| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Population Biology |
Department of Biology, Willamette University, 900 State Street, Salem, Oregon 97301 USA
Received for publication January 2, 2005. Accepted for publication July 11, 2005.
ABSTRACT
Reintroductions are increasingly used to enhance declining populations, yet comparative data for critical germination and establishment phases are seldom available for both rare and common herbaceous perennials. After introducing a total of >1800 seeds, we compared experimentally manipulated and natural populations of widespread Silene douglasii var. douglasii relative to rare S. douglasii var. oraria, known in only three coastal headlands. Despite equivalent ex situ germination, oraria field plots produced significantly fewer juveniles than douglasii plots indicating that seedling survival limits plant establishment. We also evaluated transplant vs. seed reintroductions as restoration tools, the effect of inbreeding on fitness, and the potential importance of buried seed pools. Germination declined rapidly for seeds over 1–2 years old, and only 2.2% of newly collected seeds of oraria survived as seedlings. Transplant survival over 5 years was greatest for outbred progeny; furthermore, 75% of the new seedlings emerged near outbred progeny from the original reintroduction. Despite similar ovule numbers and pollinator visitation, transplants exhibited 49–179% maladaptation in the formerly grazed site, with significantly lower fruit and seed set than adults in more diverse natural populations. This study experimentally identifies several key factors affecting plant reintroductions, facilitating effective development of large-scale reintroduction strategies for native perennials.
Key Words: Caryophyllaceae • inbreeding • maladaptation • Oregon • rarity • reintroduction • seedling survival • Silene
Conservationist biologists are increasingly faced with preserving or enhancing native plant populations that have fallen below a viable size. Reintroduction of a rare species into a habitat that it historically occupied is a tactic growing in popularity: it simultaneously increases population size and diversifies the gene pool, thereby addressing two of the most critical variables determining population viability (Maunder, 1992
; Ellstrand and Elam, 1993
; Guerrant and Pavlik, 1997
; Lande, 1999
). Even so, the challenges for successful restoration are formidable. Reintroductions enable plants to survive and reproduce successfully in a new environment, but the prerequisites are often unique to each species, and local adaptation may result in poor performance or maladaptation of geographically distant genotypes (Montalvo et al., 1997
; Jones and Hayes, 1999). Invasive species also pose a serious threat to native populations; they often thrive in disturbed habitats and may outcompete rare plants (Huenneke and Thomson, 1995
; Vitousek et al., 1997
). In native populations of Abronia, the most serious threat was the intrusion of European beach grass into inland dune habitats once occupied by native Abronia (McGlaughlin et al., 2002
). Yet habitat preferences and life history traits of native species are rarely known precisely, so reintroduction can be risky, leading to high mortality (Helenurm, 1998
; Drayton and Primack, 2000
). Despite the potential of transplantation as a conservation tool, however, few studies examine levels of maladaptation in the new environment or track the long-term viability of reintroduced populations (Maunder, 1992
; Husband and Campbell, 2004
). Even in Britain, California, and Hawaii, where the history of reintroduction extends 
100 years, the reasons for the failure of reintroductions remain largely unknown (Guerrant and Pavlik, 1997
).
The goal of reintroduction is to return a species to its historical habitat, which may have been altered in quality or extent by increased land use. In coastal to montane areas, this creation of fragmented "sky islands" has led to isolation, rarity, and a loss of genetic diversity (Young et al., 1996
; Menges and Dolan, 1998
). Most species needing restoration have declined in population size or range, which can lead to inbreeding and genetic drift (Ellstrand and Elam, 1993
; Helenurm and Parsons, 1997
; Newman and Pilson, 1997
). Inbreeding depression may occur if deleterious partially recessive or overdominant alleles are not purged from the gene pool by natural selection (Byers and Waller, 1999
; Frankham et al., 2001
). Although inbreeding greatly influences plant fitness and can affect restoration efforts and the probability of extinction (Newman and Pilson, 1997
; Bowles et al., 2002
; Brook et al., 2002
; Luijten et al., 2002
), it has rarely been considered in reintroduction projects (Sheridan and Karowe, 2000
; Kephart, 2004
). Only 11.9% and 4.8% of 42 in situ plantings reviewed by Husband and Campbell (2004)
used genetic or genecological criteria, most of which guided decisions as to source populations for collection or sampling, but were not used in tracking establishment of propagules.
In the Pacific Northwest, three diverse varieties of Douglas' catchfly [Silene douglasii (Peck) Hitch. & Maguire] provide an excellent case study for evaluating the importance of outbred progeny in seed and transplant reintroductions, as well as the relationship between rarity and seedling establishment. In S. douglasii, inbreeding depression and variable but low pollination levels (Brown and Kephart, 1999
; Kephart et al., 1999a
) may have contributed to the near extinction of the rare coastal S. douglasii var. oraria (hereafter called oraria) The mixed mating system of oraria also makes it an interesting candidate for evaluating the effect of inbreeding on reintroduction success and plant fitness. Although selfing is common now and is a likely result of pollen limitation and poor pollen quality (Brown and Kephart, 1999
), it is limited by strong protandry. When outcrossing does occur, the resulting progeny are more vigorous, but their survival in the glasshouse is similar to that of inbred progeny (Kephart et al., 1999a
).
Unfortunately, we lack comparative data on natural rates of seedling establishment of rare and common varieties of S. douglasii and other perennials differing in rarity. Even seed studies related to potential reintroduction (e.g., Drayton and Primrack, 2000
; Hamzé and Jolls, 2000
; Baskin and Baskin, 2002
; Table 12.1 in Husband and Campbell, 2004
) mostly focus on seed dormancy, germination, or seed bank characteristics (e.g., Budelsky and Galatowitsch, 1999
; Cochrane et al., 2001
). Furthermore, published studies of seed germination in Silene (e.g., Menges, 1991
, 1995
; Peroni and Armstrong, 2001
) thus far do not include the effect of inbreeding on germination and survival as part of restoration or experimental reintroductions.
Reintroducing seeds and transplants in multiple years is often essential for creating self-sustaining populations (Drayton and Primrack, 2000
), as is tracking both short-term parameters (e.g., survival, reproduction) and long-term population growth and resilience (Pavlik, 1996
; Husband and Campbell, 2004
). We followed the fates of multi-year transplant and seed reintroductions of S. douglasii relative to varietal status, vegetative cover, and progeny type, and studied juvenile and seedling establishment in natural and manipulated populations. Specifically, we explored the following questions: (1) Do seed germination and seedling establishment differ in natural populations of rare vs. common varieties or under varied competitive regimes? Prior demographic study indicated that low seedling numbers, competition, and recruitment limit population viability in the rare oraria variety (Kephart and Paladino, 1997
). To simulate differing competitive environments, we compared seed reintroduction plots in which existing vegetation was either clipped or cleared. Using seeds of varied age also allowed us to estimate the potential for a buried seed pool. (2) Do outbred progeny have greater fitness than inbred progeny in transplant and seed reintroduction treatments? This multi-year study extends research in which outbred transplants had higher 3-yr field survival and reproduction (Kephart, 2004
) to manipulative experiments designed to ascertain whether in situ seed reintroduction and seedling establishment differ for inbred vs. outbred progeny. (3) Are transplant and seed reintroductions successful as restoration strategies for oraria and rare plants in general? We measured community properties, seedling establishment, traits affecting plant fitness, and pollinator visitation in reintroduced populations of oraria against natural populations. We also estimated the percentage maladaptation to the new environment. In situ reintroductions function as colonizing populations for which the degree of maladaptation in the new environment (sensu Gomulkiewicz and Holt, 1995
; Husband and Campbell, 2004
), and the population's response to it may be critical for establishment.
MATERIALS AND METHODS
Study taxa and natural populations
We studied eight natural populations of S. douglasii varying in varietal designation, rarity, and geographic distribution within Oregon, USA (Table 1). The three varieties are morphologically distinct (Kephart and Sturgeon, 1993
; Kephart et al., 1999b
). Jack Creek, Iron Mountain, and Cone Peak exemplify east and west side populations of geographically widespread douglasii in the Cascades mountains. Silene douglasii var. rupinae, endemic to the Columbia River Gorge, grows in steep rocky habitats in both large (Angel's Rest) and small populations (Oneonta Gorge, Wahkeenah Falls). Extant in fragmented populations at only three sites, S. douglasii var. oraria is currently threatened in Oregon, of special concern in the USA, and internationally endangered (USDA, 2004
; ONHP, 2001
; Walter and Gillet, 1998
). The largest populations of oraria, at Cascade Head and Neahkahnie Mountain, are protected by The Nature Conservancy and Oregon State Parks.
|
750 m inland in a formerly grazed meadow closely resembling the natural habitat (Kephart, 2004
). Source plants for the 1998 reintroduction were 18–24 mo old, glasshouse-grown progeny (200) from in situ hand-, self- and cross-pollinations, and from open-pollinated individuals in the natural headland (Kephart, 2004
). The present study extends plant monitoring in time and spatial extent and augments the existing reintroduced populations with seed propagules.
Seed germination and establishment in rare and common varieties
Natural seedling establishment
To better understand natural levels of seedling establishment in relation to plant rarity, we compared natural seedling abundances among varieties using 0.25-m2 circular plots, placed around 15–20 randomly located adults in each of two natural populations of each variety of S. douglasii. We searched all plots intensively for seedlings and juveniles. A seedling has fewer than two leaf or cotyledon pairs, while juveniles have two or more pairs of true leaves.
Seed bank simulation: germination and seedling establishment in manipulated plots
Because seed bank characteristics and seedling establishment can determine the long-term persistence of populations, we designed a mock-seed bank experiment in field populations of rare oraria and common douglasii (Table 1). We compared in situ and glasshouse seed germination between varieties for different aged seeds, and using inbred and outbred progeny in experiments detailed below. We outplanted two sets of seeds of each variety encompassing both older seeds (e.g., 7 years) and either one year or current year seeds. In all, three study sites (Cascade Head, Jack Creek, and Cone Peak) each received at least 600 seeds. One set of seeds planted at each site also included progeny from hand-, cross-, self-, and open-pollination outplanted to plots with intact or cleared vegetation. Greenhouse studies controlled for environmental factors affecting germination, and mirrored the in situ studies by including seeds of varying age, varietal status, and progeny type.
Seed germination and inbreeding for rare oraria
We introduced the first set of seed from oraria at Cascade Head on the 28 May 2003 with 7-yr-old seeds collected from the same 1996 seed pool used in the initial reintroduction (Kephart, 2004
). To track germination and survival of the seeds of open-, self-, and cross-pollinated progeny, we planted 330 seeds in 2.5-cm color-coded plastic straws. We used gridded frames to plant 15 seeds in each standardized 40 x 24 cm plot and to randomize the location of each progeny type. Five seeds of each type were planted alternately in rows of three, spaced 8 cm apart. We split straws lengthwise for effective drainage and to facilitate their removal from the new seedlings. To assure that some progeny grew in microhabitats that currently support surviving plants from the original reintroduction, we placed half of the 22 plots 48 cm from an existing plant; the other half were placed randomly. To compare survival under differing levels of competition, we manipulated the existing vegetation. We paired two seed plots together at 11 locations; in one plot we cleared all vegetation before planting, and in the other, we clipped vegetation several centimeters above ground level. We mechanically scarified all seeds before planting to ensure germination if viable, because they were not exposed to seasonal temperature changes known to release seeds from dormancy (Baskin, 2003
). Soil levels within the straws paralleled external soil levels. One week after the seeds were planted, 0.5 L of water was applied to each plot in the early morning. We checked plots weekly for signs of germination from 28 May to 3 July and again on 23 July. Glasshouse trials confirmed that germination was equivalent for seeds planted with (37.7%, N = 53) and without (33.3%; N = 54) a surrounding straw (
2 test; P > 0.05).
On 7 February 2004 we implemented a second seed planting trial using recent seed collected July 2003 from open-pollinated plants at Cascade Head. Positioned 3 cm away from each open pollinated seed in the first seed reintroduction, we planted a new seed in a colored straw. In addition, we planted a third plot next to each of the existing plots with 15 additional open-pollinated seeds (2003); we clipped existing vegetation 3 cm from ground level. Seeds were not scarified because they were exposed to cold season temperatures. To retain and reduce disturbance to straws through natural processes and animal activity without interfering with seedling development, we covered newly clipped and prior clipped plots with 0.5 cm plastic netting, secured with wire clips. Finally, to simulate natural seeding without straws or netting, we cleared an 11 x 11 cm patch of ground in the northwest corner of each new plot, and scattered five seeds in the exposed soil. In total, we planted 330 new seeds (2003): 110 seeds in pre-existing plots, 165 seeds within the 11 new gridded plots, and 55 seeds scattered adjacent to each new plot.
Seed germination and inbreeding for common douglasii
We planted the first set of douglasii seed in July 2003 at Jack Creek and Cone Peak, using the same protocol as for oraria, in 15 sets of paired plots at each site, one cleared and one clipped of vegetation. We planted 1-yr-old seed collected in 2002 from in situ cross-, self-, and open-pollinations, placing 450 scarified seeds per study site in straws color-coded by progeny type. We checked plots for seed germination and disturbance three times within 4 weeks after planting.
The second set of seed reintroduction occurred later in the same week of July 2003 after collecting new seeds from open-pollinated flowers. Planting protocols followed those at Cascade Head: we planted each new seed in a straw 3 cm from each previously planted, open-pollinated progeny. We planted five new seeds in each of 30 plots at both Jack Creek and Cone Peak, totaling 300 seeds.
Glasshouse germination studies: age, variety, and progeny type
In glasshouse experiments we compared seed germination among varieties, progeny types, and seed ages. Seed ages and origins were dependent on availability and time constraints but ranged from 6 mo to 8 years at the time of planting (N = 15–90 mechanically scarified seeds/trial). To compare recent seed viability in oraria and in east (Cone Peak-Iron Mountain) and west cascades populations (Jack Creek) of douglasii, we planted open-pollinated seeds in 2003 and 2004 that had been collected in 2002 and 2003 from at least eight different maternal parents in natural populations of each variety.
As part of the seed age experiment, we evaluated glasshouse germination for the same cohorts as those used in the seed reintroductions at Cascade Head, Cone Peak, and Jack Creek, including open pollinated progeny. In addition, for douglasii, we planted seeds collected from hand-, self- and cross-pollinations performed in 2002 at both Jack Creek and Cone Peak (N = 32 seeds/progeny type/site). For oraria, we compared cross- and self-pollinated progeny using 1996 seed originating from the same seed pool as in germination studies of Kephart et al. (1999a)
.
Multi-year trends in inbreeding for reintroduced population of oraria
In May 2003, as part of a 5-year census of the extant reintroduced population, we relocated all oraria plants in each of the three macroplots planted in October 1998 (Kephart, 2004
). To compare the fate of inbred versus outbred progeny, on 2 June we scored plants for survival, canopy area (size), and the number of reproductive stems, herbivore-browsed stems, open flowers, and fruits as in Kephart (2004)
.
Comparison of reintroduced and natural populations, oraria
Community composition
Vegetation surveys allowed us to quantify the differing landscapes of the reintroduction site and natural headland. We surveyed all aboveground plants in 1-m2 quadrats for a total of 12 random points along each of three 50-m transects in the reintroduction site, and along four 50-m transects in the natural headland, two in each of grassy and rocky habitats. We computed species richness (S) based on transects, direct observation, and pollinator plots. We used the proportional abundances of species to calculate species diversity (Shannon-Weiner index; H' = –
pi · lnpi).
Flowering phenology and pollinator visitation
We quantified flowering phenology for 9 weeks in reintroduced and natural populations. Prior to flowering, we marked 15 randomly chosen plants along four 50-m transects in the natural site totaling 30 plants each in rocky and grassy habitats. For these plants and the 19 surviving plants at the reintroduction site, we counted the number of stems in bud, flower, fruit, the number of open flowers, senescent flowers, and fruit. Pilot studies of flower duration indicated a flower life of 4–7 days (D. Lofflin and S. Kephart, unpublished data); thus, we used weekly counts of open flowers to compute total flower production per plant across the season.
We compared pollinator activity throughout the season at Cascade Head by regularly monitoring 33, 1 m2 observation plots that included at least one Silene plant. We located 26 plots in the natural site, with 14 in grassy sites and 12 in rocky habitat. In the reintroduction site, five of seven plots occurred in grassy and two in rocky areas. For each 10-min observation period (0500– 2300 hours), we noted the number of plants and open flowers of oraria, and quantified floral densities for other co-flowering species in the same plot. For each insect visit, we recorded the time spent on the flower, plant species, insect taxon or morphotype, and type of visit (pollen or nectar collection).
Fitness components and maladaptation
To assess plant fitness and the degree of maladaptation after reintroduction (sensu Husband and Campbell, 2004
), we measured diverse traits, including plant size (i.e., canopy area, as in Kephart, 2004
), as well as flower, fruit, and seed output. To estimate total fruit production on a per plant basis, we counted all fruits weekly on 79 study plants and at the end of the season. We quantified seed output across the flowering season by marking, on each study plant in June, a senescent flower, an open flower, and a young bud with vari-colored thread. At senescence, these marked flowers were bagged with fine mesh to retain the seeds. We collected bags of mature fruit in late July and compared the number of seeds, aborted ovules, and seed mass across the two study populations for early, middle, and late season flowers.
We estimated the percentage maladaptation (i.e., "M") of the reintroduced transplants to the new environment, by comparing mean trait values for these "alien" populations (A) against resident plants in the native, or "home" (H) environment; M = [|H – A| x 100]/A is comparable to quantitative genetics estimates of the selection differential for a population under selection relative to the initial population (Husband and Campbell, 2004
; B. Husband, University of Guelph, personal communication). The resident population is assumed to have the optimum value for which fitness is maximized in measured traits (e.g., leaf canopy area, fruit and seed number).
Seedling establishment
In the reintroduced population, and in each of two natural populations (Neahkahnie Mountain, Cascade Head), we censused both seedlings (<2 leaf pairs) and juveniles (
2 leaf pairs) in 0.25-m2 circular plots placed around 20 adults. We used the same protocols for seedling establishment as before, but in the reintroduction site, half of the plots occurred in places where transplants had died.
Statistical analysis
We used SPSS statistical software packages (SPSS, Inc., Chicago, IL, USA) for analysis, including nonparametric Kruskal-Wallis tests for comparing seedling numbers among varieties and seasonal differences in fruit and seed output and Mann-Whitney or t tests for evaluating differences between reintroduced and natural populations. Chi-square (2) tests were used to detect statistical differences in survivorship of inbred, outbred, and open-pollinated progeny and to compare greenhouse germination trials by year and variety. Prior to using parametric tests (e.g., ANOVA), we transformed counts or proportions as needed and used the Levine test to verify homogeneity of variances (Zar, 1996
).
RESULTS
Seed germination and establishment
Seedlings and juveniles in natural populations
We detected significant differences among varieties in the number of juveniles (Kruskal-Wallis test, 2 = 14.4, P = 0.002) and in the total numbers of seedlings and juveniles (Kruskal-Wallis, 2 = 14.4, P = 0.01), with rare oraria exhibiting the lowest densities (Fig. 1). Seedling and juvenile densities did not vary significantly among field populations within oraria (Mann-Whitney test, P > 0.05) or rupinae (Kruskal-Wallis test, 2 = 2.2–2.9, P > 0.05), however juvenile densities were significantly greater for East Cascades douglasii (Jack Creek) than for West Cascades populations (Iron Mt., Cone Peak; Mann-Whitney test, P < 0.01, Fig. 1). Juvenile and seedling numbers for east side douglasii and rupinae were statistically indistinguishable (Fig. 1, Mann-Whitney test, P > 0.05).
|
|
Glasshouse germination of the progeny of self- vs. cross-pollinations varied with variety and site: outbred progeny germinated at significantly higher rates than progeny of self-pollination (Fig. 3) for both oraria (75.9% vs. 50.6%; P < 0.01, t test) and for East Cascades douglasii (Jack Creek) populations (48.6% vs. 27.0%; P < 0.01, 2 = 7.8, N = 37 seeds/ progeny type). In contrast, the slight differences between germination of inbred and outbred seed progeny were insignificant for West Cascades douglasii at Cone Peak (Fig. 3; N = 35 seeds/progeny type). Field tests of the effect of inbreeding on seed germination and seedling survival were inconclusive due to the 0–1% emergence rates for the older, scarified seeds.
|
) shows a 22% overall decline in the last 2 years (2001–2003, Fig. 4), with only 12% of the original transplants of oraria still extant. However, new recruitment of seedlings and juveniles in reintroduction plots increases the current population of the reintroduction from 18 to 62% of its original size. Of the surviving transplants in 2003, outbred progeny exceeded the overall population trend for the reintroduction by 15.2% in 2001 and 8.5% in 2003 (Fig. 4); however, the higher 2003 survival of outbred progeny over plants derived from of open and self-pollination only approaches significance due to the small numbers of progeny remaining (df = 2, 2 test, P = 0.17).
|
53), the progeny of self and open pollination appeared to increase in canopy size, whereas plant size remained unchanged for outbred progeny (Fig. 5). Similarly, differences in reproductive stems and open flowers were not statistically significant over time or for open vs. inbred progeny after 2001 (P > 0.05, Kruskal-Wallis; Kephart, 2004
; and this study). However, a large percentage of the smallest 2001 plants (0–100 cm size) died in 2003 among selfed (56%) and open pollinated progeny (60%), while only 37% of small-sized plants died among outbred progeny. Of surviving plants in 2003, 25% (inbred) to 62.5% (outbred) produced flowers, and a larger mean proportion of flowers set fruit in outbred (23.2%) than in inbred (3.3%) or open-pollinated plants (12.0%) despite lower mean numbers of flowers and reproductive stems for surviving outbred progeny. Mean fruit and seed number/plant did not differ among progeny types in 2003, nor did we detect differences in mean number of fruits/plant in August 1999 vs. July 2003 (P > 0.05, 2 tests). Browsing was heaviest on plants with larger numbers of open flowers (Fig. 6) regardless of progeny type or year.
|
|
Flowering phenology
Flowering phenologies differed for natural and reintroduced populations in both the magnitude and dispersion of the floral displays (Fig. 7). The mean number of open flowers/plant in the natural population was 3.9 times greater at the first sampling date (4.38 ± 0.83) and 1.7 times higher at peak flowering (9.67 ± 1.5) than for reintroduced population. The flowering curve for reintroduced plants was also less variable in amplitude than in the natural site and peaked later in the season. Total season flower production/ plant did not differ significantly between sites (P = 0.193, Mann-Whitney).
|
Pollinator visitation
The most common insect visitors to oraria in 2003 were sweat bees (Halictidae) and hover flies (Syrphidae) in natural and reintroduction sites. Visitation rates of putative pollinators were low overall, but did not differ significantly for the reintroduced vs. natural populations (P > 0.05, two proportion z test) even though visitation rates were slightly lower for Silene plants in the reintroduced population (0.13 ± 0.04 visits/10 min) compared to the natural population (0.15 ± 0.02). Co-flowering species associated with Silene in the same plots also had lower visitation in the reintroduced relative to the natural population. Both halictid bees and syrphid flies delivered some pollen on each visit (40–50 grains/ visit), but only a small proportion of the available flowers were visited. We did not visit reintroduction plots at night, but nocturnal visitation is exceedingly low for natural populations of oraria relative to other varieties of S. douglasii. In fact, no visits have been recorded to oraria by the noctuid moths (e.g., Hadena variolata) that are regular crepuscular visitors to flowers of both rupinae and douglasii (S. Kephart, unpublished data).
Fitness traits and maladaptation
For both vegetative and reproductive traits, plants from the natural headland (home environment) yielded higher estimates of fitness than did progeny transplanted to the adjacent, formerly grazed site of the 1998 reintroduction (Table 2). Estimates of the fitness differential between plants in both areas gave values of maladaptation ranging from 49–179%. Leaf canopy area was highest for plants on home soils, approaching significance (P = 0.06, Table 2). In terms of reproductive potential, total ovule number did not differ among populations (P > 0.05, Mann-Whitney), but we detected significantly higher fruit set in the natural relative to the reintroduced population (Table 2, P = 0.001, Mann-Whitney). Seed number per capsule also differed significantly between the natural and reintroduced sites (P = 0.05, Mann-Whitney), but the number of unfilled seeds, aborted seeds, and ovules did not vary between them (P > 0.05, Mann-Whitney), nor did seed or ovule number differ among early, mid, and late season flowers (P > 0.05, Kruskal-Wallis).
|
|
Reintroduction is a relatively recent and potentially important restoration tool, yet few case studies have examined the interaction of genetic and demographic factors in evaluating seed and transplant reintroductions of herbaceous perennials. Our analysis depicts differential seedling and juvenile abundances in natural and experimentally augmented populations of rare and common varieties of S. douglasii and further confirms a high success rate for outbred progeny (Kephart, 2004
), By identifying the important variables affecting plant fitness and survival, this research will facilitate the design of larger scale attempts at successful reintroduction of S. douglasii and other herbaceous perennials.
Natural seedling establishment in rare and common varieties
Our first goal was to evaluate the potential relationship between rarity and seedling establishment in natural and experimentally manipulated populations of rare and common varieties. In comparing natural populations differing in geographic range and rarity, we found significantly lower total densities of juveniles in rare oraria than in the two more common varieties, rupinae and douglasii. In addition, only in oraria did seedling densities exceed that of juveniles, indicating that its seedlings may have a comparatively lower chance of surviving to the juvenile stage. Outplanting of current year seeds in natural populations also revealed germination and establishment rates that were four times higher for populations of douglasii than for oraria, despite glasshouse germination levels in oraria that were equivalent to or exceeded those of other varieties.
The disparity among varieties in potential seedling recruitment may be due to inbreeding depression. Cumulative inbreeding depression for glasshouse-grown progeny of oraria at Cascade Head is 76% (Kephart et al., 1999a
), but levels are not yet known for rupinae and douglasii. In Silene regia, whose status is endangered or threatened in five states, Menges (1991)
found lower seed germination in small populations, presumably from an increase of inferior, geitonogamously produced seed associated with reduced interplant movements by hummingbirds. Fruit and seed set in oraria is limited by low pollen deposition arising from insufficient pollinator visitation (Brown and Kephart, 1999
): a key nocturnal pollinator (Hadena variolata) that is prevalent in rupinae and douglasii is absent thus far in oraria populations (S. Kephart et al., unpublished data). Such limited pollen dispersal may compound the localized gravity-dispersed seed from dry capsules creating a greater potential for inbreeding. That low pollen deposition is associated with diminished seedling survival is supported by significantly higher 60-d glasshouse survival of progeny derived from oraria flowers that received supplemental pollen in the field (Brown and Kephart, 1999
).
We did not detect significant differences in seedling emergence between plots with cleared or clipped vegetation, but the role of competition in seedling establishment merits further study. For instance, the highest natural abundance of juveniles occurred in populations of rupinae, which occurs in very rocky habitats, often with few other plants in close proximity. A transition, matrix-based demographic study of oraria revealed higher juvenile recruitment of oraria in rocky than grassy microsites (Kephart and Paladino, 1997
). Other silenes may be vulnerable to competition as well; the density of prairie congener S. spaldingii was also negatively correlated with the cover of Festuca (Lesica, 1997
), and, in S. dioica, population decline appears related to low seed germination in the presence of late successional plants (Giles and Goudet, 1997
). Rocky localities are harsh habitats, but the advantageous trade-off may be lower competition in these early successional environments.
Fire may also influence seedling establishment in natural populations. Based on glasshouse germination and habitat differences, we expected low seedling numbers in dry East Cascades habitats (Jack Creek), but instead recruitment was highest at this site. Unexpectedly, a forest fire burned through this site within weeks after planting the seeds. Initial analysis of seedling emergence in fully burned plots (plastic straws melted, carbonized plant material present) vs. plots with only smoke damage suggests that smoke-induced germination may have augmented seedling numbers at that site. This result is supported by a 13% higher germination and a faster rate of emergence of Jack Creek seeds experimentally exposed to smoke compared to unexposed controls (L. Jefferson, Chicago Botanic Garden, unpublished data). In two other Silenes, burning led to either greater recruitment and reproduction (S. regia; Menges, 1995
) or higher recruitment and larger population sizes (S. spaldingii, Lesica, 1999
).
Inbreeding and reintroduction
Another goal of this study was to determine the extent to which inbreeding may limit the success of plant reintroductions via transplant or seed propagules. Several lines of investigation support a role for genetics in influencing demographic patterns. First, throughout the five-year comparison of oraria, outbred progeny had a higher probability of survival in the reintroduced population. Second, even in 2003, after severe population decline, a larger proportion of surviving outbred progeny produced flowers and fruits. Third, 75% of new seedlings occurred near existing outbred transplants, indicating potentially higher seed production or viability. In addition, glasshouse seed germination rates were significantly higher for outbred than inbred progeny at both Cascade Head, and for East Cascades populations of douglasii at Jack Creek. The Jack Creek population, like oraria populations, exhibits pollen limitation of seed set in some years (S. Kephart, unpublished data). Similarly, in Schiedea membranacea (Caryophyllaceae) outbred progeny had greater seed mass and germination than those derived from self-pollination (Culley et al., 1999
).
Despite higher overall survival and percentage reproduction, outbred progeny no longer exceeded other progeny in fitness (e.g., mean plant size, number of reproductive stems) by the fifth study year (Fig. 5). This shift was not statistically significant, which may reflect low population size, or the complexity of gene–environment interactions (Newman and Pilson, 1997
; Byers and Waller, 1999
). For example, inbreeding lowered seed germination, viability, and seed mass in Silene latifolia, but only under suboptimal conditions (J. White and P. Peroni, Davidson College, unpublished data). Purging of deleterious alleles can increase fitness in inbred populations (Hedrick and Kalinowski, 2000
; Tallmon et al., 2004
), but it requires multiple generations and is thus an unlikely explanation for shifts in the relative fitness of outbred progeny of oraria in later years. Our findings are most consistent with environmental and demographic stochasticity, processes that can mimic genetic drift in populations with low recruitment (Schaal and Leverich, 2004
). Once populations decline to small sizes, chance losses of particular individuals with high or low trait values can markedly affect population means. Because more small-sized plants died from 2001–2003 among inbred than outbred progeny, the few remaining large individuals had a larger impact on mean fitness of inbred progeny for the traits measured. Thus, chance and genotype-dependent processes may interact to influence population means.
Reintroduction as a restoration strategy
A third goal of our study was to evaluate the potential role of seed and transplant reintroductions as effective strategies for restoration. Seed bank studies can reveal sources of mortality for a reintroduced population that are not evident for transplanted plants, e.g., herbaceous competitors can easily impair survival of new seedlings yet have little effect on established individuals (Guerrant and Pavlik, 1997
). Our surveys detected the highest seedling densities in the reintroduced population; however, the low ratio of juveniles to seedlings in this population (1 : 4) compared to two natural sites (1–1.3 : 1) suggests that the survival of seedlings to juvenile stage is compromised there. The low juvenile recruitment might be attributed to the high proportion of introduced species in the reintroduction site (45% of species in survey, but see Menke and Muir, 2004
). Alternatively, soil disturbance while planting or monitoring reintroduced plants may create more habitat openings for new seedlings, which later succumb to a drier, less hospitable summer microclimate prior to reaching the juvenile stage. Variables for consideration in future reintroductions given their known effects on germination include the presence of adult plants, fire, and litter or nutrient levels (Menges, 1995
; Batty et al., 2001
).
Another limiting factor in the success of reintroductions of rare plants is longevity of seeds in buried seed pools or ex situ seed storage. The rapid deterioration of germinability with seed age for rare and common varieties of S. douglasii suggests that soil seed banks are unlikely to be a significant source of recruitment. Viable seeds over 10-years old occur in widespread S. latifolia, but even these seeds developed slowly, potentially placing them at a competitive disadvantage relative to younger seeds (S. Budischak and P. Peroni, Davidson College, unpublished data). In oraria, from glasshouse studies we expected 33% germination from the 7-yr-old seeds planted in June 2003 and 63% germination for current year seeds outplanted in February 2004 to the reintroduction site, yet at best we could detect only 2.2% seedling emergence for 1-yr-old seeds planted in the field. This trend implies a 4.3% decrease per year for ex situ collections and even greater losses of viability for in situ seed populations. The 2–8.5% values detected for all varieties of S. douglasii approximates the <1% to 6% range for establishment of diverse perennials dispersed as seed in outdoor plots (Drayton and Primack, 2000
). Thus, without massive plantings from recent and genetically diverse sources, seed reintroductions are not likely to provide enough progeny reaching reproductive maturity to sustain a population. For oraria and other rare perennials, such efforts may not be feasible because removing large numbers of seeds from a few remaining in situ populations could compromise natural seedling establishment to a far greater extent than the seed collection required to establish transplant populations. Seed propagules pose some advantages, but can require 1000s of seed dispersed in multiple years (Drayton and Primack, 2000
).
Another measure of reintroduction effectiveness is the extent of maladaptation of the reintroduced population for characters affecting fitness, either those directly influencing growth (e.g., leaf biomass, canopy size, photosynthetic rate) or reproductive success (e.g., seed production). Even for single traits, few studies estimate levels of maladaptation for newly established vs. source populations (Husband and Campbell, 2004
), but the 49%, and nearly significant, deviation we observed from the optimal phenotype for leaf canopy area closely approximates the mean (53%) for plants outplanted from nearby source populations to similar habitats (reviewed in Husband and Campbell, 2004
). Estimates of maladaptation for seed and fruit set in oraria reintroductions are even higher (85–179%) but still fall below means for populations altered significantly by human activity (254%, Husband and Campbell, 2004
). The sources of maladaptation in oraria populations are likely both ecological (i.e., given the higher species diversity, richness, and native composition of the natural population) and genetic because the larger source population also exhibits considerable selfing (S = 0.34–1.6) and inbreeding depression (70–85%; Kephart et al., 1999a
). Higher than expected levels of inbreeding appear associated with low pollinator service in Scutellaria montana as well (Cruzan, 2001
).
Overall, most fitness components were either lower or comparable in value for transplanted progeny in the reintroduced site relative to the natural population, with clear differences in multiple variables (canopy area, fruit and seed set, numbers of juveniles). Although reintroduced transplants had similar levels of pollinator visitation and the same potential seed set as the natural population (i.e., ovule number), fruit and seed set were significantly lower suggesting fitness may be limited by a site-specific variable (e.g., invasive species competition, differing abiotic microenvironments, etc). Thus, population and community restoration to levels equivalent to the natural site will likely require long-term management goals that reduce invasives while augmenting native species diversity. The strong associations between diversity, productivity, and community function (Tilman et al., 1996
; Hille Ris Lambers et al., 2004
) imply that conservation biologists might invest more effort in designing simultaneous reintroductions of species that co-occur in the natural populations serving as models for restoration.
Low seed reintroduction success, as well as the declining population size and reduced fitness of the transplants all suggest that the reintroduction has not fully met the requisite parameters for a viable population (Pavlik, 1996
). Although the mean size of reintroduced plants increased over time and the new seedlings detected in our surveys raise the extrapolated population size (seedlings, juveniles, and adult transplants) to a potential of nearly 100 plants, not all young plants will survive to maturity. Also, over the 5 years, we found no new adults, and reintroduced transplants dropped to 12% of the original population, with numbers well below the 50 minimum viable population (MVP) set forth by the Center for Plant Conservation (1991). Interestingly, the mean 17.6% mortality/year approached mortality rates in the natural population (2–16.2%; Kephart and Paladino, 1997
). With numbers that are only stable to declining in the natural population (Kephart and Paladino, 1997
), however, the fecundity of reintroduced plants is probably insufficient for population persistence without additional reintroduction or management.
This investigation presents quantitative data on how demography and genetics interact in seed vs. transplant reintroductions of rare plants, as well as comparative data on levels of maladaptation to a new environment. High maladaptation is expected for non-local propagule sources, but it is influenced also by the genetic makeup of locally derived transplants, which showed reduced performance relative to the original source population. Present guidelines for plant reintroductions (e.g., Guerrant and Pavlik, 1997
; Vitt and Havens, 2004
) encourage broad, randomized collections of source material across habitats and years, from small and large individuals exhibiting genetic diversity for traits that affect plant fitness. Although resilient, self-sustaining populations are the ultimate goal of reintroduction, we hope this investigation will help instigate new scientific studies that link in situ differences in the success of reintroductions of rare and common plants to field surveys of quantitative traits and fitness components at each developmental stage. In plants and animals, we need more multi-year quantitative analyses that relate the genetic structure of reintroductions from source populations to in situ seedling, juvenile, and adult recruitment. In S. douglasii, inbreeding and/or rarity are closely associated with low in situ seedling emergence and juvenile densities, implying that genetic diversity is an important criterion for consideration in future reintroductions of herbaceous perennials. Low pollinator visitation is currently under study as an important factor contributing to inbreeding.
FOOTNOTES
1 The authors thank J. Butler, T. Culley, K. Theiss, and B. Pavlik for helpful comments on this paper and L. Rieseberg for suggesting in situ seedling studies. We had valuable field and logistical help from E. Christophersen, A. Lovejoy, L. Bernacchi, C. Brown, A. Countner, D. Pickering, K. Reuland, F. Russell, J. Shinn, M. Weeber, R. Murtha, D. Wentworth, and volunteers. Earthwatch Institute, Willamette University, The Oregon Nature Conservancy, and the M. S. Rogers Foundation gave financial support. 
2 Present address: 5133 Costabella Lane, Las Vegas, NV 89130 USA
3 Author for correspondence (e-mail: skephart{at}willamette.edu
)
LITERATURE CITED
Baskin C. J. Baskin 2002 Achene germination ecology of the federally threatened floodplain endemic Boltonia ducurrens (Asteraceae). American Midland Naturalist 147: 16-24[CrossRef][ISI]
Baskin C. 2003 Breaking physical dormancy in seeds: focusing on the lens. New Phytologist 168: 229-233
Batty A. K. Dixon M. Brundrett K. Sivasithamparam 2001 Constraints to symbiotic germination of terrestrial orchid seed in a Mediterranean bushland. New Phytologist 152: 511-520[CrossRef][ISI]
Bowles M. K. Jacobs L. Zettler T. Delaney 2002 Crossing effects on seed viability and experimental germination of the federal threatened Platanthera leucophalea (Orchidaceae). Rhodora 104: 14-30[ISI]
Brown E. S. Kephart 1999 Variability in pollen load: implications for reproduction and seedling vigor in a rare plant, Silene douglasii var. oraria. International Journal of Plant Sciences 160: 1145-1152[CrossRef][ISI][Medline]
Brook B. D. Tonkyn J. O'Grady B. Miller 2002 Contribution of inbreeding to extinction risk in threatened species. Conservation Ecology 6: 16 [online, accessed 17 May 2005], website, http://www.consecol.org/vol6/iss1/art16/
Budelsky R. S. Galatowitsch 1999 Effects of moisture, temperature, and time on seed germination of five wetland Carices: implications for restoration. Restoration Ecology 7: 86-97[CrossRef][ISI]
Byers D. D. Waller 1999 Do plant populations purge their genetic load? Effects of population size and mating history on inbreeding depression. Annual Reviews of Ecology Systematics 30: 479-513[CrossRef]
Center for Plant Conservation. 2001 Sampling strategies for genetic variation in ex situ collections of endangered plant species. In D. Falk and K. Holsinger [eds.], Genetics and conservation of rare plants. Oxford University Press, New York, New York, USA
Cochrane A. K. Brown S. Cunneen A. Kelly 2001 Variation in seed production and germination in 22 rare and threatened Western Australian Verticorida (Myrtaceae). Journal of the Royal Society of Western Australia 84: 103-110
Cruzan M. 2001 Population size and fragmentation thresholds for the maintenance of genetic diversity in the herbaceous endemic Scutellaria montana (Lamiaceae). Evolution 55: 1569-1580[CrossRef][ISI][Medline]
Culley T. S. Weller A. Sakai A. Rankin 1999 Inbreeding depression and selfing rates in a self-compatible, hermaphroditic species, Schiedea membranacea (Caryophyllaceae). American Journal of Botany 86: 980-987
Drayton B. R. Primack 2000 Rates of success in the reintroduction by four methods of several perennial plant species in Eastern Massachusetts. Rhodora 102: 299-331[ISI]
Ellstrand N. D. Elam 1993 Population genetic consequences of small population size: implications for plant conservation. Annual Reviews of Ecology Systematics 24: 217-242
Frankham R. M. Gilligan D. Morris D. Briscoe 2001 Inbreeding and extinction: effects of purging. Conservation Genetics 2: 279-284[CrossRef]
Giles B. J. Goudet 1997 A case study of genetic structure in a plant metapopulation. In I. Hanski and M. Gilpin [eds], Metapopulation biology: ecology, genetics, and evolution, 429–454. Academic Press, San Diego, California, USA
Gomulkiewicz R. R. Holt 1995 When does evolution by natural selection prevent extinction?. Evolution 49: 201-207[CrossRef][ISI]
Guerrant E. B. Pavlik 1997 Reintroduction of rare plants: genetics, demography, and the role of ex situ conservation methods. In P. Fiedler and P. Kareiva [eds.], Conservation biology, 80–108. Chapman & Hall, New York, New York, USA
Hamzé S. C. Jolls 2000 Germination ecology of a federally threatened endemic thistle, Cirsium pitcheri, of the Great Lakes. American Midland Naturalist 143: 141-153[CrossRef][ISI]
Hedrick P. S. Kalinowski 2000 Inbreeding depression in conservation biology. Annual Reviews of Ecology Systematics 31: 139-162[CrossRef]
Helenurm K. 1998 Outplanting and differential source population success in Lupinus guadalupensis. Conservation Biology 12: 118-127[CrossRef][ISI]
Helenurm K. L. Parsons 1997 Genetic variation and the reintroduction of Cordylanthus maritimus ssp. maritimus to Sweetwater Marsh, California. Restoration Ecology 5: 236-244
Hille Ris Lambers J. W. Harpole D. Tilman J. Knops P. Reich 2004 Mechanisms responsible for the positive diversity-productivity relationship in Minnesota grasslands. Ecology Letters 7: 661-668[CrossRef][ISI]
Huenneke L. J. Thomson 1995 Potential interference between a threatened endemic thistle and an invasive nonnative plant. Conservation Biology 9: 416-425[CrossRef][ISI]
Husband B. L. Campbell 2004 Population responses to novel environments: implications for ex situ plant conservation. In E. Guerrant, K. Havens, and M. Maunder [eds.], Ex situ plant conservation: supporting species survival in the wild, 231–266. Island Press, Washington, D.C., USA
Jones A. M. Hayes 2001 Increasing floristic diversity in grasslands: the effects of management regime and provenance on species reintroduction. Biological Conservation 87: 318-390
Kephart S. 2004 Inbreeding and reintroduction: progeny success in rare Silene populations of varied density. Conservation Genetics 5: 1-13
Kephart S. E. Brown J. Hall 1999a Inbreeding depression and partial selfing: evolutionary implications of mixed-mating in a coastal endemic, Silene douglasii var. oraria (Caryophyllaceae). Heredity 82: 543-554
Kephart S. C. Paladino 1997 Demographic change and microhabitat variability in a grassland endemic, Silene douglasii var. oraria (Caryophyllaceae). American Journal of Botany 84: 179-189[Abstract]
Kephart S. K. Sturgeon 1993 A new varietal designation in Silene douglasii. Madroño 42: 90-100
Kephart S. K. Sturgeon J. Lum K. Bledsoe 1999b Varietal relationships in Silene douglasii (Caryophyllaceae): morphological variability at the population level. Systematic Botany 24: 529-544[CrossRef][ISI]
Lande R. 1999 Extinction risks from anthropogenic, ecological, and genetic factors. In L. Landweber and A. Dobson [eds.], Genetics and the extinction of species, 1–22. Princeton University Press, Princeton, New Jersey, USA
Lesica P. 1997 Demography of the endangered plant Silene spaldingii (Caryophyllaceae) in Northwest Montana. Madroño 44: 347-358
Lesica P. 1999 Effects of fire on the demography of the endangered, geophytic herb Silene spaldingii (Caryophyllaceae). American Journal of Botany 86: 996-1002
Luijten S. J. Marckéry J. Oostermeijer J. Den Nijs 2002 Demographic consequences of inbreeding and outbreeding in Arnica montana: a field experiment. Journal of Ecology 90: 592-603
Maunder M. 1992 Plant reintroduction: an overview. Biodiversity and Conservation 1: 51-61
McGlaughlin M. K. Karoly T. Kaye 2002 Genetic variation and its relationship to population size in reintroduced population of the pink sand verbena, Abronia umbellata subsp. breviflora. Conservation Genetics 3: 411-420[CrossRef][ISI]
Menges E. 1991 Seed germination percentage increases with population size in a fragmented prairie species. Conservation Biology 5: 158-162[CrossRef][ISI]
Menges E. 1995 Factors limiting fecundity and germination in small populations of Silene regia (Caryophyllaceae), a rare hummingbird-pollinated prairie forb. American Midland Naturalist 133: 242-255[CrossRef][ISI]
Menges E. R. Dolan 1998 Demographic viability of populations of Silene regia in midwestern prairies: relationships with fire management, genetic variation, geographic location, population size and isolation. Journal of Ecology 86: 63-78[CrossRef]
Menke C. P. Muir 2004 Patterns and influences of exotic species invasion into the grassland habitat of the threatened plant Silene spaldingii. Natural Areas Journal 24: 119-128[ISI]
Montalvo A. S. Williams K. Rice S. Buchmann C. Cory S. Handel G. Naban R. Primack R. Robichaux 1997 Restoration biology: a population biology perspective. Restoration Ecology 5: 277-290
Newman D. D. Pilson 1997 Increased probability of extinction due to decreased genetic effective population size: experimental populations of Clarkia pulchella. Evolution 51: 354-362[CrossRef][ISI]
ONHP. 2001 Rare, threatened and endangered plants and animals of Oregon. Oregon Natural Heritage Program, Portland, Oregon, USA
Pavlik B. 1996 Defining and measuring success. In D. Falk, C. Millar, and P. Olwell [eds.], Restoring diversity: strategies for reintroduction of endangered plants, 127–155. Island Press, Washington, D.C., USA
Peroni P. R. Armstrong 2001 Density, dispersion, and population genetics of a Silene latifolia seed bank form southwestern Virginia. Journal of Torrey Botanical Society 128: 400-406[CrossRef]
Schaal B. W. Leverich 2004 Population genetic issues in ex situ plant conservation. In E. Guerrant, K. Havens, and M. Maunder [eds.], Ex situ plant conservation: supporting species survival in the wild, 267– 284. Island Press, Washington, D.C., USA
Sheridan P. D. Karowe 2000 Inbreeding, outbreeding, and heterosis in the yellow pitcher plant, Sarracenia flava (Sarraceniaceae), in Virginia. American Journal of Botany 87: 1628-1633
Tallmon D. G. Luikart R. Waples 2004 The alluring simplicity and complex reality of genetic rescue. Trends in Ecology and Evolution 19: 489-496[CrossRef]
Tilman D. D. Wedin J. Knops 1996 Productivity and sustainability influenced by biodiversity in grassland ecosystems. Ecology 78: 81-92[ISI]
USDA, NRCS. 2004 The PLANTS database, version 3.5 [online, accessed 3 March 2004], National Plant Data Center, Baton Rouge, Louisiana, USA, website, http://plants.usda.gov
Vitousek P. C. D'antonio L. Loope M. Rejmanek R. Westbrooks 1997 Introduced species: a significant component of human-caused global change. New Zealand Journal of Ecology 21: 1-16
Vitt P. K. Havens 2004 Integrating quantitative genetics into ex situ plant conservation and restoration practices. In E. Guerrant, K. Havens, and M. Maunder [eds.], Ex situ plant conservation: supporting species survival in the wild, 286–304. Island Press, Washington, D.C., USA
Walter K. H. Gillett 1998 IUCN red list of threatened plants. World Conservation Union, Cambridge, UK
Young A. T. Boyle T. Brown 1996 The population genetic consequences of habitat fragmentation for plants. Trends in Ecology and Evolution 11: 413-417[CrossRef]
Zar J. 1996 Biostatistical analysis. Prentice Hall, Upper Saddle River, New Jersey, USA
| ||||||||||||
