insect. We placed each insect in 3 ml of detergent solution in a vial and shook the vial until all pollen appeared to be removed from the insect. We removed pollen sacs from the hind legs of Hymenopterans prior to putting them in the detergent solution because this pollen would be unavailable to stigmas. For each individual, we then examined 10 0.05 ml pollen suspension samples at 100 X and counted E. laevigata and non--E. laevigata grains. The mean of the 10 samples per individual was determined, [mean.sub.individual]. For each insect group, we determined the [mean.sub.group], average of the [means.sub.individual], and its confidence interval (Table 3). We tested for a pollinator group effect on number of E. laevigata pollen grains per 0.05 ml sample ([mean.sub.group]) and on proportion of E. laevigata grains per sample with 1-way ANOVA using SAS PROC GLM (SAS Institute Inc., 2008). Data were pooled over observation plots, days, and years. Pollen counts were square root transformed and proportional data were arcsine transformed for these analyses. Pollinator group differences were determined with Tukey's studentized range test at an alpha level of 0.05.

We determined the relative coneflower pollinator importance for each insect visitor group as the sum of the ranks for the four component criteria. We also described the behavior of flower visitors while on coneflower heads, but we did not incorporate this behavior in the linear combination of pollination variables.

Number of insect flower visits and visitor species richness may be reduced in small populations relative to large populations (Sih and Baltus, 1987; Jennersten, 1988; Steffan-Dewenter and Tscharntke, 1999; Suzuki, 2000; Fausto et al., 2001). To determine if this was the case in the large and small coneflower populations, we observed insect flower visitors during the second--fourth weeks of Jun. 2004 and 2005. We counted visits by each insect group during 15 min observation periods. We used the insect counts from PC that were also used to determine pollinator importance. We did not use multiple sampling plots in small populations because they were composed of so few plants. Instead, we counted visits from one vantage point. Number of daily observation periods in the small populations was limited by the small number of flowering heads and their irregular daily occurrence.

For each year, insect visit counts and species richness were square root transformed and then modeled through mixed effects linear model with site as fixed effect and days within site as random effect using SAS PROC MIXED (SAS Institute Inc., 2008). Day within site variance was estimated separately for each site. Differences among sites were determined with Tukey's studentized range test at an alpha level of 0.05.

FACTORS AFFECTING SEED PRODUCTION

Pollen viability.--Low pollen viability has been associated with reduced seed production (Demchik and Day, 1996; Clevenger et al., 2004; Prasad et al., 2006). To estimate pollen viability in the study populations, we determined the % of pollen that stained dark blue in cotton blue-lactophenol, a correlate of pollen viability (Kearns and Inuoye, 1993), during 79 Jun. 2004, for PC, SHR, KOR and during 6-9 Jun. 2005, for BD, FB, and LS. We sampled pollen from eight randomly selected flowering heads at PC and from three randomly selected heads at each of the small populations. We collected anthers from four disk flowers of the day from each head, macerated them together in a drop of stain on a microscope slide and counted presumably viable (dark blue and unshriveled) and non-viable (light blue and/or shriveled) grains at 100 X.

Compatibility.--Seed production by self-incompatible plant species is highly dependent on pollinator activity while that of self-compatible species is largely independent of pollinators. To confirm our suspicion that coneflower is self-incompatible, we compared seed production following three bagging/pollination treatments, bagged heads/no artificial pollination (testing auto-geitonogamy combined with genetic self-compatibility), bagged/ artificial geitonogamy (testing genetic self-compatibility), and unbagged heads (testing combined effects of geitonogamy and cross pollinations). We applied these treatments in PC throughout the second-third weeks of Jun. 2004. Substantial seed production in all three treatments would suggest apomixis.

For the bagged/no artificial pollination treatment, we bagged 10 heads randomly selected from the central part of the population. We did not manipulate these heads until their contents were harvested in Sep. We randomly selected 10 heads from the central part of the population for the bagged/artificial geitonogamy treatment. When the stigmas of the outer ring of disk flowers in each head were receptive, we began geitonogamy pollinations by gently transferring pollen to stigmas with toothpicks. We reduced the possibility of experimental manipulation effects by minimizing stigma--toothpick tip contact. We repeatedly pollinated each head daily until all of its disk flowers had been pollinated, rebagging the head after each pollination. Immediately after artificial pollinations were completed during the third week of Jun., we randomly selected and bagged 24 heads from among those in the central part of the population that had partly fresh ray flowers, indicating that they had recently completed flowering. These heads were the unbagged treatment. We harvested the 24 heads of the unbagged treatment on 2 Sep. 2004, and the 20 heads of the two bagged treatments on 23 Sep. 2004. We counted plump and non-plump seeds from each head. We also looked for indications of seed herbivory in the harvested heads. To determine if pollination treatment affected the frequency of heads that produced seeds, we performed Fisher's Exact Test using SAS PROC FREQ (SAS Institute Inc., 2008). We also tested for an effect of pollination treatment on numbers of seeds per seed-producing head with a 1-way ANOVA using SAS PROC GLM (SAS Institute Inc., 2008) and untransformed data. Differences among treatments were determined with Tukey's studentized range test at an alpha level of 0.05.

Pollen limitation.--Apart from possible effects of pollen viability and compatibility, pollinator activity may limit seed production by failing to deposit on stigmas that amount of pollen required for maximal seed production. This phenomenon, pollen limitation, has frequently been noted in small populations. To determine if pollen limitation was affecting the coneflower populations, we compared seed production following open-pollination (heads unmanipulated) and cross-pollination (unbagged heads receiving artificial crosspollination) treatments. During the second week of Jun. of both 2004 and 2005, we randomly selected heads which were near anthesis and which appeared equal in size for the two pollination treatments in each population. The number of heads in the small populations was limited by the low number of heads in these populations. In each population, the open-pollination heads remained unmanipulated until they were bagged during the second week of Jul. We artificially cross-pollinated heads throughout the second and third weeks of Jun. of both 2004 and 2005. In PC we cross-pollinated each of 10 heads daily throughout its flowering duration leaving it unbagged between pollinations. For each daily cross-pollination, we used pollen from three plants that were randomly selected from 3-5 m from the maternal plant. We reduced the possibility of experimental manipulation effects by minimizing stigma--toothpick tip contact. In each small population, the number of cross-pollinated heads was <10 because there were so few heads in these populations, SHR (n = 4), KOR (n = 4), BD (n = 8), LS (n = 4), FD (n = 6). Also, the small number of plants flowering daily in these populations limited the number of pollen donors to one on some days and two on other days. We used toothpicks for transferring pollen. During the second week of Jul. of 2004 and 2005, we bagged the cross-pollinated and open-pollinated heads in each population. We harvested these heads during the first week of Sep. of both 2004 and 2005. We determined seed production % for each head by counting plump and shriveled seeds. We had previously determined that plump seeds sliced open and soaked in a 0.1% solution of 2,3,5 triphenyltetrazolium chloride (Kearns and Inouye, 1993) turned pink, indicating viability, while no shriveled seeds turned pink.

For each year and pollination treatment combination, we tested for a population effect on seed production. Also, for each population, we tested for a pollination treatment effect on seed production. We did each of these tests with 1-way ANOVA using SAS PROC GLM. Population and pollination treatment differences were determined with Tukey's studentized range test at an alpha level of 0.05. (SAS Institute Inc., 2008).

Nutrients.--Bosch and Waser (1999) suggested resource limitation as a mechanism of reduced seed production in local populations. A positive correlation between soil nutrient cation amount and seed production would suggest that cations might be limiting seed production. We estimated amounts of [Ca.sup.++], [K.sup.+], and [Mg.sup.++] in the soil for each population. From each population, we collected five soil subsamples from the top 25 mm of the soil profile (plant root zone). Each subsample location was within 1 m ofa coneflower plant and [greater than or equal to]3 m distant from other subsample locations. We combined the five subsamples in a plastic pail, air dried it, pulverized it by hand, and passed it through a 2 mm sieve. The North Carolina Department of Agriculture and Consumer Services, Agronomic Division--Soil Testing Section analyzed the soil samples. They determined the pH of a 1:1 soil:distilled water suspension, the quantities of available calcium, magnesium, and potassium using a Mehlich-3 extraction (Mehlich, 1984), and then calculated cation exchange capacity (CEC) and % base saturation (Mehlich, 1976). We used correlation analysis (Excel Data Analysis Tools, Microsoft Office Corporation, 2003) to assess the relationship between cation amount and seed production following open-pollination (see methods under Flower visitation and pollen limitation below).

We also explored plant size and seed production data for patterns suggesting resource limitation (Griffin and Barrett, 2002). We tested for plant size difference between PC and the combined small populations. The number of disk flowers per cross-pollination head (see methods under Flower visitation and pollen limitation below) was the measure of plant size (Wagenius, 2004). These disk flower numbers were square root transformed for analysis. We pooled data of small populations from both years and PC data from both years. We also tested for plant size difference between pollination treatments (cross- and open-) in each population. For both of the preceding tests, we used a linear model, 1-way ANOVA, with SAS PROC GLM (SAS Institute Inc., 2008). Finally, we assessed the relationship between plant size and seed production of cross-pollination heads for both PC and the small populations. We pooled PC data of 2004 and 2005 and pooled data of all small populations from both years. We used correlation analysis (Excel Data Analysis Tools, Microsoft Office Corporation, 2003) and determined P values from Table Y of Rohlf and Sokal (1969).

PERFORMANCE OF SEEDS AND SEEDLINGS FROM A LARGE AND A SMALL POPULATION

A direct relationship between population size and seed and seedling performance has been been described for several plant taxa (Menges, 1991, 1995; Fischer and Matthies, 1998; Bruna, 2002; Donaldson et al., 2002; Hooftman et al., 2003; Kolb, 2005). We looked for differences in coneflower seed performance and seedling size between the large PC population and the small SHR population. During the first week of Sep. 2004, we harvested randomly selected open-pollinated heads from PC and available heads from SHR and selected the plump seeds from each head. We pooled the plump seeds of each population and moist-stratified them (4 C for 50 d). On 4 Mar. 2005, we planted them at PC in 10 4 X 2 m field blocks at a depth of 0.1-0.2 mm. To minimize the chances of coneflower seedlings emerging from a soil seed bank, we located the blocks approximately 100 m from the edge of the coneflower population. Prior to planting, we cleared vegetation from the blocks with herbicide and clipping. Within each block, we planted 20 PC seeds in one 0.5 X 0.5 m plot, 20 SHR seeds in another 0.5 x 0.5 m plot and left the third plot unplanted as a control. We recorded numbers of new seedlings (seed germinations) and seedling deaths in each plot weekly from 11 Mar. to 29 Jul. On 19 Aug. we counted the leaves on each surviving seedling.

We tested for a population effect on field seed germination and seedling survivorship with a randomized block ANOVA using SAS PROC ANOVA (SAS Institute Inc., 2008). Data for these analyses were arc sine transformed. We tested for a population effect on seedling leaf number with a mixed model ANOVA using SAS PROC MIXED (SAS Institute Inc., 2008). Leaf numbers were square root transformed for this analysis. Means separation for these analyses was determined with Tukey's studentized range test at an alpha level of 0.05.

[FIGURE 1 OMITTED]

On 5 Mar. 2005, we planted 10 open-pollination seeds from PC at a depth of 0.1-0.2 mm in each of 20 greenhouse pots containing 4 : 1 loam : sand substrate. We planted 10 seeds from SHR in 20 other pots containing the same substrate. Plants were maintained on a greenhouse bench, lightly watered to maintain a moist substrate surface, and rotated weekly to avoid position effects. Greenhouse temperatures ranged from 23 C to 29 C. We recorded numbers of germinations and seedling deaths for each pot weekly until no more germination occurred for two weeks in a row. Then all seedlings except the largest in each pot were removed. We harvested the seedlings on 22-24 Aug. 2005, dried them at 43 C to constant weight and recorded their dry weight.

We tested for a population effect on greenhouse seed germination, seedling survivorship and seedling dry weight with 1-way ANOVA using SAS PROC GLM. For these analyses, germination and survivorship data were arc sine transformed; dry weight data was untransformed. Means separation for these analyses was determined with Tukey's studentized range test at an alpha level of 0.05.

RESULTS

FLOWERING PHENOLOGY AND FLOWER VISITORS

Flowering phenology.--Coneflower was first seen flowering at PC on 25 May 2004, and by 3 Jun. it was the most abundant blooming species in the community (Fig. 1). It remained most abundant throughout Jun. At its blooming peak, coneflower accounted for 68% of the flowering individuals in the community. Blephila ciliata (Linnaeus) Bentham, Penstemon australis Small, and Rosa caroliniana Linnaeus bloomed along with coneflower during late May-early Jun. and Houstonia longifolia Gaertner did so during early Jun.--early Jul. In 2004, Echinacea laevigata flowered at SHR and KOR from 4 to 26 Jun. and 4 to 28 Jun., respectively. In 2005, flowering occurred at BD, LS, and FB from 3 to 23 Jun., 5 to 30 Jun., and 6 to 27 Jun., respectively. Blephila ciliata, P. australis, R. caroliniana, and H. longifolia did not occur with any of the small coneflower populations.

Flower Visitors.--We saw bees visiting coneflower in the PC study transects during 25 May-6 Jul. Bees visited Blephila ciliata, Penstemon australis, and Rosa caroliniana until 3 Jun., but not later. We did not see bees visit Houstonia longifolia. Bees visited one plant of Cirsium caroliniana (Walter) Fernald and B.G. Schubert located outside of the study transects on 3 Jun., but not later. Bees began visiting Centrosema virginiana (Linnaeus) Bentham and Liatris squarrosa (Linnaeus) Michaux in Jul. after coneflower had completed flowering.

The order of coneflower pollinator importance was Bombus > Megachile = Xylocopa = Hesperiidae > Lygaeus = Nymphalidae. Herrera (1987) also found that butterflies were relatively unimportant pollinators compared to various bees. Bumblebees ranked higher than all other insect groups for number of visits to coneflower (Table 3). They accounted for 73% of all visits during 2004 and 93% during 2005. Additionally, individual bumblebee visitors carried nearly as many Echinacea laevigata pollen grains as carpenter bees (Table 3). There were no differences among pollinator groups for proportion of coneflower pollen in pollen load or pollen location on insect body. All bee and seed bug visitors crawled over flower heads and contacted receptive stigmas with their mouth parts, ventral thorax, abdomen, and legs. Skippers and butterflies remained stationary while they probed disk flowers with their proboscides. Other insects which visited coneflower occasionally included Hymenopterans Apis meUifera L. (Apidae), several Halictidae species, and Ammophila sp. Dahlbom (Sphecidae); Coleopterans, Brachyloptura vegans Olivier and Typocerus zebra Olivier (Cerambycidae); Lepidopterans, Eusarca confusaria Hubner (Geometridae) and Harrisina sp. (Zygaenidae); and Dipterans, Eristalis transverses Wiedemann, Milesia virgineinsis Drury, and Toxomerus sp. (Syrphidae).

During 2004, insects visited heads in PC five and 16 times as often as in KOR and SHR, respectively ([F.sub.2,17] = 11.93, P < 0.001, Table 4). However, during 2005, insects visited heads in FB more often than in PC ([F.sub.3,21] = 27.22, P < 0.001, Table 4). We often saw Bombus pollinators emerging from the ground, apparently from a nest, at the edge of the FB population. Insects visited heads in PC three and four times as often as in LS and BD, respectively. During both study years, insect visitor species richness was higher in PC than in each small population ([F.sub.2,17] = 18.51, P < 0.0001 and [F.sub.3.21] = 9.05, P = 0.001).

FACTORS AFFECTING SEED PRODUCTION

Pollen viability.--Pollen % viability for PC was 94.6 [+ or -] 1.19 (n = 8); SHR, 98.3 [+ or -] 0.58 (n = 3); KOR, 93.2 [+ or -] 1.53 (n = 3); BD, 98.7 [+ or -] 1.15 (n = 3); LS, 97.7 [+ or -] 0.58 (n = 3) and FB, 99.3 [+ or -] 0.58 (n = 3). These data suggest that low pollen viability did not limit coneflower seed production.

Compatibility.--All 24 of the unbagged, open-pollinated heads produced seeds, but only 20% (2/10) of the heads receiving either self pollen or no pollen produced seeds. The probability of seed production depended on pollination treatment (P < 0.001). Comparing those heads that did produce seeds, % production was greater ([F.sub.2.25] = 13.1, P < 0.001) for open-pollinated heads, 42.0 [+ or -] 15.7 (n = 24), than those receiving self pollen, 2.0 [+ or -] 0.19 (n = 2), or no pollen, 4.3 [+ or -] 0.31 (n = 2). We saw no evidence of seed herbivory. We concluded that coneflower is nearly exclusively outcrossing but occasionally produces a few seeds via geitonogamy. These results did not indicate apomictic seed production. This compatibility system is like that previously reported for other Echinacea species (McGregor, 1968; Parrish and Bazzaz, 1979; Leuszler et al., 1996; Wagenius, 2004, 2006).

Pollen limitation.--Seed production patterns differed between years. During 2004, there was no difference for either open-pollination or cross-pollination between PC and the small populations ([F.sub.2,24] = 1.90, P = 0.16 and [F.sub.1,10] = 1.36, P = 0.27, Table 5). However, open-pollination seed production in PC exceeded that in all of the small populations during 2005 ([F.sub.3,35] = 17.69, P < 0.001). Seed set was lower in BD than in LS and FB. Cross-pollination seed set also was lowest in BD ([F.sub.3,16] = 3.39, P = 0.04, Table 5). We concluded that seed production during 2005 was limited in the small populations relative to PC and this limitation was strongest in BD.

In PC during 2005, the relatively low cross-pollination seed production was unexpected given relatively high open-pollination seed set in this population (Table 5). Since we barely touched stigmas when we cross-pollinated flowers, it seems unlikely that stigma damage caused this result. However, the cross-pollinated heads had only 67% the number of disk flowers as open-pollinated heads. It is possible that these smaller plants were less reproductively fit than the larger open-pollinated plants. Because of this equivocal result, we do not make inferences based on comparisons of cross-pollination seed production between PC and LS or FD. However, we feel justified in concluding that cross-pollination seed production was higher in PC than in BD.

In the small populations of both 2004 and 2005, cross-pollination failed to increase seed production relative to open-pollination (Table 5). This indicated that pollinators were realizing the seed producing potential of these populations; seed production was not pollen-limited. The seed production limitation in small populations relative to PC during 2005 was caused by a mechanism(s) other than pollen limitation.

Nutrients.--Each cation, [Ca.sup.++], [K.sup.+], and [Mg.sup.++], was present in appreciable amounts in soils supporting both the large and small coneflower populations (Table 2). Cation exchange capacities of all these soils were near the maximum for North Carolina soils and base saturations were high (Hardy et al., 2003). These high base conditions were comparable with those previously reported for soils supporting Echinacea laevigata (Dayton, 1966). There was not a relationship between total soil cation amount and open-pollination seed production ([r.sub.5] = 0.612, P = 0.14; Table 2).

There was no difference in plant size between PC and the pooled small populations ([F.sub.1,30] = 0.402, P = 0.53). Plant size in open- and cross-pollination treatments differed in only PC during 2005 ([F.sub.l,16] = 5.195, P = 0.04). In this population, cross-pollinated heads had only 67% the number of disk flowers, suggesting that they could be less reproductively fit than the open-pollination heads. There was no relationship between plant size and seed production in the cross-pollination treatment in PC ([r.sub.16] = 0.195, P = 0.44) or in the pooled small populations ([r.sub.12] = 0.342, P = 0.23). These results suggested that, with the possible exception of PC during 2005, resources that regulate plant size did not limit seed production in the study populations.

PERFORMANCE OF SEEDS AND SEEDLINGS FROM A LARGE AND A SMALL POPULATION

Field seed germination for SHR and PC did not differ ([F.sub.1,9] = 4.61, P = 0.06, Table 6), but seedling survivorship was greater for SHR ([F.sub.1,9] = 5.11, P = 0.05). On average, each fieldraised PC seedling had 25% more leaves than each SHR seedling ([F.sub.1,19] = 8.50, P = 0.02). SHR and PC greenhouse germination ([F.sub.1,38] = 0.88, P = 0.35) and survivorship ([F.sub.1,38]= 3.04, P = 0.09) did not differ. Individual Picture Creek seedlings weighed 52% more than those from SHR ([F.sub.1,36] = 22.89, P < 0.001).

DISCUSSION

Flowering phenology and flower visitors.--The most important coneflower pollinators we saw were species of bumblebees that have been reported to visit more than 75 plant genera (Mitchell, 1962). Additionally, Bombus foragers often visit several plant species during individual foraging flights (Mosquin, 1971; Kunin, 1993); they are generalist flower visitors. In the current study, few plant species that attracted bumblebees flowered at the same time as coneflower. However, at other locations, co-flowering species may compete with coneflower for Bombus pollinators (Mosquin, 1971; Gross and Werner, 1983; Campbell, 1985; Jennersten and Nilsson, 1993; Kunin, 1993; Ramsey, 1995) or facilitate its pollination (Feldman et al., 2004; Ghazoul, 2006; Lopezaraiza-Mikel et al., 2007; La'zaro et al., 2009). Wagenius (pers. comm.) indicated that co-flowering Amorpha canescens Pursh competed with Echinacea angustifolia De Candolle for Bombus pollinators rather than facilitating its pollination. Accordingly, the importance of generalist bumblebees as coneflower pollinators may differ from that reported in the current study (Mitchell et al., 2009), depending on community flowering phenology. It is important that species' abundances and flowering phenologies are included in descriptions of rare species' pollination biology and seed production.

Factors affecting seed production.--During 2005, seed production was limited in LS, FD, and more so in BD. Since there was no indication that soil cations, other resources that regulate plant size, pollen viability, or pollen limitation reduced seed production, we speculate that a genetic mechanism caused this limitation. Peters et al. (2009) determined from genetic data that inbreeding was not elevated in small coneflower populations relative to larger populations but did consider loss of S alleles via genetic drift a serious threat to small population viability. Accordingly, we suggest that a paucity of S alleles was the most likely reason for the reduced seed set. Future studies of seed set following artificial crosspollinations between neighboring coneflower populations and determining frequencies of inter-population pollinator flights would help interpret patterns of seed production among populations.

Although a direct relationship between levels of pollinator service and seed production has been described by recent workers (Yates and Ladd, 2005; Aguilar et al., 2006; Campbell and Husband, 2007), neither the current study or Wagenius and Lyon (2010) found a direct relationship. We showed that coneflower seed set in KOR and SHR during 2004 equaled that in PC even though flower visits were 5 and 16 times more frequent, respectively, in PC. Wagenius and Lyon (2010) reported abundant solitary Halictidae bee visits to but reduced seed production in small, isolated populations of Echinacea angustifolia. A consideration of pollinator behavior and size can contribute to explaining the different seed set--pollinator service trends in these two studies. Halictid bees typically remain within local flower patches (Ginsberg, 1985) and respond to landscape features at a fine ([less than or equal to] 250 m) spatial scale (Steffan-Dewenter et al., 2002). This behavior tends to promote pollination among plants that may be closely related or share common S alleles. Alternatively, Bombus intersperses long inter-flower flights with short flights (Ginsberg, 1985; Schulke and Waser, 2001; Kraus et al., 2009) and, in general, responds to conditions at a coarser landscape scale (Steffan-Dewenter et al., 2002). Steffan-Dewenter and Tscharntke (1999) showed that Bombus pollinator abundance decreased less than solitary bees at plant isolation distances up to 1000 m. Behavior of bumblebees, more than halictid bees, tends to promote pollination among plants that have different S alleles. Additionally, since bumblebees are larger than halictids, they can accumulate larger pollen loads from more pollen donor plants. These large pollen loads accumulated from several donor plants may mean that few bumblebee visits are required for maximal seed production. In combination, these results suggest that halictid bees would be less effective than Bombus in cross-pollinating isolated plants and may partly explain the different results noted by Wagenius and Lyon (2010) and the current study. In summary, both pollinator foraging behavior and effectiveness in response to plant community composition and blooming phenologies must be considered in interpretations of population size and isolation on seed production.

Performance of seeds and seedlings from a large and a small population.--SHR seedlings were smaller than PC seedlings. Inbreeding has previously been suggested as a likely cause for relatively small seedlings in small populations (Menges, 1991, 1995; Young et al., 1996; Fischer and Matthies, 1998; Bruna, 2002; Donaldson et al., 2002; Hooftman et al., 2003; Severns, 2003; Hensen and Oberpreiler, 2005; Kolb, 2005; Ward and Johnson, 2005). However, inbreeding was previously shown not to be elevated in SHR or other small coneflower populations (Peters et al., 2009). Alternatively, chronically open habitat can select for relatively small seedlings (Crawley and Nachapong, 1985; McGinley et al., 1987; Tripathi and Khan, 1990). The amount of litter and living vegetation has been controlled at SHR by annum mowing. Since the mid-1980s, PC has been mowed only occasionally (Walker, 2009) so dense annual stands of grasses and forbs have produced a substantial litter layer. Accordingly, we suggest adaptation to local site conditions as a more likely explanation than inbreeding for the relatively small SHR seedlings. Future studies are needed of possible demographic consequences of variable coneflower seedling sizes in habitats that differ in amount of vegetative and litter cover.

Conservation recommendations.--Based on results of our study, we make the following management suggestions: (1) Conservation workers should maintain Bombus habitat near coneflower populations (Sipes and Tepedino, 1995; Tepedino et al., 1997). Habitat should include nest sites (Alford, 1975) and spring-blooming species which provide queens with pollen and nectar as they establish nests early in the season and help habituate Bombus foragers to coneflower habitat (Waser and Real, 1979; Tepedino et al., 1997). (2) If coneflower seed set is low, workers could reduce the flower abundance of those co-flowering species in the community which attract bumblebees (Lopezaraiza-Mikel et al., 2007). Coneflower seed production in response to flower reduction should be monitored because it could be reduced instead of increased. (3) If outbreeding depression is not indicated among neighboring coneflower populations (Waser and Price, 1989; Tallmon et al., 2004; Seltmann et al., 2009), workers should manage habitat separating these populations to encourage inter-population pollinator flights (Cane, 2001), thereby reducing the probability of local S allele loss.

Acknowledgments.--We thank Dr. Cavell Brownie and Wade Wall for initial statistical assistance; Andy Walker, Angela Richardson, Julie Roszco, Jennifer Petite, and Caitlin Elam for their help with data collection; Dave Stephan for insect identification; Jenny Xiang for use of her lab facilities; Donna Wright for providing greenhouse space; and the NCSU Phytotron for research space.

LITERATURE CITED

ALFORD, D. V. 1975. Bumblebees. Davis-Poynter, Ltd., London, England. 352 p.

ALLEY, H. AND J. AFFOLTER. 2004. Experimental comparison of reintroduction methods for the endangered Echinacea laevigata (Boyton and Beadle) Blake. Nat. Areas J., 24:345-350.

--, M. REIGER, AND J. AFFOLTER. 2005. Effects of development light level on photosynthesis and biomass production in Echinacea laevigata, a federally listed endangered species. Nat. Areas J., 25:117-122.

ANDERSON, G.J., G. BERNARDELLO, T. F. STUESSY, AND D.J. CRAWFORD. 2001. Breeding system and pollination of selected plants endemic to Juan Fernanded Islands. Am. J. Bot., 88:220-233.

APSIT, V. J. AND P. M. DIXON. 2001. Genetic diversity and population structure in Echinacea laevigata (Boynton and Beadle) Blake, and endangered plant species. Nat. Areas J., 24:345-350.

AGUILAR, R., L. ASHWORTH, L. GALETTO, AND M. A. AIZEN. 2006. Plant reproductive susceptibility to habitat fragmentation: review and synthesis through a meta-analysis. Ecol. Let., 9:968-980.

ASHWORTH, L., R. AGUILAR, L. GALETTO, AND M. A. AIZEN. 2004. Why do pollination generalist and specialist plant species show similar reproductive susceptibility to habitat fragmentation? J. Ecol, 92:717-719.

BOSCH, M. AND N. M. WASER. 1999. Effects of local density on pollination and reproduction in Delphinium nuttallianum and Aconitum columbianum. Am. J. Bot., 86:871-879.

BRUNA, E. 2002. Effects of forest fragmentation on Heliconia acuminata seedling recruitment in central Amazonia. Oecologia, 132:235-243.

BUCHANAN, M. F. AND J. T. FINNEGAN. 2010. Natural heritage program list of rare plant species of North Carolina. North Carolina Natural Heritage Program, Department of Environment and Natural Resources, Raleigh. 125 p.

BYERS, D. L. AND T. R. MEAGHER. 1992. Mate availability in small populations of plant species with homomorphic sporophytic self-incompatibility. Hered., 68:353-359.

CAMPBELL, D. R. 1985. Pollinator sharing and seed set of Stellaria pubera: Competition for pollination. Ecology, 66:544-553.

CAMPBELL, L. G. AND B. C. HUSBAND. 2007. Small populations are mate-poor but pollinator-rich in a rare, self-incompatible plant, Hymenoxys herbacea. New Phytol., 174:915-925.

CANE, J. H. 2001. Habitat fragmentation and native bees: a premature verdict? Conserv. Ecol., 5(1):3. <http://www.consecol.org/vol5/issl/art3>

CLEVENGER, D. J., J. E. BARRETT, H. J. KILEE, AND D. G. CLARK. 2004. Factors affecting seed production in transgenic ethylene-insensitive petunias. J. Amer. Soc. Hort. Sci., 129:401-406.

CRAWLEY, M. J. AND M. NACHAPONC. 1985. The establishment of seedlings from primary and regrowth seeds of ragwort (Senecio jacobaea). J. Ecol., 73:255-261.

CUNNINGHAM, S. A. 2000. Effects of habitat fragmentation on the reproductive ecology of four plant species in mallee woodland. Conserv. Biol., 14:758-768.

DAYTON, B. R. 1966. The relationship of vegetation to Iredell and other Piedmont soils in Granville County, North Carolina. J. Elisha Mitchell Sci. Soc., 82:108-118.

DEMCHIK, S. M. AND T. A. DAY. 1996. Effect of enhanced UV-B radiation of pollen quantity, quality, and seed yield in Brassica rapa (Brassicaceae). Amer. J. Bot., 83:573-579.

DONALDSON, J., I. NANNI, C. ZACHARIADES, AND J. KEMPER. 2002. Effects of habitat fragmentation on pollinator diversity and plant reproductive success in renosterveld shrublands of South Africa. Conserv. Biol., 16:1267-1276.

EDWARDS, L. AND D. MADSEN. 1993. Technical draft recovery plan for smooth coneflower (Echinacea laevigata). Class project for Wildlife Biology 813. Conservation and Ecology of Endangered Species. Clemson University, Clemson, South Carolina. 18 p.

ESTILL, J. C. AND M. B. CRUZAN. 2001. Phytogeography of rare plant species endemic to the southeastern United States. Castanea, 66:3-23.

FARJI-BRENER, A. G. AND L. GHERMANDI. 2008. Leaf-cutting ant nests near roads increase fitness of exotic plant species in natural protected areas. P. Roy. Soc. Lond. B Bio., 275:1431-1440.

FAUSTO, J. A., V. M. ECKHART, AND M. A. GEBER. 2001. Reproductive assurance and the evolutionary ecology of self-pollination in Clarkia xantiana (Onagraceae). Am. J. Bot., 88:1794-1800.

FELDMAN, T. S., W. F. MORRIS, AND W. G. WtLSON. 2004. When can two plant species facilitate each other's pollination? Oikos, 105:197-207.

FENNEMAN, N. M. 1938. Physiography of the Eastern United States, 1st ed. McGraw Hill Book Company, Inc., New York, New York and London. 714 p.

FISCHER, M. AND D. MATTHIES. 1998. Effects of population size on performance in the rare plant Gentianella germanica. J. Ecol., 86:195-204.

GHAZOUL, J. 2006. Floral diversity and the facilitation of pollination. J. Ecol., 94:295-304.

GINSBERG, H. S. 1985. Foraging movements of Halictus ligatus (Hymenoptera: Halictidae) and Ceratina calcarata (Hymenoptera: Anthophoridae) on Chrysanthemum leucanthemum and Erigeron annuus (Asteraceae). J. Kansas Entomol. Soc., 58:19-26.

GRIFFIN, S. R. AND S. C. H. BARRETT. 2002. Factors affecting low seed:ovule ratios in a spring woodland herb, Trillium grandiflorum (Melanthiaceae). Int. J. Plant Sci., 163:581-590.

GROSS, R. S. AND P. A. WERNER. 1983. Relationships among flowering phenology, insect visitors, and seedset of individuals: Experimental studies on four co-occurring species of goldenrod (Solidago: Compositae). Ecol. Monogr., 53:95-117.

HENSEN, I. AND C. OBERPREILER. 2005. Effects of population size on genetic diversity and seed production in the rare Dictamnus albus (Rutaceae) in central Germany. Conserv. Gen., 6:63--73.

HERRERA, C. M. 1987. Components of pollinator "quality": comparative analysis of a diverse insect assemblage. Oikos, 50:79-90.

HOOFTMAN, D. A. P., M. VAN KLEUNEN, AND M. DEIMER. 2003. Effects of habitat fragmentation on the fitness of two common wetland species, Carex davalliana and Succisa pratensis. Oecologia, 134:350-359.

HORVITZ, C. C. AND D. W. SCHEMSKE. 1988. A test of the pollinator limitation hypothesis for a neotropical herb. Ecol., 69:200-206.

JENNERSTEN, O. 1988. Pollination in Dianthus deltoides (Caryophyllaceae): effects of habitat fragmentation on visitation and seed set. Conserv. Biol., 2:359-366.

-- AND S. G. NILSSON. 1993. Insect flower visitation frequency and seed production in relation to patch size of Viscaria vulgaris (Caryophyllaceae). Oikos, 68:283-292.

JURAS, P. 1997. The presettlement piedmont savanna: A model for landscape design and management. C:\DataDrive\D\RESEARCHA\Echinacea\The Presettlement Piedmont Savanna--Philip Juras MLA Thesis 1997.mht. [Accessed 27 Oct. 2010].

KAYE, T. N. 1999. From flowering to dispersal: Reproductive ecology of an endemic plant, Astragalus australis var. olympicus (Fabaceae). Am. J. Bot., 86:1248-1256.

KEARNS, C. A. AND D. W. INUOYE. 1993. Techniques for pollination biologists. University Press of Colorado, Niwot, Colorado. 583 p.

KOLB, A. 2005. Reduced reproductive success and offspring survival in fragmented populations of the forest herb Phyteuma spicatum, J. Ecol., 93:1226-1237.

KRAUS, F. B., S. WOLF, AND R. F. A. MORITZ. 2009. Male flight distance and population substructure in the bumblebee Bombus terrestris. J. Anim. Ecol., 78:247-252.

KUNIN, W. E. 1993. Sex and the single mustard: population density and pollinator behavior effects on seed-set. Ecology, 74:2145-2160.

LAMONT, B. B., P. G. L. KLINKHAMER, AND E. T. F. WITKOWSKI. 1993. Population fragmentation may reduce fertility to zero in Banksia goodii--a demonstration of the Allee effect. Oecologia, 94:446-450.

LA'ZARO, A., R. LUNDGREN, AND O. TOTLAND. 2009. Co-flowering neighbors influence the diversity and identity of pollinator groups visiting plant species. Oikos, 118:691-702.

LEBLOND, R.J. 2001. Endemic plants of the Cape Fear Arch Region. Castanea, 66:83-97.

LEUSZLER, H. K., V.J. TEPEDINO, AND D. G. ALSTON. 1996. Reproductive biology of purple coneflower in southwestern North Dakota. Prair. Nat., 28:91-102.

LOPEZARAIZA-MIKEL, M. E., R. B. HAYES, M. R. WHALLEY, AND J. MEMMOTT. 2007. The impact of an alien plant on a native plant-pollinator network: an experimental approach. Ecol. Letters, 10:539-550.

MARTEN, S. AND M. QUESADA. 2001. Phenology, sexual expression, and reproductive success of the rare neotropical palm Geonoma epetiolata. Biotropica, 33:596-605.

MCGINLEY, M. A., D. H. TEMME, ANn M. A. GEBER. 1987. Parental investment in offspring in variable environments: Theoretical and empirical considerations. Am. Nat., 130:370-298.

MCGREGOR, R. L. 1968. The taxonomy of the genus Echinacea (Compositae). Univ. Kansas Sci. Bull., 48:113-142.

MEHLICH, A. 1976. New buffer method for rapid estimation of exchangeable acidity and lime requirement. Commun. in Soil Sci. Pl. Anal., 7:637-652.

--. 1984. Mehlich 3 soil test extractant: a modification of the Mehlich 2 extractant. Commun. in Soil Sci. and Pl. Anal., 15:1409-1416.

MENGES, E. S. 1991. Seed germination percentage increases with population size in a fragmented prairie species. Conserv. Biol., 5:158-164.

1995. Factors limiting fecundity and germination in small populations of Silene regia (Caryophyllaceae), a rare hummingbird prairie forb. Am. Midl. Nat., 133:242-55.

MICROSOFT OFFICE CORPORATION. 2003. EXCEL [computer program].

MITCHELL, R. J., R. J. FLANAGAN, B. J. BROWN, N. M. WASER, AND J. D. KARRON. 2009. New frontiers in competition for pollination. Ann. Bot., 103:1403-1413.

MITCHELL, T. B. 1962. Bees of the eastern United States. II. Technical Bulletin No. 152. North Carolina Agricultural Experiment Station, 557 p.

MORGAN, J. W. 1999. Effects of population size on seed production and germinability in an endangered, fragmented grassland plant. Conserv. Biol., 13:266-273.

MOSQUIN, T. 1971. Competition for pollinators as a stimulus for the evolution of flowering time. Oikos, 22:398-402.

MUOTOZ, A. A. AND M. T. ARROYO. 2006. Pollen limitation and spatial variation of reproductive success in the insect-pollinated shrub Chuquiraga oppositifolia (Asteraceae) in the Chilean Andes. Arct. Antarct. Alp. Res., 38:608-613.

MURDOCK, N. A. 1992. Endangered and threatened wildlife and plants: Echinacea laevigata (Smooth Coneflower) determined to be endangered. Fed. Reg., 57:46340-46344.

--. 1995. Recovery plan for smooth coneflower (Echinacea laevigata). United States Fish and Wildlife Service, Southeast Region, Atlanta, Georgia. 31 p.

NE'EMAN, G., R. NE'EMAN, AND A. M. ELLISON. 2006. Limits to reproductive success of Sarracenia purpurea (Sarraceniaceae). Am. J. Bot., 93:1660-1666.

NORTH CAROLINA NATURAL HERITAGE PROGRAM. 2001. Guide to Federally Listed Endangered and Threatened Species of North Carolina. Division of Parks and Recreation, North Carolina Department of Environment and Natural Resources, Raleigh. 134 p.

PARRISH, J. D. A. AND F. A. BAZZAZ. 1979. Difference in pollination niche relationships in early and late successional plant communities. Ecology, 60:597-610.

PETERS, M. D., Q. XIANG, D. T. THOMAS, J. STUCk, AND N. K. WHITEMAN. 2009. Genetic analyses of the federally endangered Echinacea laevigata using amplified fragment length polymorphisms (AFLP)--Inferences in population genetic structure and mating system. Conserv. Genet., 10:1-14.

PHILIPPI, T. B., S. COLLINS, S. GUISTI, AND P. M. DIXON. 2001. A multistage approach to population monitoring for rare plant populations. Nat. Areas J., 21:111-116.

PRASAD, P. V. V., K. J. BOOTE, AND L. H. ALLEN, JR. 2006. Adverse high temperature effects on pollen viability, seed-set, seed yield and harvest index of grain-sorghum [Sorghum bicolor (L.) Moench] are more severe at elevated carbon dioxide due to higher tissue temperatures. Agric. For. Meteorol., 139:237-251.

RAMSEY, M. 1995. Causes and consequences of seasonal variation in pollen limitation of seed production in Blandfordia grandiflora (Liliaceae). Oikos, 73:49-58.

RATSIRAR SON, J. AND J. A. SILANDER. 1996. Reproductive biology of the threatened Madagascar Triangle Palm: Neodypsis decaryi Jumelle. Biotropica, 28:737-745.

REINARTZ, J. A. AND D. H. LES. 1994. Bottleneck-induced dissolution of self-incompatibility and breeding system consequences in Aster furcatus (Asteraceae). Am. J. Bot., 81:446-455.

ROHLF, J. F. AND R. H. SOKAL. 1969. Statistical Tables. W.H. Freeman and Co., San Francisco, California. 253 p.

SAS INSTITUTE INC. 2008. SAS/STAT[R] 9.2. User's Guide. SAS Institute Inc, Cary, NC. 104 p.

SCHULKE, B. AND N. M. WASER. 2001. Long-distance pollinator flights and pollen dispersal between populations of Delphinium nuttallianum. Oecologia, 127:239-245.

SELTMANN, P., D. COCUCCI, A. RENISON, A. CIERJACKS, AND I. HENSEN. 2009. Mating system, outcrossing distance effects and pollen availability in the wind-pollinated treeline species Polylepis australis Bitt. (Rosaceae). Basic Appl. Ecol., 10:52-60.

SEVERNS, P. 2003. Inbreeding and small population size reduce seed set in a threatened and fragmented plant species, Lupinus sulphureus ssp. kincaidii (Fabaceae). Biol. Conserv., 110:221-229.

SHORE, J. S. 1993. Pollination genetics of the common milkweed, Asclepias syriaca L. Hered., 70:101-108.

SIH, A. AND M. BALTUS. 1987. Patch size, pollinator behavior, and pollinator limitation in catnip. Ecology, 68:1679-1690.

SIPES, S. D. AND V. J. TEPEDINO. 1995. Reproductive biology of the rare orchid, Spiranthes Diluvialis: breeding system, pollination, and implications for conservation. Conserv. Biol., 9:929-938.

SOKAL, R. R. AND F.J. ROHLF. 1969. Biometry. W.H. Freeman and Co., San Francisco, California. 776 p.

SORRIE, B. A. AND A. S. WEAKLEY. 2001. Coastal Plain vascular plant endemics: Phytogeographic patterns. Castanea, 66:50-82.

-- AND --. 2006. Conservation of the endangered Pinus palustris ecosystem based on Coastal Plain centers of plant endemism. App. Veg. Sci., 9:59-66.

STEFFAN-DEWENTER, I., U. MUNZENBERG, C. BUERGER, C. THIES. AND T. TSCHARNTKE. 2002. Scale-dependent effects of landscape context on three pollinator guilds. Ecology, 83:1421-1432.

-- AND Z. TSCHARNTKE. 1999. Effects of habitat isolation on pollinator communities and seed set. Oecologia, 121:432-440.

SUZUKI, N. 2000. Pollinator limitation and resource limitation of seed production in the Scotch broom, Cytisus scoparius (Leguminosae). Plant Spec. Biol., 15:187-193.

TALLMON, D. A., G. LUIKART, AND R. S. WAPLES. 2004. The alluring simplicity and complex reality of genetic rescue. Trends Ecol. Evol., 19:289-296.

TEPEDINO, V. T., S. D. SIPES, J. L. BARNES, AND L. L. HICKERSON. 1997. The need for "extended care" in conservation: Examples from studies of rare plants in the Western United States, p. 245-248.

In: K. W. Richards (ed.). Seventh Internatinoal Symposium on Pollination: From Theory to Practise. J. Acta Hort., 437:1-461.

TIMMERMAN-ERSKINE, M. AND R. S. BOYD. 1999. Reproductive biology of the endangered plant Clematis socialis (Ranunculaceae). J. Torr. Bot. Soc., 126:107-116.

TRIPATHI, R. S. AND M. L. KHAN. 1990. Effects of seed weight and microsite characteristics on germination and seedling fitness in two species of Quercus in a subtropical wet hill forest. Oikos, 57:289-296.

TSALIKI, M. AND M. DIEKMANN. 2009. Fitness and survival in fragmented populations of Narthecium ossifragum at the species' range margin. Acta Oecol., 35:415-421.

UNITED STATES DEPARTMENT OF AGRICULTURE, NATURAL RESOURCES CONSERVATION SERVICE. 2006. Plants Database. http://plants.usda.gov. [Accessed 15 Apr. 2010].

URBATSCH, L. E., K. M. NEUBIG, P. B. Cox. 2006. Echinacea, p. 88-92. In: Flora of North America Editorial Committee. Flora of North America North of Mexico, vol. 21. 16 + vols. New York and Oxford.

WAGENIUS, S. 2004. Style persistence, pollen limitation, and seed set in the common prairie plant Echinacea angustifolia (Asteraceae). Int. J. Plant Sci., 165:595-603.

--. 2006. Scale dependence of reproductive failure in fragmented Echinacea populations. Ecology, 87:931-941.

-- AND S. P. LYON. 2010. Reproduction of Echinacea angustifolia in fragmented prairie is pollen-limited but not pollinator-limited. Ecology, 91:733-742.

WALKER, A. S. 2009. Soil seedbank of rare plant communities associated with diabase soils in Durham and Granville Counties, North Carolina. MS thesis, North Carolina State University, Raleigh, North Carolina. 22 p.

WALKER, J. 1993. Rare vascular plant taxa associated with the Longleaf Pine ecosystems: patterns in taxonomy and ecology. Proc. Ann. Tall Timbers Fire Ecol. Conf, 18:105-126.

WARD, M. AND S. D. JOHNSON. 2005. Pollen limitation and demographic structure in small fragmented populations of Brunsvigia radulosa (Amaryllidaceae). Oikos, 108:253-262.

WASER, N. M. AND M. V. PRICE. 1989. Optimal outcrossing in Ipomopsis aggregata: Seed set and offspring fitness. Evol., 43:1097-1109.

-- AND L. A. REAL. 1979. Effective mutualism between sequentially flowering plant species. Nature, 281:670-672.

WEAKLEY, A. S. 2010. Flora of the Carolinas, Virginia, Georgia and surrounding areas. Working Draft of 15 May 2011. University of North Carolina, Chapel Hill, North Carolina. 1072 p.

WIGGAM, S. AND C.J. FERGUSON. 2005. Pollinator importance and temporal variation in a population of Phlox divaricata L. (Polemoniaceae). Am. Midl. Nat., 154:42-54.

YATES, C.J. AND P. G. LADD. 2005. Relative importance of reproduction biology and establishment ecology for persistence of a rare shrub in a fragmented landscape. Conserv. Biol., 19:239-249.

YOUNG, A., T. BOYLE, AND T. BROWN. 1996. The population genetic consequences of habitat fragmentation for plants. Trends Ecol. Evol., 11:413-418.

SUBMITTED 13 DECEMBER 2010

ACCEPTED 26 JANUARY 2012

J. M. STUCKY (1)

Department of Plant Biology, Box 7612, North Carolina State University, Raleigh 2769-5

L. E. GADD

North Carolina Natural Heritage Program, 1601 MSC Raleigh 27699

AND

C. ARELLANO

Department of Statistics, Box 8203, North Carolina State University, Raleigh 27695

(1) Corresponding author: Telephone: (919) 515-2227; e-mail: jmstucky@ncsu.edu
TABLE 1.--Experimental populations of Echinacea laevigata

                      General location;
Population          (county, coordinates)            Habitat

Picture Creek     Granville, 36.1[degrees]N,   60 m wide powerline
(PC)              78.7[degrees]W               right-of-way

Snow Hill Rd.     Durham, 36.1[degrees]N,      roadside adjacent to
(SHR)             78.9[degrees]W               oak forest

Knap of Reeds     Granville, 36.1[degrees]N,   10 m wide abandoned
(KOR)             78.8[degrees]W               powerline right
                                               of-way adjacent
                                               to oak forest

Briardale (BD)    Durham, 36.1[degrees]N,      roadside adjacent to
                  78.9[degrees]W               oak forest

Lakeside (LS)     Durham, 36.1[degrees]N       roadside adjacent to
                  78.9[degrees]W               oak forest

Freudenberg       Durham, 36.1[degrees]N       edge of oak forest
(FB)              78.9[degrees]W

                  Distance to                          Daily no.
                  nearest E.                           flowering
                   laevigata     No. plants that    stems/[m.sup.2],
                  population    flowered during      mean [+ or -]
Population           (km)       growing season        [SD.sup.1]

Picture Creek     6.2           [approximately     7.2 [+ or -] 2.02
(PC)                                 equal to]
                                       50,000

Snow Hill Rd.     0.8                      23     0.04 [+ or -] 0.220
(SHR)

Knap of Reeds     6.2                      31     0.08 [+ or -] 0.052
(KOR)

Briardale (BD)    0.4                      15     0.10 [+ or -] 0.076

Lakeside (LS)     0.4                      17     0.09 [+ or -] 0.057

Freudenberg       0.8                      48     0.78 [+ or -] 0.564
(FB)

(1) Stems/[m.sup.2] determined as number of flowering plants
divided by rectangular dimensions of area  occupied for SHR and
KOR during 2004 and BD, LS and FB during 2005; determined for PC
as number  of flowering plants in 2 X 2 m pollinator observation
plots during 2004 and 2005

TABLE 2.--Nutrient status of soils and open-pollination seed
production for experimental populations

                                     Base
                                  saturation           Total
                                    (% of CEC     concentration of
                     CEC (1)        occupied        base cations
                    (meq/100         by base          (meq/100
Population   pH    [cm.sup.3])    cations) (2)      [cm.sup.3])

  PC         6.3      25.2             94               23.7
  SHR        6.5      20.5             96               19.7
  KOR        6.1      21.4             93               19.9
  BD         6.1      18.6             89               16.6
  LS         5.9      17.5             91               15.9
  FB         6.5      23.2             95               22.0

             Open-pollination
             seed production
Population       (%) (3)

  PC            42, 56 (4)
  SHR               52
  KOR               50
  BD                22
  LS                38
  FB                42

(1) Cation exchange capacity

(2) Base cations include [Ca.sup.++] , [K.sup.+], and [Mg.sup.++]

(3) From Table 5

(4) Seed production for 2004 and 2005

TABLE 3.--Pollinator importance of Echinacea laevigata flower
visitors

                             Location 2 of pollen
                                loads on body;
                             mean rank (3), range

Pollinator group     n (1)                  Rank

Bombus spp. (8)         13   3.8 (a), 3-4   3.5
Xylocopa virginica
 (L.)                    5   2.7 (a), 2-4   3.5
Lygaeus kalmii
 Stal                    5   3.3 (a), 3-4   3.5
Megachile spp (9)       13   3.1 (a), 0-4   3.5
Hesperiidae (10)         5   1.2 (a), 0-4   3.5
Nymphalidae (11)         5   1.0 (a), 0-2   3.5

                     No. of E. laevigata pollen
                     grains per 0.05 ml sample;
                       [mean.sub.group] (4,5)
                               95% CL

Pollinator group                            Rank

Bombus spp. (8)      39.3 (ab), 4.0-110     4.25
Xylocopa virginica
 (L.)                97.4 (a), 22.3-222.3   5.5
Lygaeus kalmii
 Stal                12.9 (b), 5.2-24.8     3
Megachile spp (9)    13.4 (b), 6.9-21.9     3
Hesperiidae  (10)     1.5 (b), 0-6.4        3
Nymphalidae (11)      2.8 (a), 0-11.4       3

                            Proportion of
                            E. laevigata
                            pollen grains
                        per 0.05 ml sample;
                       [mean.sub.group] (4,6),
                                95% CL

Pollinator group                           Rank

Bombus spp. (8)      76.3 (a), 55.8-91.9   3.5
Xylocopa virginica
 (L.)                86.6 (a), 77.8-93.2   3.5
Lygaeus kalmii
 Stal                95.3 (a), 72.4-100    3.5
Megachile spp (9)    73.7 (a), 61.4-84.4   3.5
Hesperiidae (10)     51.4 (a), 16.3-99.8   3.5
Nymphalidae (11)     60.6 (a), 0-100       3.5

                       Visits/head/15 min;
                        mean 4,7, 95% CL

Pollinator group                             Rank   Sum of ranks

Bombus spp. (8)      2.3 (a), 1.7-2.6        6          17.3
Xylocopa virginica
 (L.)                0.011 (c), >0.01-0.01   2          14.5
Lygaeus kalmii
 Stal                >0.01 (c), >0.01-0.01   2          12
Megachile spp (9)    0.1 (b), 0.07-0.15      4.5        14.5
Hesperiidae (10)     0.1 (b), 0.05-0.13      4.5        14.5
Nymphalidae (11)     >0.01 (c), 0->0.01      2          12

(1) Number of insects evaluated for the first three criteria

(2) Location scores: 4--all pollen on body parts most likely to
contact stigmas; 3 most pollen on body parts most likely to
contact stigmas but some on dorsal surface, wings, or in pollen
sacs; 2--all pollen on body parts not likely to contact stigmas;
1 < 10 pollen grains on the body; 0 no pollen evident

(3) Kruskal-Wallis test, [H.sub.(5)] = 7.620, P > 0.05, no
differences among mean ranks

(4) Means in a column with the same superscript are not different
according to ANOVA with Tukey's multiple pairwise mean
comparison; [alpha] = 0.05

(5) ANOVA used square root transformed data for [mean.sub.group]
[F.sub.5,35] = 6.23, P < 0.001

(6) ANOVA used arcsine transformed data for [mean.sub.group],
[F.sub.5,35] = 1.11, P = 0.37

(7) ANOVA used square root transformed data, n = 76 observation
periods, [F.sub.5,449] = 246.29, P < 0.001

(8) Includes B. bimaculatus Cresson, B. citrinus Smith, B.
griseocollis Degeer, B. impatiens Cresson, B. pennsylvanicus
Degeer

(9) Includes M. brevis Say, M.mendica Cresson, M. sculpturalis
Smith, M. texana Cresson, and M. xylocopoides Smith

(10) Includes Atrytone logan Edwards, Atrytonopsis hianna
Scudder, Epargyreus clarus Cramer, Polites sp. probably
themistocles Latreille and Thorybes bathyllus J.E. Smith

(11) Includes Euptoieta claudia Cramer, Speyeria cybek F., and
Vanessa virgineinsis Drury

TABLE 4.--Insect visitation in Echinacea laevigata populations

                                   No. visits/      Visitor species
                      # 15 min.    head/15 min.     richness/15 min.
Population          intervals (n)  (SE) (1)              (SE)

2004
  Picture Creek          42        1.6 (0.28) (a)   2.8 (0.21) (a)
  Knap of Reeds           7        0.3 (0.20) (a)   0.8 (0.41) (b)
  Snow Hill Road          5        0.1 (0.24) (a)   0.6 (0.45) (b)
2005
  Picture Creek          36        4.8 (0.19) (a)   2.5 (0.17) (a)
  Briardale Road         12        1.2 (0.32) (c)   1.5 (0.26) (b)
  Lakeside Drive          6        1.6 (0.76) (c)   1.0 (0.38) (b)
  Freudenberg             6        8.7 (1.54) (a)   1.0 (0.38) (b)

(1) Tukey's means groupings, a-value for Tukey's <0.05

(2) Means with the same superscript in a column within individual
years are not different using Tukey's studentized test

TABLE 5.--Seed production in open-and supplemental-pollination
treatments in each population

                                                 % viable seeds
                                      No.         head (1,2);
                     Pollination   surviving        per mean
Year    Population    treatment      heads        [+ or -] SD

2004        PC           Open          24     42 [+ or -] 15 (a)
                         Cross         10     47 [+ or -] 18 (A)
           SHR           Open           6     52 [+ or -] 21 (a)
                         Cross         23     62 [+ or -] 5 (A)
           KOR           Open          15     50 [+ or -] 14 (a)
                         Cross          3           -
2005        PC           Open          10     56 [+ or -] 13 (a)
                         Cross         83     35 [+ or -] 9 (A)
            BD           Open          15     22 [+ or -] 13 (c)
                         Cross         43     27 [+ or -] 14 (B)
            LS           Open           8     38 [+ or -] 12 (b)
                         Cross         23     45 [+ or -] 8 (A)
            FD           Open          10     42 [+ or -] 5 (b)
                         Cross          6     49 [+ or -] 13 (A)

                            ANOVA;
                          Pollination
                           treatment
                         effect within
Year    Population        population

2004        PC       [F.sub.1,32] = 0.79,
                       P = 0.38
           SHR       [F.sub.1,16] = 0.41,
                       P = 0.54
           KOR          -

2005        PC       [F.sub.1,16] = 14.42,
                       P = 0.002
            BD       [F.sub.1,17] = 0.43,
                       P = 0.52
            LS       [F.sub.1,4]  = 0.52,
                       P = 0.51
            FD       [F.sub.1,14] = 2.81,
                       P = 0.12

(1) Populations with the same lower case superscript within a
year do not differ for open-pollination seed set

(2) Populations with the same upper case superscript within a
year do not differ for supplemental pollination seed set

(3) Longhorn beetle bored into peduncles of some heads in this
treatment. Seed production in these heads not included in
statistical analyses

TABLE 6.--Seed germination, seedling survivorship, and
seedling size in the field and greenhouse

Experimental                  % Seed germination;
environment      Population    mean [+ or -] SD

Field                PC      40.5 [+ or -] 12.12 (a)
                    SHR      54.5 [+ or -] 17.56 (a)

Greenhouse           PC      65.5 [+ or -] 28.57 (a)
                    SHR      55.5 [+ or -] 23.95 (a)

                                    % Seedling
Experimental                       survivorship;
environment      Population      mean [+ or -] SD

Field                PC      36.0 [+ or -] 12.87 (b)
                    SHR      49.5 [+ or -] 16.91 (a)

Greenhouse           PC      97.4 [+ or -] 11.23 (a)
                    SHR      98.8 [+ or -] 3.80 (a)

Experimental                  Seedling size (2);
environment      Population   mean [+ or -] SD

Field                PC       5.1 [+ or -] 1.56 (a)
                    SHR       4.1 [+ or -] 1.23 (b)

Greenhouse           PC      22.0 [+ or -] 4.31 (a)
                    SHR      14.4 [+ or -] 5.40 (b)

(1) Means with the same superscript in a column within an
individual experimental environment are not different

(2) Number of leaves per field seedling; dry weight (g) of
individual greenhouse seedlings
COPYRIGHT 2012 University of Notre Dame, Department of Biological Sciences
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2012 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Stucky, J.M.; Gadd, L.E.; Arellano, C.
Publication:The American Midland Naturalist
Article Type:Report
Geographic Code:1USA
Date:Jul 1, 2012
Words:10263
Previous Article:The distribution of Quercus rubra in the Maumee Lake Plain of Southeastern Michigan.
Next Article:Effects of hemlock mortality on streams in the Southern Appalachian Mountains.
Topics:

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters