Abstract
With their high reproductive capacity and disturbance-oriented life history strategies, Great Basin native annual forbs have the potential to be more successful in colonizing burned sagebrush sites than the later seral species that are commonly used in restoration. Seed of native annual forb species is usually not commercially available, however, and information on how to increase seeds of these species in agricultural settings is limited to nonexistent. Our objective was to start a small, native, annual forb production program, testing strategies to increase 6 species over 4 y. We found that seed of native annual forbs can be wild-collected in years with sufficient precipitation, and that they can be propagated using fall planting between rows of landscape fabric, without irrigation. Low-tech harvest methods, including vacuuming or sweeping off landscape fabric, and in some cases collecting and drying entire plants, can work for small-scale increases. For seed cleaning, a variety of sieves are key, and a seed blower can be helpful. We also present our plans for extending this work by testing the establishment of annual forbs in post-fire restoration projects. Our results demonstrate that seed production of native annual forbs can be achieved without specialized equipment or irrigation. For the majority of species, we were able to increase seed sufficiently so that it could be sown in larger agricultural-increase fields.
Collecting data on field trial plots. Photo by Alan de Queiroz
The Great Basin, a “sagebrush ocean” occupying most of Nevada as well as parts of Utah, Idaho, Oregon, and California, is a large expanse of undeveloped land where native shrubs still dominate many areas. Despite its drab appearance at first glance, this sagebrush ocean is an important source of species diversity and is particularly rich in endemics (Nachlinger and others 2001), making it a priority for conservation.
THE PROBLEM
Unfortunately, the Great Basin ecosystem also ranks highly for number of threats (Nachlinger and others 2001). A history of grazing mismanagement, as well as current threats from climate change (Bradley 2009), altered fire regimes, and invasive species are converting the sagebrush ocean into an ocean of exotic annual grasses, such as cheatgrass (Bromus tectorum L. [Poaceae]) and medusahead rye (Taeniatherum caput-medusae (L.) Nevski [Poaceae]). The rate of this conversion is accelerating, and it has dire consequences for fire cycles, wildlife, and soil ecology (D’Antonio and Vitousek 1992; Mack and others 2000).
In response to undesirable changes to natural resources, Great Basin public land managers have been seeding sagebrush areas since the 1940s. Seedings have had a variety of goals including forage production, soil stabilization, and wildlife habitat restoration (Pilliod and others 2017). Seedings initially focused on non-native grasses, with the proportion of native shrubs and grasses and, more recently, native forbs increasing over time (Pilliod and others 2017). Current efforts are also focusing on the importance of local genetics (Baughman and others 2019), though the availability of local seed for restoration is still limited.
Despite progress, seedings are often unsuccessful. A dry and variable climate is one barrier (Hardegree and others 2018), as is poor availability of locally sourced seeds for many Great Basin projects (Leger and Baughman 2015). Another problem is the rapid occupation of disturbed sites by cheatgrass, particularly after fires (Humphrey and Schupp 2004), which is transitioning Great Basin communities from shrubs, perennial grasses, and perennial forbs to annual grasses. As an annual, cheatgrass differs from historically dominant native perennial grasses and forbs in both life history and physiological traits. Cheatgrass uses soil water and nitrogen more efficiently than perennial species (Melgoza and others 1990; Lowe and others 2003), has an earlier phenology, and can have overwhelmingly high reproductive output. Thus, with slower, later seral life histories, seedlings of many Great Basin natives are not well-equipped to compete.
However, a group of native species have a life history similar to cheatgrass: native annual forbs. In the Great Basin, annual forbs readily occupy disturbed habitats in sites where their seedbanks persist (Figure 1). Prior to European settlement, they were likely an important component of post-fire communities, and we can infer from responses of sites that are closer to reference condition that native annuals can be an important component of post-fire recovery (Goergen and Chambers 2009). Experimental results have shown that these species are able to reduce the fecundity of annual weeds such as cheatgrass (Herron and others 2013; Leger and others 2014). Nonetheless, a combination of factors that include overgrazing, fire suppression, and aggressive invasion of exotic annuals has extirpated these weedy native annuals from many sites.
Amsinckia menziesii growing with cheatgrass near Reno, Nevada (left); honeybee visiting Amsinckia tessellata in our propagation field (right).
Germinating seeds (left) and seedlings (right) of Amsinckia menziesii.
Despite their apparent potential, native annuals have rarely been used in restoration and habitat improvement projects in the Great Basin. In a list of priority species for native seed production in Nevada, Davison (2002) did not include any annuals. Seed source development has focused on native perennials, and great strides have been made for many perennial forb species (Shaw and others 2005). Native annuals, however, have not been targeted for commercial development for at least 2 reasons. First, restoration objectives often include the creation of ground cover and forage within a short time frame, and Great Basin annual forbs are not robust in size nor good forage for livestock (although they may provide valuable resources for wildlife) (Dumroese and others 2015, 2016). Second, certain aspects of these species’ biology can make them difficult to grow commercially; producing their seed in quantity and at a reasonable cost may be seen as barriers.
Some recent priority shifts are increasing the profile of native annuals though: These species are an important food source for Sage-grouse chicks (Centrocercus spp. [Phasianidae]) and are also crucial nectar sources for pollinators (Figure 1). Concerns about Sage-grouse populations (Drut and others 1994; Rhodes and others 2010) and declines in pollinator populations have highlighted the need for diversification of native seed sources and, in combination with other concerns, led to the creation of the National Seed Strategy (Oldfield and Olwell 2015). Wildlife agencies are increasingly willing to contribute resources to increase the abundance of forb seeds in restoration mixes (Nevada Wildlife Action Plan Team 2012) and to support research to increase the availability of forbs for these mixes, including this project. Recognition of the potential value of native forbs has helped create an opportunity for our research group to make wild collections and to increase small quantities of seed for native annual forb species that appear to have restoration potential. While native shrub seed can usually be field-collected in quantities large enough to use in restoration, native forb seeds need to be wild-collected, then agriculturally increased to obtain sufficient seed (Shaw and others 2005). A critical step to ensure enough seed is available for large-scale production is to create small G1 (first-generation) stock seed, increased from wild-collected seed. The goals of our work were 1) to determine how to best wild-collect, clean, and increase seeds of a variety of native annual forbs to create stock seed; and 2) to create a list of best practices for conducting these activities for a variety of types of forbs.
PROPOSED SOLUTIONS
Wild Seed Collection
To obtain the seeds needed to begin the increase process, we collected wild seeds over 4 y in western and central Nevada (Table 1). We identified target species through field observations, selecting native annuals that we saw growing in quantities large enough to allow for wild seed collection. This was largely a subjective judgment but involved locating at least several hundred fruiting individuals in an area. We also targeted species that we saw co-occurring with invasive species, particularly cheatgrass. Using these native forb species, we performed research on their competitive abilities (Leger and others 2014) and germination requirements (Forbis 2010) to screen for cooperative and competitive species.
Species targeted for wild seed collections, including nomenclature, counties collected, years collected, and total weight of wild-collected seeds.
Native annual forbs present some unique seed collection challenges. Most wildland seed collection has focused on shrubs and grasses, species that vary in their seed production from year to year but are relatively consistent in appearing aboveground and producing at least some quantity of seed on an annual basis. In contrast, native annual forbs appear above-ground only in years with suitable timing and amount of precipitation, and in some years, it is impossible to obtain seed. Species’ requirements for timing and amount of precipitation are not well understood, but most of these annual forb species seem to require at least some fall to early winter precipitation, based on our observations.
Timing of seed harvest is also variable because it occurs at the onset of hot weather. This unpredictability makes planning for personnel and timing of fieldwork difficult. Therefore, wild collection of annual species usually requires multiple years. It is most efficient to concentrate collection efforts on species that are producing many seeds in a particular year, rather than spending a lot of time collecting few seeds from few plants in unfavorable years. Many native annuals are also indeterminate, meaning they mature gradually over time. Some have capsules or other fruiting structures that shatter, leading to seed loss, and some even have explosive dehiscence, meaning that seeds are flung away from the parent plant when capsules mature, all of which make seed collection challenging.
We used standard seed collection protocols for wild seed collection (Haidet and Olwell 2015), including taking less than 25% of seed to retain genetic diversity in wild populations. Seed-cleaning techniques were the same for wild collections as for harvested seeds from our seed increase and are discussed below (Table 2). Despite the challenges of collecting native annual seed, we were able to collect sufficient seed from all our species of interest for propagation.
Seed harvest and cleaning recommendations for species with different fruit and seed types.
Seed mixture for field trials.
Small-Scale Seed Production
We increased our wild-collected seeds at the University of Nevada, Reno Agricultural Experiment Station, in a field that had been used as an irrigated pasture before being tilled and fallowed. Therefore, this site had an extensive weed seedbank. For our small-scale efforts, we had neither planting nor harvesting equipment and were doing most tasks by hand. We did not use irrigation because we have observed these species growing in sagebrush sites near our field even in years with relatively low precipitation, and we assumed that irrigation would mean increased weed pressure. Also, if these species can be grown without irrigation, that fact might make them a more attractive option for growers who do not have ready access to water for irrigation or who lose their water allocation in drought years. While these species remain in the seedbank in wild sites in some years, we were able to grow them at the Agricultural Research Station without irrigation in all years of this study, which suggests crop stability to growers even without irrigation.
Plantings were done in October or November of each year, in anticipation of the typical timing for fall or winter storms in our region. We grew these species over 4 seasons: 2015–16, 2016–17, 2017–18, and 2018–19. Each year, we rotated species among different regions of the field to avoid mixing generations of the same species and to allow seeds to be certifiable.
For the first 3 y of the propagation project, we used row cover cloths (Figure 2). Row covers are used for establishing perennial forbs at the Oregon State University Malheur Field Station, which successfully increases many species each year (Shock and others 2015). Using row covers gives the seed and seedling a moister, warmer environment. We used Dewitt row cover cloth of 1 oz thickness that was 1.8 m (6 ft) wide.
Planting seeds and installing row cover cloth at the University of Nevada Agricultural Experiment Station.
In 2015–16, we planted a single row under each cover cloth. Seeds were sprinkled on the ground surface and tamped in by hand before covering with cloth. Our seeding rate for the first 3 y was consistent among species: We planted 7.6 seeds/m (25 seeds/ft). We changed the seeding rate for the 2018–19 year, at which time we doubled the rate for species that had had lower densities in the first planting year: Gilia inconspicua and Mentzelia veatchiana (see Table 1 for full nomenclature). The edges of the row cover cloth were securely buried under the soil.
In 2016–17 and 2017–18, we planted 3 rows of seeds 7.6 cm (3 in) apart under each row cover cloth (Figure 3). Planting triple rows allowed us to have more plants per unit area of row cover cloth. The area under the cloth had to be weeded by hand, whereas we were able to disk between the triple rows (which were 1.8 m [6 ft] apart) to control weeds between the planted rows.
Triple rows with uncovered (foreground) and covered (background) sections (left); Layia glandulosa growing in triple rows (center); harvesting seeds of Microsteris gracilis from triple rows into boxes as the capsules begin to mature (right).
Using row covers did give our plants a head start, but it also created problems. With a dense weed seedbank, our planted native annuals were surrounded by a dense crop of a weedy mustard, herb sophia (Descurainia sophia (L.) Webb ex Prantl [Brassicaceae]), by the time we took off the cover cloth during the seedling stage each year (usually in February). The intense weeding required once the cloth had been removed led us to ask whether the cloths were required for germination of these species at our site. So, during the 2017–18 growing year, we divided each of our planted rows into covered and uncovered sections (Figure 3). For both covered and uncovered sections, seeds were planted on the surface and tamped in by hand. Germination and growth did not differ between covered and uncovered sections for any species, so we decided to discontinue the use of row cover cloth because of the extensive growth of weeds. However, using row cover cloth may be advantageous for situations where the weed seedbank is less dense, and the cover cloth would likely be beneficial in drought years.
Germinating seeds (left) and seedlings (right) of Microsteris gracilis.
Small-Scale Seed Harvesting Methods
As a small-scale operation, harvesting was one of our challenges. We tried hand-stripping, threshing, and vacuuming, using either a leaf vacuum or a battery-powered shop vacuum. For species that have indehiscent fruits such as nutlets (Amsinckia spp.), we found that it is best to cut the inflorescence after all or most of the flowers have matured and to process cut plant material. If plant material must be stored, it is important to ensure air circulation to prevent mold. We also tried vacuuming seeds directly off the inflorescenses of Amsinckia spp. using a leaf vacuum and were able to harvest seed off the plants fairly easily; however, much of the seed was damaged by the impeller in the vacuum. For species in the Asteraceae, we have been able to use a battery-powered shop vacuum to more efficiently collect seeds.
For species that have explosive dehiscence (Microsteris gracilis) and for small-seeded species that drop their seeds (Gilia inconspicua and Collinsia parviflora), we found that it is best to collect entire plants at the point when some of the capsules have begun to dehisce but the plant retains multiple intact capsules. We let plants dry in boxes and then shake out the seeds (Figure 4, Table 2). One additional issue with Microsteris is that it has mucilaginous seed. If seed is allowed to fall onto landscape fabric and becomes wet, it will stick to the fabric when the mucilage dries out. Finally, for species that grow in dense stands and have small seeds held in capsules, such as Mentzelia veatchiana, it is efficient to use a broom to thresh seeds from mature plants, then sweep up the seeds.
Microsteris gracilis plants are harvested when capsules begin to mature, and then plants are dried in cloth-lined boxes, where seeds drop and can easily be collected.
For all methods, seed loss into the field and the formation of a seedbank of these weedy natives for subsequent growing seasons created 2 problems: decrease in seed harvest efficiency and competition from the previous years’ seed with the current crop. With indeterminate species maturing over a long period of time, multiple harvest dates for each species were necessary, and many seeds dropped to the ground before we could harvest them. Seed recovery could be improved if harvesting equipment were available. For example, using a swather/windrower could be a more efficient method for harvesting some species such as Amsinckia menziesii and Amsinckia tessellata. For other species, growing on landscape fabric might facilitate more efficient harvesting and reduce seed loss (see below).
Seed Conditioning and Cleaning
We found that the most important tools for seed cleaning were a variety of sieves and access to a large seed blower. If seed-cleaning equipment is not available, seed can be cleaned off-site, for example at the USFS Bend Seed Extractory for government entities, or by contract with seed producers or university facilities.
Two main steps are involved in the process of separating seed from harvested material: conditioning and cleaning. Seed conditioning involves breaking up plant material to free the seeds, while cleaning seed entails the separation of seed from other plant material. In the early years of this project, both steps were done primarily by hand, using kitchen sieves and other available tools. Subsequently, we were able to buy seed conditioning and cleaning equipment, which greatly increased the volume and efficiency. For most species, conditioning by hand or with a small brush machine (Westrup LA-H, Hoffman Manufacturing, Corvallis, Oregon) is followed by cleaning through sieving and the use of a continuous seed blower (Mater Continuous Seed Blower, Corvallis, Oregon). The crucial equipment for efficiently cleaning seed from these species includes a set of sieves and a seed blower. We use a combination of sieves from Seedburo Equipment Company (Des Plaines, Illinois), including perforated metal sieves ranging from 0.3 to 0.5 cm (7/64 to 12/64 in) round holes, 0.2 cm (5/64 in) triangles, and 0.2 × 2.0 cm (5/64 × 3/4 in) slots as well as fine wire mesh sieves (ranging from No. 12 to No. 30). By trial and error, we have developed species-specific recommendations for cleaning the seeds of these species (see Table 2).
Growing on Landscape Fabric
Because of problems that arose in previous years, for the 2018–19 growing season we decided to try a seed production technique suggested to us by Lynda Boyer of Heritage Seedlings and Kelsey Prickett of Benson Seeds. We hoped to find a method that might alleviate 2 of our main problems: inefficient seed harvest for species with indeterminate growth and the need for intensive hand-weeding. Following their advice, we used 0.9 m (3 ft) wide Belton weed barrier fabric (Beltech 3859; Belton Industries, Honea Path, South Carolina), which we stapled to the ground at 0.45 m (1.5 ft) intervals using eight gauge, 20.3 cm (8 in) × 2.5 cm (1 in) sod staples (Grower’s Nursery Supply, Salem, Oregon). Between each 0.9 m (3 ft) fabric, we left a 5.1 cm (2 in) gap into which we planted our seeds (Figure 5). The need for hand-weeding was dramatically reduced when planting was confined to this small area (as compared to a 30.5 cm [12 in] wide planting row when planting under row cover cloth). Because native annual forbs are such small plants, we recommend leaving only a 2.5 cm (1 in) gap for each planting row (rather than a 5.1 cm [2 in] gap, as we did), primarily to reduce seed lost into the soil of the row at harvest time. When using weed fabric with indeterminate species, seed can be allowed to fall onto the fabric and then swept or vacuumed up at weekly intervals. These techniques increase the efficiency of seed harvest and prevent the creation of a seedbank of annuals in the propagation field (Figure 5).
Stapling down weed barrier fabric prior to planting in fall 2018 (left); 3 native annual species growing in 5.08 cm (2 in) rows between 0.9 m (3 ft) weed barrier fabric. From left to right: Mentzelia veatchiana, Microsteris gracilis, and Blepharipappus scaber.
Seed harvest in spring 2019 allowed us to see the results of growing on weed barrier fabric, both in terms of weed control and harvest efficiency (Table 3). For some of our species, yield was dramatically increased by growing on weed barrier fabric: Amsinckia menziesii, Blepharipappus scaber, Layia glandulosa, and Mentzelia veatchiana all had increased yields. In fact, for Amsinckia spp., which had been notoriously difficult to collect, seed recovery increased 100 times when grown on weed barrier fabric.
Comparison of yield without versus with weed barrier fabric.
Yield was lower for Microsteris gracilis. For this species, almost no seed was lost in previous years because whole plants were harvested into boxes, and seeds were allowed to drop, but when we grew plants on weed barrier fabric, we allowed seeds to drop and then swept them up. Thus, for this species, harvesting entire plants ensured higher yield than when we allowed plants to disperse seeds onto the weed barrier fabric to be swept up. Yield for Gilia inconspicua also decreased in 2018–19, likely because plant density was lower. We are not sure if lower plant density was caused by lower seed viability or if the timing of precipitation did not meet the germination requirements of the species, but because there were several other possible reasons for lower plant density, we did not attribute this reduction in yield to the fact that plants were grown on weed barrier fabric.
Potential Applications
The strategies we recommend to others interested in creating a native annual forb seed increase program are summarized in Table 4. We found that seeds of these “native weeds” were relatively easy to grow and required no irrigation. In our field, weed control was a major issue. The use of weed barrier fabric helped minimize the area we had to hand-weed and improved our harvest efficiency. With very little equipment (other than our seed-cleaning equipment) and a modest amount of labor, we were able to create a native annual forb production program, increasing wild-collected seeds to a volume needed for larger field increases (Figure 6).
Native annuals growing in our propagation field. From left to right: Layia glandulosa, Gilia inconspicua, Mentzelia veatchiana, and Amsinckia tessellata.
Strategies for success for each step of a native annual seed program, based on our 4 y of increasing native annual forb seeds.
In the Great Basin, the Bureau of Land Management (BLM) is the major restoration practitioner buying and using native seed. Already the seeds from this increase have been used by BLM to contract a larger seed increase on one occasion. In 2016, the Ely, Nevada, BLM contracted a growout with BFI Native Seed in Washington, using G1 seed we increased for 2 species: Amsinckia menziesii and Gilia inconspicua. These seeds were grown by BFI and harvested in spring 2017. From 400 g (0.9 lb) of Amsinckia menziesii seed, they were able to harvest 23.1 kg (51 lb) on 0.04 ha (0.1 ac) for a yield of 561.5 kg/ha (501 lb/ac). From 600 g (1.3 lb) of Gilia, they harvested 3.5 kg (7.6 lb) on 0.04 ha (0.1 ac) for a yield of 85.2 kg/ha (76 lb/ac). These G2 seeds are available for post-fire seedings, and the Gilia has been used in some large-scale plots, described below.
NEXT STEPS
We plan to continue to collect wild native annual seed, particularly in good years when they are available in large quantities; these collections will be the seed for further increase efforts. We also plan to make the G1 seed available to agency personnel who can contract large-scale professional seed increases. Through these efforts, we will ensure that substantial quantities of seed of annual forbs can be available for management-scale, post-fire restoration treatments.
Professional growers doing second-level, mechanized increase will also need to document cultural practices in commercial settings, as well as develop information on potential annual yields. The market potential of these species is a question deserving of economic research.
We are also conducting field trials with these species on freshly burned sites. We have installed small research plots to test seeds from this project on the 2019 Long Valley fire on BLM lands managed by the Sierra Front field office in Carson City, Nevada. Preliminary results from spring 2020 indicate that all of the 6 native forb species we seeded (Mentzelia veatchiana, Gilia inconspicua, Microsteris gracilis, Collinsia parviflora, Layia glandulosa, and Blepharipappus scaber) emerged (Figure 7). We will continue to follow these plots to determine whether these species establish on-site, and whether they affect the density of cheatgrass and other invasive species. We also hope to determine whether seeded species remain in the seedbank in unfavorable years and are then able to emerge in subsequent years.
Three 20 × 20 cm (8 × 8 in) square plots seeded in November 2019 following the Long Valley Fire, which burned in August 2019 on the California–Nevada border north of Reno, Nevada. These photos were taken in April 2020 and show the germination of seeded native annuals, including Mentzelia veatchiana, Blepharipappus scaber, Microsteris gracilis, Layia glandulosa, and Collinsia parviflora, alongside a few cheatgrass (Bromus tectorum) plants.
Project partners Rob Burton (BLM) and Sarah Kulpa (USFWS) with Layia glandulosa (foreground) and Blepharipappus scaber (background).
Seedlings of Microsteris gracilis growing between landscape fabric.
We are using seeds from this project in a larger-scale test on the 2019 Sheep Canyon Fire on BLM lands managed by the Humboldt field office in Winnemucca, Nevada. Seeds from our native annual forb species increase were added to a native seed mix and seeded using a Truax drill. We plan to continue to facilitate the increased use of native annual forb seeds for management-scale seedings and to monitor their success. This information will help land managers include the use of native annuals as one of several strategies that will improve post-fire seeding success in the Great Basin.
ACKNOWLEDGMENTS
This work has been funded by the Nevada Department of Wildlife (Agreement 17–56), the Bureau of Land Management (Agreement L16AC00318), the U.S. Fish and Wildlife Service (Award F15AC00789), and the University of Nevada, Reno. Field and logistical help was provided by Dash Hibbard, Sarah Barga, Charlene Duncan, Jamey McClinton, Cody Ernst-Brock, Alison Agneray, Sage Ellis, Trevor Carter, Sam Bohannon, Matt Church, Stephanie Freund, Marenna Disbro, Chase Estes, Tony Kamikawa, Dan Boatman, Fred Boyles, and Scott Huber. Owen Baughman, Lynda Boyer, and two anonymous reviewers provided helpful suggestions on an earlier version of this paper.
Footnotes
Photos by the authors unless otherwise noted
