Abstract
Indian paintbrush, a parasitic member of the family Scrophulariaceae, is difficult to propagate due to dependence on an appropriate host plant species for long-term survival and health. A series of experiments identified prospective host species and procedures to improve viability for horticultural propagation of paintbrush (Castilleja spp.). Thirty-four paintbrush species were evaluated for tolerance to horticultural propagation techniques. Castilleja integra A. Gray, C. scabrida Eastw., and C. chromosa A. Nelson (Scrophulariaceae) demonstrated superior tolerance to transplanting and survived as seedlings in flats and pots when grown without a host. In subsequent preliminary and culminating experiments, compatibility of paintbrush/host species combinations was evaluated using a range of production conditions. Castilleja integra plants survived better after pot-to-pot transplanting with a host species when larger seedlings were utilized. Direct-seeding paintbrush and host plants into a common pot was approximately half as effective for establishing healthy C. integra compared with transplanting the species together after germination in independent flats. Castilleja integra showed a strong host-specific survival and vigor response. Higher levels of paintbrush survival, plant vigor, and aesthetic compatibility occurred when C. integra was paired with Artemisia frigida Willd. (Asteraceae), Artemisia michauxiana Besser (Asteraceae), Ericameria nauseosa Willd. (Asteraceae), Penstemon pinifolius Greene (Scrophulariaceae), Calylophus serrulatus (Nutt.) P.H. Raven (Onagraceae), Eriogonum strictum Benth. (Polygonaceae), and Eriogonum jamesii Benth. (Polygonaceae).
Horticulturally compatible paintbrush/host pair involving Castilleja integra and Penstemon pinifolius.
Indian paintbrush (Castilleja spp.) offers strong market potential for the landscape nursery industry provided plants can be successfully and consistently propagated and grown in pots to saleable size. Consumer acceptance of paintbrush as a landscape plant is evidenced by the results of a Montana homeowner survey, wherein 65% of survey participants indicated they were “very likely” to incorporate native plants in home landscapes as replacements for more traditional exotic materials (Zadegan and others 2008). Among the plants mentioned as desirable in the survey was paintbrush, a genus of plants many people in the survey recognized by name and wished to grow.
Species in the genus Castilleja are known to be very difficult to propagate as potted plants, making them less desirable for the retail nursery market (Meyer and Carlson 2004; Backes and Hoch 2010). Propagative recalcitrance is thought, in part, to be the result of Castilleja species’ parasitic nature, a characteristic documented by Heckard (1962). Subsequent research affirmed this parasitism and better defined paintbrush relationships with parasitized companion species (Malcolm 1966; Dobbins and Kuut 1973; Montes-Hernández and others 2015).
Paintbrush is considered to be a hemi-parasite rather than an obligate parasite because plants from this genus can germinate without a host and photosynthesize to produce much of their own energy. Yet, researchers have theorized that parasitism is beneficial for mineral and water uptake, resulting in vigorous growth and increased sexual reproduction (Heckard 1962; Malcolm 1966; Matthies 1997). Further research has shown that paintbrush derives a secondary benefit from many parasitized hosts, that is, the uptake of anti-herbivory compounds such as alkaloids that are associated with tolerance to pests (Stermitz and others 1993; Marvier 1996; Adler 2003).
Castilleja species have been categorized as “generalist” parasites, meaning paintbrushes can establish successful parasitic relationships with a wide range of host species (Heckard 1962; Clancy and others 2013; Montes-Hernández and others 2015). A number of studies, however, have shown that paintbrush response is both parasite- and host species-specific (Matthies 1997; Adler 2003; Nelson 2005). Nelson (2005) tested 8 Penstemon species against 5 Castilleja species and found the best host to be Penstemon strictus Benth. (Scrophulariaceae).
Although revelatory research has helped define the unique parasitic characteristics of Castilleja species, essential information is lacking that will allow commercial-scale production of this species. Determining efficacious host/parasite species pairs for Castilleja propagation has been explored to a limited degree (Heckard 1962; Clancy and others 2013; Montes-Hernández and others 2015), but more research is needed to identify aesthetically compatible pairs that can be successfully grown in pots and ultimately outplanted to the garden.
Our research refines Castilleja propagation procedures through the use of paired hosts as horticultural tools. Specifically, the following questions were posed: Which paintbrush species are most amenable to culture in a garden setting? Which host species are most efficacious for producing healthy paintbrush plants in pots? Is it better to establish host/paintbrush pairs by direct-seeding into pots or by paired transplanting? Which host species are optimal for survival of paintbrush plants after outplanting to the garden?
METHODS
Paintbrush Species Evaluation: A Prelude Experiment
Beginning in 2005 and continuing through 2014, 34 species of Castilleja (Table 1, includes full nomenclature) were evaluated to determine their tolerance to horticultural propagation protocols. Castilleja seeds were collected from the wild or obtained from commercial suppliers. Seeds were stratified and planted using procedures described by Luna (2005) and Meyer and Carlson (2004). Following 8-wk stratification at 4.4 °C (40 °F), seeds were planted in flats and germinated in a climate-controlled greenhouse. Daytime greenhouse temperature was set for 27 °C (80 °F) and nighttime temperature for 13 °C (55 °F). Seedlings were transplanted into cone trays (cells 5.1 cm wide × 10.2 cm deep [2 × 4 in]) and allowed to grow for 6 to 8 wk, after which we planted survivors into a common garden.
Taxonomic label, relative germination rate, and survival of Castilleja species during greenhouse and garden transplanting.
Field evaluation of Castilleja species was conducted on the Aberdeen Research and Extension Center, Aberdeen, Idaho. Soil at the research site consisted of a Declo silt loam with a pH of 8.4. The center is located in the high, dry desert of southeastern Idaho. The region is within USDA Hardiness Zone 5 and has annual precipitation of 233 mm (9.2 in). The evaluation field was provided with supplemental water to bring the total of rainfall and irrigation to approximately 381 mm (15 in).
Individual plots consisted of up to 40 plants (final number depended on germination rate) arranged in rows. Plots were unreplicated and treated like breeding blocks. During the propagation process, observations and measurements were made of germination, survival during flat-to-pot transplanting, and survival following outplanting to the field. During the propagation process, paintbrush plants were isolated from potential host plants, thus allowing assessment of host-free propagation potential through traditional nursery-adopted bedding plant production methods.
Evaluation of Host Performance and Transplant Size: Prelude 2012 Experiment
In June 2012, a prelude host-response experiment was initiated to evaluate plants of Castilleja integra in association with 12 potential host species (Table 2, includes full nomenclature). Seeds of C. integra were cold-stratified for 8 wk in preparation for planting. Seeds of potential host species were stratified for 3 wk (Eriogonum brevicaule, Eriogonum douglasii, Eriogonum strictum, Erigeron compositus), 8 wk (Penstemon pinifolius, Penstemon confertus), or left non-stratified (Ericameria nauseosa, Poa secunda, Tetraneuris acaulis, Achillea millefolium, Sphaeralcea munroana, Agastache cusickii). Planting (in Fafard #2 soilless potting mix, Sungro Horticulture, Vilna, Alberta, Canada) of each seedlot was completed on 2 dates with an 8-d interval, resulting in larger early-planted seedlings and smaller late-planted seedlings. Seed was planted in 5.1 cm (2 in) pots and allowed to grow in a greenhouse for 1 mo before transplanting into larger pots, side-by-side with host species. At the time of transplanting, the early-planted paintbrush seedlings were approximately 10.2 cm (4 in) tall, and late-planted seedlings were approximately 5.1 cm (2 in) tall. A single plant of C. integra along with a single host plant were transplanted into black plastic gallon pots (Figure 1), filled with peat-based potting soil (Fafard #2, Sungro Horticulture), and placed in a greenhouse to grow. Daytime greenhouse temperature was set for 27 °C (80 °F) and nighttime temperature for 13 °C (55 °F). Six replicates of single pots were planted and arranged in the greenhouse in a completely randomized design.
Prelude host evaluation of 2012: paintbrush survival during establishment in greenhouse pots and after field outplanting.
Seedlings of Castilleja integra and Eriogonum brevicaule being pot-to-pot transplanted into a common container.
Individual greenhouse-grown plants were evaluated for survival 1 mo after transplanting and again in May 2013, 11 mo post-transplanting after seedlings were overwintered under non-heated conditions. Pots with surviving paintbrush plants were planted in the field at the Aberdeen Research & Extension Center in late May 2013 (Figure 2). On 15 July 2014 and 1 June 2015, paintbrush plants were evaluated for survival in the field.
Castilleja integra paired with host Eriogonum brevicaule in the 2012 host evaluation study 2 y after establishment in the field.
Evaluation of Host Performance and Direct Seeding: Prelude 2013 Experiment
A second prelude experiment was initiated in early April 2013. Seeds of 6 prospective host species (Table 3) and C. integra were non-stratified (Agastache cusickii, Artemisia frigida, Poa secunda) or cold-stratified for 3 wk (Eriogonum brevicaule) or 8 wk (Penstemon pinifolius, Calylophus lavandulifolius). Ten pretreated seeds of both host and paintbrush were direct-seeded into a black plastic gallon pot filled with a peat-based potting soil (Fafard #2, Sungro Horticulture). Paintbrush and host plants were allowed to germinate and grow together. After plants were established, pots containing 3 or more plants of either host or paintbrush were thinned to a maximum of 4 plants (2 paintbrush and 2 host plants). Six replicates of single pots were arranged in a climate-controlled greenhouse in a completely randomized design. During this experiment, greenhouse daytime temperature was set for 27 °C (80 °F) and a nighttime temperature for 13 °C (55 °F).
Prelude host evaluation of 2013: paintbrush survival and vigor during establishment in greenhouse pots and after field outplanting.
Individual paintbrush plants were evaluated for survival and vigor in the greenhouse 2 mo after seeding. Paintbrush plants that survived the 15 mo of culture in the greenhouse were outplanted to the field at the Aberdeen R&E Center in late May 2014 (Figure 3). During June 2014, June 2015, and June 2016, field-established paintbrush plants were evaluated for survival and vigor.
Preliminary paintbrush/host evaluation from 2013, 1 y after being outplanted to the field.
Evaluation of Host Performance: Culminating 2015 Experiment
The principal experiment for assessing host plants for paintbrush production was initiated late March 2015. This experiment incorporated several propagation concepts learned from the prelude studies. Superior hosts identified in the paintbrush species evaluation trials were included in this experiment. Additional host species were also included, with emphasis on those lacking seed pretreatment requirements, which was a logical concept designed to simplify the commercial production process. Castilleja integra and prospective host plants (Table 4) were germinated in separate flats and later transplanted as species pairs into square 10.2 cm (4 in) pots in the greenhouse. During this experiment, greenhouse daytime temperature was set for 27 °C (80 °F) and nighttime temperature for 13 °C (55 °F). Paintbrush plants were allowed to grow to a height of approximately 10.2 cm (4 in) prior to transplanting. Two plants each of paintbrush and host species were transplanted into a single pot in an attempt to enhance the ultimate number of pots with a healthy pair of species.
Culminating host evaluation of 2015: paintbrush survival during establishment in greenhouse pots and after field outplanting.
Eleven species of potential host plants were chosen for this experiment (see the list of species in Table 4 and Table 5). Seeds of Castilleja integra, Penstemon virens, Liatris spicata, Antennaria media, and Campanula rotundifolia were stratified for 8 wk prior to seeding into germination flats. Seeds of Eriogonum jamesii and Penstemon ramaleyi were stratified for 3 wk. Seeds of Artemisia frigida, Artemisia michauxiana, Erigeron elatior, Tetraneuris scaposa, and Zinnia grandiflora remained unstratified. A 12th treatment was created by transplanting paintbrush plants without a prospective host. Host and paintbrush species were seeded into individual, separate flats. When plants reached a height of approximately 10.2 cm (4 in), 2 paintbrush and 2 host plants were transplanted together into 10.2 cm (4 in) pots (Figure 4). The choice of size and style of pot was based on common usage in the industry. Growing medium was a peat-based potting mix (Fafard #2, Sungro Horticulture) that was modified to improve drainage by adding 25% (by volume) horticultural perlite. Three replicates of 10 pots each were planted and arranged in the greenhouse in a completely randomized design. During this experiment, greenhouse daytime temperatures were set for 27 °C (80 °F) and a nighttime for 13 °C (55 °F).
Culminating host evaluation of 2015: paintbrush vigor as measured by plant height (cm) in the greenhouse and total stem length (plant height x number of stems) after field outplanting.
Arrangement of paintbrush (C. integra) and host (Antennaria media) plants in greenhouse pots in the 2015 culmimating host evaluation experiment.
Individual paintbrush plants were evaluated for survival and plant height in the greenhouse 1 mo and 3 mo after transplanting. When plants in the pots were of saleable size (10.2–15.3 cm tall [4–6 in]), we calculated the percentage of pots with marketable, vigorous plants of both paintbrush and host (Figure 5).
Culminating paintbrush/host evaluation experiment of 2015 at the point in time when we assessed percent saleable greenhouse-grown plants.
Pots containing live paintbrush plants were outplanted into a field at the Aberdeen R&E Center in late May 2015. During June 2016 and July 2016, paintbrush plants in the field were evaluated for survival and vigor. Vigor was assessed by measuring plant height and multiplying this value by the number of stems. During bloom in 2016, surviving paintbrush/host pairs were rated for aesthetic value (mutual appearance-based bloom color complementation, compact host growth habit that enhanced paintbrush visibility, and competitive similarity between host and paintbrush) as a measure of species compatibility (Figure 6).
Aesthetically compatible paintbrush/host pair in summer 2016 involving Castilleja integra and Erigononum jamesii. High ratings of aesthetic value were associated with complementary flower color, compact growth habit that enhances paintbrush visibility, and competitive similarity.
Evaluation of Host Performance: Supplemental 2015 Experiment
Another greenhouse experiment was conducted in 2015 to provide more information about the performance of prospective host species. Seed preparation and propagation procedures were identical to those used in the previously described 2015 experiment. This supplemental 2015 experiment included a no-host control and 11 species of prospective host plants (Table 6). Prior to planting, seeds of Calylophus serrulatus, Eriogonum jamesii, Penstemon ramaleyi, Penstemon heterophullus, Horkelia fusca, Eriogonum umbellatum, and Geum triflorum were stratified for 3 wk. Seeds of Erigeron elatior, Gaillardia spathulata, Dalea purpurea, and Melampodium leucanthum remained unstratified. Each plot consisted of a single 10.6 cm square (4 in) pot replicated 10 times and arranged in a completely randomized design. Individual paintbrush plants were evaluated for survival and plant height in the greenhouse 2 mo and 3 mo after transplanting into a common pot, after which the experiment was terminated.
Supplementary host evaluation of 2015: paintbrush survival in the greenhouse after transplanting into pots with host species.
Data Analyses
Replicated, quantitative data for all studies were analyzed by simple analysis of variance using the Systat (Systat Software Inc, Chicago, Illinois) statistical program.
RESULTS
Paintbrush Species Evaluation: A Prelude Experiment
Preliminary evaluation of all Castilleja species was completed in a series of unreplicated, common gardens, making direct comparison of species performance inappropriate. Of the 34 Castilleja species included in the experiment, evaluations of 17 were limited to only 1 or 2 accessions, meaning assessment of propagation potential was partial by nature. Germination was nil from seeds for 10 of these 17 limited-observation species (see Table 1). In spite of design limitations, adequate information was generated to allow selection of some species that were responsive to seed-based propagation and transplantation.
The majority of Castilleja species evaluated showed limited or no germination, making them poor candidates for horticultural use (see Table 1). Seeds of Castilleja angustifolia, C. linariifolia, C. miniata, and C. sessiliflora tended to germinate at high rates, but plant survival declined during the propagation process and most died during pot-to-pot transplanting or pot-to-field outplanting. Three species, Castilleja integra, C. chromosa, and C. scabrida, showed moderate levels of survival during transplanting, even without the presence of host plants, providing evidence for good horticultural potential.
Evaluation of Host Performance and Transplant Size: Prelude 2012 Experiment
Larger paintbrush seedlings employed at transplanting resulted in statistically (P = 0.05) higher survival rates observed in the greenhouse and numerically superior survival rates in the field: large plants 90% compared to small plants 50% 1 mo after transplanting into pots in the greenhouse; 43% compared to 23% after 9 mo in the greenhouse; 27% compared to 17% after 1 y in the field; 21% compared to 10% after 3 y in the field.
Given the preliminary nature of the experiment—designed with a single plant per replication—and the qualitative nature of the survival measurements, insufficient data points were available to complete an effective chi-square distribution test as a statistical comparison. We present the host evaluation data, without statistical analysis, in Table 2. After 1 mo in the greenhouse, potential host species supported C. integra plants at a rate of 63% or higher.
Penstemon confertus and Tetraneuris acaulis were poor hosts for C. integra, and 100% of paintbrush plants hosted by these 2 species died in the greenhouse before pots were outplanted to the field. Poa secunda, Eriogonum strictum, and Erigeron compositus provided the highest levels of short-term survival for paintbrush plants in this experiment. Plants paired with P. secunda eventually died. Eriogonum strictum and Erigeron compositus were more effective as long-term hosts. Paintbrush plants grew very large and vigorous when paired with Achillea millefolium, but the aggressive nature of the host created a level of competition the paintbrush plants could not withstand on a long-term basis. Sphaeralcea munroana was also too aggressive in the field to serve as a good long-term host. The wild buckwheat species in this preliminary experiment (Eriogonum brevicaule, E. douglasii, and E. strictum) plus Ericameria nauseosa tended to be the best long-term host species for C. integra.
Evaluation of Host Performance and Direct-Seeding: Prelude 2013 Experiment
Direct-seeding, at best, was a marginal method for establishing host pairs in pots, partly because of incomplete germination of paintbrush and host seeds. Castilleja integra failed to emerge or survive as seedlings in 8 of the 42 pots. Conversely, in 11 of the pots, the host failed to emerge or died as seedlings. All plants died in 3 of the pots. Right from the beginning, 16 of 42 pots (38%) included in the experiment had no opportunity to develop into saleable products.
As with the 2012 experiment, the design of this experiment did not allow the application of statistical methods for assessment of survival for paintbrush growing over a range of hosts. Trends in the data, however, provided some insights as to paintbrush response to host species (see Table 3). Castilleja integra survived at 83% or higher with all host species through 3 mo in the greenhouse. As time went on, large differences in host-related response began to emerge. By the time the paintbrush plants were 3 y old and had been in the field for 2 y, there was no paintbrush survival within the no-host treatment or when C. integra was paired with Poa secunda. Long-term survival of C. integra was best when paired with Penstemon pinifolius and Artemisia frigida. Based on survival rates in the field, Calylophus lavandulifolius and Eriogonum brevicaule also served as acceptable long-term hosts.
Statistically significant differences were noted for host-dependent paintbrush vigor at each stage of evaluation (see Table 3). Castilleja integra plants paired with Artemisia frigida and Penstemon pinifolius expressed good vigor. Paintbrush plants associated with all other species tended to lack vigor throughout the experiment or declined in vigor after being established in the field. Although average vigor of C. integra plants hosted by C. lavandulifolius and Agastache cusickii was somewhat poor due to low scores contributed by non-survivors, paintbrush plants that did survive were healthy and attractive in the 2nd y in the field. Based on survival in the greenhouse and field, and on vigor of C. integra plants, Poa secunda was the poorest performer of the host species included in the 2013 experiment.
Evaluation of Host Performance: Culminating 2015 Experiment
Castilleja integra exhibited a wide range of survival responses to the 11 prospective host species (see Table 4). Liatris spicata, Penstemon virens, and Campanula rotundifolia failed to host C. integra adequately to maintain survival beyond the first few weeks of pot culture. A higher percentage long-term survival of paintbrush plants occurred when paired with Artemisia frigida, Artemisia michauxiana, and Eriogonum jamesii. Tetraneuris scaposa, Antennaria media, and Erigeron elatior provided good host support, 63 to 70% survival, for paintbrush plants during the first month in the greenhouse, but after 14 mo in the field, paintbrush plants showed 0 to 3% survival. Zinnia grandiflora and Penstemon ramaleyi provided only 17% paintbrush survival by the end of the experimental period.
During the first month in the greenhouse, we noted no significant differences in paintbrush plant height as a result of pairing with any prospective host species (see Table 5). Thereafter, paintbrush plant vigor, as indicated by plant height (greenhouse) and (or) total stem length (field), largely mirrored host-related survival. The most vigorous plants were paired with Artemisia frigida, Artemisia michauxiana, and Eriogonum jamesii. Moderate vigor was associated with pairings involving Zinnia grandiflora and Penstemon ramaleyi. The remaining host species failed to provide a parasitic relationship sufficient to keep C. integra plants healthy.
A higher percentage (43%) of saleable pots was evident when paintbrush plants were paired with Artemisia frigida or Artemisia michauxiana (see Table 5) as compared to other host/paintbrush pairings. Eriogonum jamesii, Zinnia grandiflora, and Penstemon ramaleyi provided 27%, 17%, and 20% saleable pots, respectively, and had comparatively higher levels of saleable pots in contrast to the remainder of prospective hosts. The remaining 6 species provided 3% or less saleable product.
All 11 prospective host species produced inadequate survival rates to be efficacious with respect to commercial propagation. In part, low survival may have been related to fluoride toxicity, an issue often blamed on the presence of perlite in potting mixes (Yang and others 2016). It was outside the objectives of this study to confirm the cause of toxicity symptoms, but significant barriers to normal growth were evident. Although this toxicity problem did not kill plants outright, it did weaken the plants during the first few months of growth in the greenhouse and seemed to slightly reduce the ability of paintbrush to establish a successful parasitic relationship.
Host/paintbrush pairs were rated for aesthetic value after a year in the field (Figure 7). The C. integra–Eriogonum jamesii pair had the highest aesthetic value ratings, followed by paintbrush paired with Artemisia frigida and Artemisia michauxiana. Penstemon ramaleyi and Zinnia grandiflora had aesthetic value ratings between 5 and 6 compared with Eriogonum jamesii that had an aesthetic rating of nearly 10. All other prospective host/paintbrush pairs had aesthetic value ratings less than 2, primarily the result of weak paintbrush plants and (or) poor paintbrush survival.
Subjective aesthetic value rating of paintbrush/host combinations. Paired combinations were rated 1 to 10, with 10 = best. High ratings of aesthetic value were associated with complementary flower color, compact growth habit that enhances paintbrush visibility, and competitive similarity.
Evaluation of Host Performance: Supplemental 2015 Experiment
Eriogonum jamesii, Penstemon ramaleyi, and Calylophus serrulatus supported C. integra with relatively high survival rates in greenhouse pots during this final experiment (see Table 6). Calylophus serrulatus was not only a good host but also had favorable flower color, compact stature, and appropriate competitiveness, making this species a potentially ideal host. Use of C. serrulatus as a commercially valuable host pair may warrant further investigation.
As with the culminating experiment, plants in this supplemental experiment exhibited minor issues with what appeared to be fluoride toxicity (Yang and others 2016). This injury affected survival of the paintbrush plants, meaning the potential of the evaluated hosts is likely better than what is reflected in these results.
Commercial Vetting of Propagation Procedures
The information garnered in our study was used to propagate attractive, vigorous paintbrush/host pairs for distribution through an industry partner. Castilleja integra was paired with Artemisia frigida, a species with good host potential as shown in our experiments. Castilleja integra and Artemisia frigida were germinated in separate flats before transplanting paintbrush/host pairs into 7.6 cm wide × 15.2 cm deep (3 in × 6 in) pots. For commercial propagation, pots were deeper than those employed in our experiments. Additionally, perlite added to the soil mix was screened prior to mixing to eliminate the fines and, thereby, reduce exposure to fluoride. Under these propagation conditions, 96% of paintbrush plants reached saleable size. Of pots containing paintbrush/host pairs, 93% survived overwintering in the greenhouse and were delivered as potentially profitable nursery products.
DISCUSSION
Lack of paintbrush survival and (or) very low plant vigor when plants were isolated from potential host plants supported the concept of parasitism in paintbrush as reported by Heckard (1962). The hemi-parasitic nature of paintbrush, as described by Malcolm (1966) and Matthies (1997), was affirmed in that some paintbrush plants emerged, and sometimes grew to establishment, in the absence of host plants. However, isolated paintbrush plants lacked vigor, usually died while young, and very rarely bloomed or produced seed.
Findings in our study, and experiences gained by growing a number of paintbrush species, do not fully support the concept expressed by Heckard (1962), Clancy and others (2013), and Montes-Hernández and others (2015) that Castilleja is a generalist parasite. For example, we found that C. integra can efficiently parasitize approximately 20% of the species evaluated as hosts. Another 20 to 30% of species tested served as moderately effective hosts. The remaining, approximately half of species tested, were poor hosts or even non-hosts. Our results are more in agreement with those reported by Matthies (1997), Adler (2003), and Nelson (2005) who documented a range of responses relative to species evaluated as paintbrush hosts. The evident inference is that a limited number of plant species will serve as effective hosts for propagating Castilleja species.
Of the prospective host species evaluated as part of this study, Artemisia frigida, Artemisia michauxiana, Ericameria nauseosa, Penstemon pinifolius, Calylophus serrulatus, Eriogonum strictum, and Eriogonum jamesii produced higher rates and longer terms of survival for partner C. integra plants. Any of these species could be used to develop marketable potted products and to help maintain healthy paintbrush plants in the garden. In the horticulture industry, however, aesthetics are as important as survival. Aesthetically compatible host species that did not overwhelm the paintbrush plants after field establishment included Artemisia frigida, Artemisia michauxiana, Penstemon pinifolius, Eriogonum jamesii, and Eriogonum strictum. Based on plant vigor, flower color, and plant texture, Calylophus serrulatus would possibly be another good paintbrush host, although we derived this assumption without a long-term field evaluation.
Useful general conclusions can be made with regard to prospective host species. Grass species, including Poa secunda, one of the host species of this study, and Festuca idahoensis Elmer (Poaceae), a species included in evaluations not reported here, were poor hosts for C. integra. Plants paired with grasses very seldom survived the first winter in the field and frequently did not survive shorter-term in greenhouse-grown pots. Wild buckwheat (Eriogonum) species varied widely in their ability to act as effective hosts. Buckwheat species that were efficient hosts produced moderate rates of paintbrush survival in pots, but excellent survival once established in a garden. Buckwheat species are typified by very slow seedling growth and a narrow taproot, both traits that could limit the ability of paintbrush to establish a timely parasitic relationship.
Observations suggest that most of the efficacious prospective hosts had fibrous root systems. Fibrous-root plants possibly provide better opportunities in pots for haustorial attachment by the paintbrush parasite and also create a stronger root-ball in the pots. During the outplanting process, paintbrush plants with disturbed or damaged root-balls rarely survived more than a few days.
Ease of propagation is an important consideration when choosing a host species for potted paintbrush production. Host species exhibiting rapid rates of emergence and lacking the requirement for stratification will allow a simpler and more efficient propagation system. Evaluated host species with these advantages include Artemisia frigida, Artemisia michauxiana, Calylophus serrulatus, Eriogonum strictum, and Eriogonum jamesii. These 5 species are easy to grow and aesthetically compatible, making them ideal hosts for Castilleja integra.
CONCLUSIONS AND APPLICATION
When paired with an appropriate host during propagation, paintbrush plants produce horticulturally viable nursery products.
Because of high germination rates, tolerance to independent seed propagation and transplanting techniques, and ability to survive short-term without a host, the most viable species for paintbrush production on a horticultural scale include Castilleja integra, C. chromosa, and C. scabrida.
Of the host species evaluated for supporting C. integra during horticultural propagation, Artemisia frigida, Artemisia michauxiana, Calylophus serrulatus, Eriogonum strictum, and Eriogonum jamesii performed better than did the others we tested.
Suggested guidelines for successfully producing marketable paintbrush plants include: 1) select a paintbrush species that is tolerant of common horticultural propagation techniques; 2) pair paintbrush with an appropriate host species; 3) transplant pairs into common pots rather than direct-seeding; 4) grow paintbrush seedlings to a height of approximately 10.2 cm (4 in) before pairing them with hosts in a common pot; and 5) place multiple paintbrush and host plants in a single pot to improve the number of pots with surviving host/parasite pairs.
Untested procedures that may help produce a high percentage of marketable plants include transplanting pairs into deep pots and avoiding the use of soil mixes with a high percentage of small-diameter perlite particles (possibly a source of fluoride toxicity).
ACKNOWLEDGMENTS
Thanks to Ben Pierce for his diligent maintenance of greenhouse facilities and research sites and for timely data collection. This research was funded by grants from the Idaho State Department of Agriculture: Nursery, Landscape, and Floral Program and by USDA Hatch Funds.
Footnotes
Photos by Stephen L Love and Tony A McCammon
