DEMOGRAPHIC ANALYSISOF
ERIGERON DECUMBENS VAR. DECUMBENS,
AN ENDANGERED PLANT SPECIES OF
THE WILLAMETTE VALLEY, OREGON
1999 Field Studies
Prepared by
Deborah L. Clark
Prepared for
U.S. Fish and Wildlife Service
Western Oregon NWR Refuge Complex
Order No. 1448-13590-9-M047A
February 2000
Erigeron decumbens var. decumbens1Clark, D.L.
SUMMARY
Erigeron decumbens var. decumbens, known only from the Willamette Valley of western Oregon, is listed as Endangered by the U.S. Fish and Wildlife Service and by the state of Oregon. In 1993 a long-term population monitoring program for Erigeron decumbens var. decumbens was initiated at Baskett Slough National Wildlife Refuge (Polk County). The objectives were to assess long-term population trends, particularly patterns of reproduction, recruitment, and mortality.
This report presents the results of the demographic monitoring from 1993-1999. Considered together the following demographic parameters indicate that the measured population at Baskett Butte (Baskett Slough National Wildlife Refuge) is declining.
1. The population size decreased every year but one with the 1999 population only 48% of the 1993 population.
2. Increasing mortality rates exceeded recruitment rates every year but one.
3. The percentage of the population that flowered generally decreased every year.
4. The average number of heads per flowering plant decreased each year beginning in 1995 to 3.4 heads per flowering plant in 1999.
5. The pattern of average basal area per plant varied but the 1999 average basal area is only 63% of that measured in 1993.
To stabilize the population, adaptive management strategies need to be developed in conjunction with experimental approaches that will determine the factors responsible for the population decline. A second approach to aid in the recovery of Erigeron decumbens var. decumbens is to increase the population by transplanting cuttings or plants grown from seed under greenhouse conditions or by sowing seeds in the field under more controlled conditions. It is strongly recommended that future efforts to stabilize the population at Baskett Slough National Wildlife Refuge be coordinated with other agencies in the Valley who are also responsible for management of Erigeron decumbens var. decumbens populations.
INTRODUCTION
Erigeron decumbens var. decumbens Nutt., the Willamette daisy, is known only from the Willamette Valley of western Oregon. Herbarium records from the 1800’s through the early 1900’s indicate that the Willamette daisy was once common and widely distributed throughout the Willamette Valley (Clark et al. 1993). However, between 1934 and 1980, this taxon was not observed or collected and was considered to be extinct until its rediscovery in 1980 at Willow Creek (Lane County) by James Kagan, and at Finley National Wildlife Refuge (Benton County) by Robert Meinke (Clark et al. 1993). Only 18 known populations remain in the Willamette Valley, ranging from Grand Ronde in the north to Goshen in the south (Clark et al.1993). Most populations are small, consisting of fewer than 50 plants, with the largest concentration of sites occurring in the wetland prairies of west Eugene. Erigeron decumbens var. decumbens is listed as endangered by the U.S. Fish and Wildlife Service and by the state of Oregon (Oregon Natural Heritage Program, 1995).
Erigeron decumbens var. decumbens is a perennial herb, 15-60 cm in height, with erect or somewhat prostrate, upward-curving leaves. The basal leaves are mostly linear (5-12 cm) and narrow in width (<1 cm) (Clark et al. 1993). It appears to spread vegetatively, often forming large clumps (Clark et al. 1993). Flowering stems appear primarily in June and July, producing 1-5 heads, each of which is daisy-like with pinkish to pale blue ray flowers and yellow disk flowers. Of six observed pollinator species, the native butterfly Phyciodes campestris (the field crescent), commonly found in the Northwest, appears to be the most effective pollinator, with syrphid fly Toxomerous occidentalis also important (Jackson 1996). The variety of insects visitors suggest that Erigeron decumbens var. decumbens is a generalist in regard to its pollinators (Jackson 1996). However, these pollinators tended to favor Erigeron decumbens var. decumbens, and were frequently observed traveling around flowers of other species to reach Erigeron decumbens var. decumbens, suggesting the insects are specialists in regard to the food sources they select (Jackson 1996).
Wind-dispersed seeds, produced in July and August, are small with a pappus consisting of capillary-like bristles (Clark et al.1993). The estimated average dispersal distance is 94.24 cm (Jackson 1996). This estimate does not include convection, nor the possibility of rare long-distance transport, which is almost certain to occur in some circumstances (Jackson 1996). The presence of a persistent seedbank for Erigeron decumbens var. decumbens is unknown.
Laboratory studies indicate, that despite abundant achenes, few seeds are filled (<11%) and only about 65% are viable (Clark et al. 1997). Germination under laboratory conditions required seed coat scarification (Clark et al.1997). Seed germination in pots placed outdoors occurred in April and May (Clark et al. 1997).
The primary habitat for Erigeron decumbens var. decumbens is wetland prairies, dominated by Deschampsiacaespitosa (tufted hairgrass). Seasonal flooding occurs in these low, flat regions, creating anaerobic and strongly reducing soils during winter and spring months (Clark et al. 1993, Finley 1994). The Willamette daisy also grows at a few upland prairie sites, which have moderate to well-drained soils and are characterized by a mix of native bunchgrasses, including Festuca roemeri, Bromus carinatus (California brome), and Elymus glaucus (blue wild rye). Scattered Quercus garryana (white oak) are often present. One such population of Willamette daisy exists at the summit of Baskett Butte, within Baskett Slough National Wildlife Refuge south of Monmouth. Erigeron decumbens var. decumbens usually occurs were woody cover is slow or absent, and where the cover of herbaceous vegetation is low relative to surrounding areas (Clark et al.1993). It has not been considered to be the dominant species at a site.
Before the mid-1800’s the Kalapuya, indigenous people of the Valley, maintained both wetland and upland prairies by setting seasonal fires, facilitating hunting and gathering of food (Boyd 1986, Toepel 1991, Boag 1992). The frequent fires prevented widespread abundance of woody species. Today, native prairies are considered among the rarest of Oregon’s ecosystems and are refuges for many other rare species in addition to Erigeron decumbens var. decumbens (Noss 1995). Factors that contributed to the destruction of these unique prairies include urban and agricultural development, threats that are also prominent today (Johannessen et al. 1971, Towle 1982). Succession to shrublands and forests threaten most of the few remnants of undeveloped prairie. With the suppression of the large-scale fires in the early 1800’s, woody plants increased in abundance, eliminating suitable growing conditions for the prairie species (Habeck 1962, Johannessen et al. 1971, Towle 1982). In addition, weedy non-native species introduced with wide-spread settlement in the mid-1800s have spread rapidly, often dominating and suppressing native vegetation.
In 1993 D. Clark, C. Ingersoll, and K. Finley, funded by Oregon Department of Agriculture, completed the status report of Erigeron decumbens var. decumbens and initiated a long-term population monitoring program (Ingersoll et al. 1993). Permanent monitoring plots were initially established at three sites: Fisher Butte (Fern Ridge Research Natural Area managed by the U.S. Army Corps of Engineers), Bald Hill City Park (managed by City of Corvallis), Baskett Butte (Baskett Slough National Wildlife Refuge, managed by the U.S. Fish and Wildlife Service). The objectives of this monitoring program were to assess long-term population trends of Erigeron decumbens var. decumbens, particularly patterns of reproduction, recruitment, and mortality. Past results were reported in Ingersoll et al.(1995), Finley and Ingersoll (1995), and Finley (1998). This present report presents the results of monitoring during the 1999 field season at one of the sites, Baskett Slough National Wildlife Refuge (Polk County, Oregon).
METHODS
Study site
The study area for the monitoring is located to the southeast of the south peak (Mt. Baldy) of Baskett Butte, a promontory within the Baskett Slough National Wildlife Refuge. The dry slope with shallow rocky soil mapped as Chehulpam and Stiewer/Santiam series, supports a mixture of native herbaceous species Festuca roemeri and Danthonia californica and weedy non-native species such as Arrhenatherum elatius with encroachment by woody species Pseudotsuga menziesii, Rhus diversiloba, Rosa eglanteria, Rubus discolor, and Crataegus spp.
Study Design
Approximate boundaries of the Willamette daisy population at Baskett Butte were delineated in 1993. Macroplots, encompassing areas of somewhat homogeneous vegetation, were selected to represent the range of Erigeron decumbens var. decumbens densities at the site. One high-density, two moderate-density, and one low density macroplots were established (Appendix A). Within each macroplot, the number of Erigeron decumbens var. decumbens flowering clumps was estimated, then 44 individual clumps were selected at random. Each selected clump was included in a 1.0 x 0.5 m quadrat. Diagonal (upper right and lower left) corners of each quadrat were permanently marked with a metal ring stake and numbered copper tag.
Notes on the number of macroplots and quadrats in each macroplot are provided in Appendix A. Rough maps for relocation of macroplots and approximate locations of quadrats within macroplots are provided in Appendix B.
Data Collection
Measurements of Erigeron decumbens var. decumbens plants were made during the peak of blooming period, which varied between mid-June and mid-July, each year 1993-1999 with the exception of 1998. Using a PVC quadrat frame (0.5 m x 1.0 m) with a 25 cm wire grid, all Erigeron decumbens var. decumbens plants within a quadrat were mapped, and data collected indicating which plants had disappeared (died) and which plants were new since the previous year’s mapping. Recruits were not distinguished as to whether they were seedlings and vegetative sprouts.
An “individual” plant was defined as a basal clump at least 5 cm from the nearest neighbor. For each plant, the following measurements were taken: longest basal dimension (diameter 1), basal dimension 90 degrees from the first dimension (diameter 2),stem height (height to the tallest flowering head or leaf with the stem stretched out), number of flowering stems per plant, number of flowering heads per plant, and number of flower buds per plant. Because the number of flowering stems had a nearly 1:1 correspondence with the number of flowering heads, flowering stems were not counted after 1995 (Ingersoll et al. 1995). In addition, cover of the weedy non-native grass Arrhenatherum elatius (tall oatgrass) and cover of woody species as a group were measured in each quadrat in 1999.
Data analysis
The results from the summaries of population patterns apply only to the measured population and do not necessarily match patterns occurring throughout the entire biological population at Baskett Butte. However, quadrats were located in areas representative of the range of Erigeron decumbens var. decumbens densities.
Twelve of the original plots set up in 1993 could not be relocated in 1999 as a result of loss of corner stakes. The data for this 1999 report include only quadrats that could be relocated during the entire monitoring period from 1993-1999, and thus, summaries of results included in this report may differ slightly with previous reports. Occasionally, plants growing close together to one another were no longer distinguishable, and so they were combined in our data, but were not included in our assessment of mortality.
The population size was measured as the number of plants counted in the quadrats (n=32). Mortality was calculated as the percentage of the population of yeart that died the following year t+1, except for the 1997-1999 data, which measured the percentage of the population that died during a two year period. Because new plants were not distinguished as to whether they were seedlings or vegetative recruits, recruitment was calculated as the percentage of new plants measured in (year t+1) produced by the entire population of yeart except for the 1997-1999 data, which reflects recruitment over two years. Flowering was calculated as the percentage of the current population that had flowers or buds. The number of heads per flowering plant was calculated as the average number of flowers produced by flowering plants in the current population. The basal area per plant was calculated as an ellipse using the two diameter measurements for each plants. For analysis of population size structures, sampled populations were separated into nine size classes based on basal areas (Ingersoll et al. 1995).
RESULTS AND DISCUSSION
Demographic monitoring, which follows the fates of individual plants in a population and makes repeated in situ measurements, is used to determine whether a population has the potential for growth or stability in the near future (Pavlik 1994, Elzinga et al.1998). Demographic analyses can use either integrative or non-integrative approaches (Pavlik 1994). Integrative techniques generate a single numerical index indicating population stability, growth, or decline. Non-integrative approaches, which this report uses, rely on parallel interpretation of data from several demographic parameters, such as recruitment, mortality, and population structure. These parameters vary depending on the study and no single parameter is used to predict the stability of the population.
The results of the measurements of the demographic parameters made in this study are summarized below. Considered together they indicate that the measured population of Erigeron decumbens var. decumbens at Baskett Butte is declining.
Total population The measured population decreased every measurement period, except 1994-1995, with the current 1999 population only about 48% of the original population in 1993 (Table 1).
Mortality rates Mortality, with one exception, increased each year from 25.8% in 1993-1994 to 32.8% in 1996-1997 (Table 1). The mortality for 1997-1999 (64.5%) reflects the number of plants dying over two years instead of one, but it is still consistent with increasing mortality (Table 1).
Recruitmentrates The rate of recruitment varied from year to year, generally within 10% of the previous year (Table 1). As with the 1997-1999 mortality, the 1997-1999 recruitment (29.1%) measures the number of new plants over a two year period (Table 1).
Comparison of mortality and recruitment rates Mortality exceeded recruitment every year except 1994-1995 (Table 1).
Reproduction The percentage of the population that flowered decreased every year from 73.4% in 1993 to 32.0% in 1997 (Table 1). The percentage of flowering plants in 1999 increased to 41.4%, but is still only about 56% of the percentage of flowering plants in 1993 (Table 1). Although the percentage of flowering plants generally decreased each year, the mean number of heads per flowering plant initially increased from 5.5 heads in 1993 to 9.3 heads in 1995, but then steadily decreased to 3.4 heads per flowering plant in 1999 (Table 1).
Mean basal area per plant The mean basal area per plant increased each year between 1993 and 1996 from 5.6 cm2 to 11.1 cm2 but then decreased during the next two measurement periods to 3.5 cm2, only 63% of the size measured in 1993 (Table 1).
Population size structure More than 60% of the plants in 1999 had a basal area less than 1.0 cm2 (Table 2), a pattern which can indicate a stable or growing population. This pattern is similar to that reported in 1994 and 1995, at which time about 50% of the plants were in the two smaller size classes (Ingersoll etal.1995, Finley and Ingersoll 1995).
Recovery of the declining population of Erigeron decumbens var. decumbens depends on identifying and prioritizing the factors responsible for the population decline. Some of the demographic data already collected suggests possible mechanisms and thus, can form a framework for adaptive management approaches.
Patterns of mortality Of the plants that died between 1997 and 1999, most (60%) were smaller than 1 cm2 in basal area (Table 2). The pattern of mortality is generally that of small but consistent increases, suggesting mortality factors that increase with time such as increasing competition from other herbaceous or woody plants are the cause. However, cover of woody species and cover Arrhenatherum elatius(tall oatgrass) in 1999 were not significantly correlated with the 1997-1999 mortality rate (P=0.31, R2= 3.43% and P=0.71, R2 = 0.48% respectively).
Patterns of recruitment Most recruits (72%) measured in 1999 were <1cm2 even though some of them may have been two years old as data were not collected in 1998 (Table 2). The general decline in percentage of flowering plants (Table 1) is not translated into a similar pattern of decline in recruitment rates, which varied year to year (Table 1), suggesting recruitment is partially dependent on vegetative regeneration. Cover of Arrhenatherum elatius and woody species explained essentially none of the variation in yearly recruitment rates (P=0.92, R2=0.03%, and P= 0.49, R2=1.63%, respectively).
Patterns of sexual reproduction Flowering occurred with plants of all size classes, even those of the smallest size class (<0.2 cm2) (Table 2). Woody cover in 1999 showed a significant relationship with percentage of flowering plants in 1999 with woody cover explaining about 21% of the variation in the percentage of flowering plants (P=0.02, R2 =21.43%), suggesting competition from woody plants may be reducing the number of plants that flower. Cover of Arrhenatherum elatius in 1999 was not significantly related to the percentage of plants flowering in 1999 (P= 0.47, R2 = 2.37%).
The measured population at Baskett Butte is declining due mortality rates exceeding recruitment rates. One key to understanding the mechanisms for this pattern is identification of the roles of sexual reproduction and vegetative regeneration in maintaining the population. If sexual reproduction is more important in population recruitment than is vegetative regeneration, then managers should focus on strategies that promote flowering, seed production, and seedling establishment. In contrast, if vegetative regeneration is prevalent, then managers should focus on those factors which promote ramet production.
Experimental approaches are often the only way to identify the specific factors contributing to the population decline of rare plants (Pavlik 1994). However, adaptive management strategies based on our current understanding should be implemented in conjunction with these experimental approaches. As understanding increases of the processes involved in population growth based on data from the experimental approaches and management responses of Erigeron decumbens var. decumbens, management strategies can be revised or adapted.