Malheur Experiment Station Information for Sustainable Agriculture

Native Wildflower Seed Production With low levels of Irrigation

Clint Shock, Erik Feibert, and Lamont Saunders, Malheur Experiment Station, Oregon State University, Ontario, OR

Nancy Shaw, U.S. Forest Service, Rocky Mountain Research Station, Boise, ID

Ram S. Sampangi, University of Idaho, Parma, ID

Introduction

Native wildflower seed is needed to restore rangelands of the Intermountain West. Commercial seed production is necessary to provide the quantity of seed needed for restoration efforts. A major limitation to economically viable commercial production of native wildflower (forb) seed is stable and consistent seed productivity over years.

In natural rangelands, the natural variations in spring rainfall and soil moisture result in highly unpredictable water stress at flowering, seed set, and seed development, which for other seed crops is known to compromise seed yield and quality.

Native wildflower plants are not adapted to croplands. Native plants are often not competitive with crop weeds in cultivated fields. Poor competition with weeds could also limit wildflower seed production. Both sprinkler and furrow irrigation could provide supplemental water for seed production, but these irrigation systems risk further encouraging weeds. Also, sprinkler and furrow irrigation can lead to the loss of plant stand and seed production due to fungal pathogens. By burying drip tapes at 12-inch depth and avoiding wetting of the soil surface, we hoped to assure flowering and seed set without undue encouragement of weeds or opportunistic diseases. The trials reported here tested the effects of three low rates of irrigation on the seed yield of 13 native forb species.

Materials and Methods

Plant Establishment

Seed of the seven Intermountain West forb species (the first seven species in Table 1) was received in late November in 2004 from the Rocky Mountain Research Station (Boise, ID). The plan was to plant the seed in the fall of 2004, but due to excessive rainfall in October, the ground preparation was not completed and planting was postponed to early 2005. To try to ensure germination the seed was submitted to cold stratification. The seed was soaked overnight in distilled water on January 26, 2005, after which the water was drained and the seed soaked for 20 min in a 10 percent by volume solution of 13 percent bleach in distilled water. The water was drained and the seed was placed in thin layers in plastic containers. The plastic containers had lids with holes drilled in them to allow air movement. These containers were placed in a cooler set at approximately 34°F. Every few days the seed was mixed and, if necessary, distilled water added to maintain seed moisture. In late February, seed of Lomatium grayi and L. triternatum had started to sprout.

In late February, 2005 drip tape (T-Tape TSX 515-16-340) was buried at 12-inch depth between 2 30-inch rows of a Nyssa silt loam with a pH of 8.3 and 1.1 percent organic matter. The drip tape was buried in alternating inter-row spaces (5 ft apart). The flow rate for the drip tape was 0.34 gal/min/100 ft at 8 psi with emitters spaced 16 inches apart, resulting in a water application rate of 0.066 inch/hour.

On March 3, seed of all species was planted in 30-inch rows using a custom-made plot grain drill with disk openers. All seed was planted at 20-30 seeds/ft of row. The Eriogonum umbellatum and the Penstemon spp. were planted at 0.25-inch depth and the Lomatium spp. at 0.5-inch depth. The trial was irrigated with a minisprinkler system (R10 Turbo Rotator, Nelson Irrigation Corp., Walla Walla, WA) for even stand establishment from March 4 to April 29. Risers were spaced 25 ft apart along the flexible polyethylene hose laterals that were spaced 30 ft apart and the water application rate was 0.10 inch/hour. A total of 1.72 inches of water was applied with the minisprinkler system. Eriogonum umbellatum, Lomatium triternatum, and L. grayi started emerging on March 29. All other species except L. dissectum emerged by late April. Starting June 24, the field was irrigated with the drip system. A total of 3.73 inches of water was applied with the drip system from June 24 to July 7. The field was not irrigated further in 2005.

Plant stands for Eriogonum umbellatum, Penstemon spp., Lomatium triternatum, and L. grayi were uneven. Lomatium dissectum did not emerge. None of the species flowered in 2005. In early October, 2005 more seed was received from the Rocky Mountain Research Station for replanting. The blank lengths of row were replanted by hand in the E. umbellatum and Penstemon spp. plots. The Lomatium spp. plots had the entire row lengths replanted using the planter. The seed was replanted on October 26, 2005. In the spring of 2006, the plant stands of the replanted species were excellent, except for P. deustus.

On April 11, 2006 seed of three globemallow species (Sphaeralcea parvifolia, S. grossulariifolia, S. coccinea), two prairie clover species (Dalea searlsiae, D. ornata), and basalt milkvetch (Astragalus filipes) was planted at 30 seeds/ft of row (Table 1). The field was sprinkler irrigated until emergence. Emergence was poor. In late August of 2006 seed of the three globemallow species was harvested by hand. On November 9, 2006 the six forbs that were planted in 2006 were mechanically flailed. On November 10, 2006 the six forbs were replanted. On November 11, the Penstemon deustus plots were also replanted at 30 seeds/ft of row.

Table 1. Forb species planted in the drip irrigation trials at the Malheur Experiment Station, Oregon State University, Ontario, OR.

Irrigation for Seed Production

In April, 2006 each strip of each forb species was divided into plots 30 ft long. Each plot contained four rows of each species. The experimental designs were randomized complete blocks with four replicates. The three irrigation treatments were a non-irrigated check, 1 inch per irrigation, and 2 inches per irrigation. Each treatment received 4 irrigations that were applied approximately every 2 weeks starting with flowering of the forbs. The amount of water applied to each treatment was calculated by the length of time necessary to deliver 1 or 2 inches through the drip system; the amount was measured by a water meter and recorded after each irrigation to ensure correct water applications. Irrigations were controlled with a controller and solenoid valves.

In March of 2007, the drip-irrigation system was modified to allow separate irrigation of the species due to different timings of flowering. The three Lomatium spp. were irrigated together and Penstemon deustus and P. speciosus were irrigated together, but separately from the others. Penstemon acuminatus and Eriogonum umbellatum were irrigated individually. In early April, 2007 the three globemallow species, two prairie clover species, and basalt milkvetch were divided into plots with a drip-irrigation system to allow the same irrigation treatments that were received by the other forbs.

Irrigation dates can be found in Table 2. In 2007, irrigation treatments were inadvertently continued after the fourth irrigation. In 2007, irrigation treatments for all species were continued until the last irrigation on June 24.

Soil volumetric water content was measured by neutron probe. The neutron probe was calibrated by taking soil samples and probe readings at 8-, 20-, and 32-inch depths during installation of the access tubes. The soil water content was determined volumetrically from the soil samples and regressed against the neutron probe readings, separately for each soil depth. Regression equations were then used to transform the neutron probe readings during the season into volumetric soil water content.

Flowering, Harvesting, and Seed Cleaning

Flowering dates for each species were recorded (Table 2). The Eriogonum umbellatum and Penstemon spp. plots produced seed in 2006, in part because they had emerged in the spring of 2005. Each year, the middle two rows of each plot were harvested when seed of each species was mature (Table 2), using the methods listed in Table 3. The plant stand for P. deustus was too poor to result in reliable seed yield estimates. Replanting of P. deustus in the fall of 2006 did not result in adequate plant stand in the spring of 2007.

Eriogonum umbellatum seeds did not separate from the flowering structures in the combine; the unthreshed seed was taken to the U.S. Forest Service Lucky Peak Nursery (Boise, ID) and run through a dewinger to separate seed. The seed was further cleaned in a small clipper seed cleaner.

Penstemon deustus seed pods were too hard to be opened in the combine; the unthreshed seed was precleaned in a small clipper seed cleaner and then seed pods were broken manually by rubbing the pods on a ribbed rubber mat. The seed was then cleaned again in the small clipper seed cleaner.

Penstemon acuminatus and P. speciosus were threshed in the combine and the seed was further cleaned using a small clipper seed cleaner.

Cultural Practices in 2006

On October 27, 2006, 50 lb phosphorus (P)/acre and 2 lb zinc (Zn)/acre were injected through the drip tape to all plots of Eriogonum umbellatum, Penstemon spp., and Lomatium spp. On November 11, 100 lb nitrogen (N)/acre as urea was broadcast to all Lomatium spp. plots. On November 17, all plots of Eriogonum umbellatum, Penstemon spp. (except P. deustus), and Lomatium spp. had Prowl^® at 1 lb ai/acre broadcast on the soil surface. Irrigations for all species were initiated on May 19 and terminated on June 30. Harvesting and seed cleaning methods for each species are listed in Table 3.

Cultural Practices in 2007

Penstemon acuminatus and P. speciosus were sprayed with Aza-Direct^® at 0.0062 lb ai/acre on May 14 and 29 for lygus bug control. Irrigations for each species were initiated and terminated on different dates (Table 2). Harvesting and seed cleaning methods for each species are listed in Table 3. All plots of the three Sphaeralcea species were flailed on November 8, 2007.

Cultural Practices in 2008

On November 9, 2007 and on April 15, 2008, Prowl at 1 lb ai/acre was broadcast on all plots for weed control. Capture^® 2EC at 0.1 lb ai/acre was sprayed on all plots of Penstemon acuminatus and P. speciosus on May 20 for lygus bug control. Irrigations for each species were initiated and terminated on different dates (Table 2). Harvesting and seed cleaning methods for each species are listed in Table 3.

Cultural Practices in 2009

On March18, Prowl at 1 lb ai/acre and Volunteer^® at 8 oz/acre were broadcast on all plots for weed control. On April 9, 50 lb N/acre and 10 lb P/acre were applied through the drip irrigation system to the three Lomatium species.

The flowering, irrigation timing, and harvest timing were recorded for each species (Table 2). Harvesting and seed cleaning methods for each species are listed in Table 3.

Cultural Practices in 2010

On December 4, 2009, Prowl at 1 lb ai/acre was broadcast on all plots for weed control. Flowering, irrigation timing, and harvest timing were recorded for each species (Table 2). Harvesting and seed cleaning methods for each species are listed in Table 3.

Table 2. Native forb flowering, irrigation, and seed harvest dates by species in 2006, 2007, and 2008, Malheur Experiment Station, Oregon State University, Ontario, OR.

Table 3. Native forb seed harvest and cleaning by species, Malheur Experiment Station, Oregon State University, Ontario, OR.

^a Wintersteiger Nurserymaster small-plot combine with dry bean concave.

^b Specialized seed threshing machine at USDA Lucky Peak Nursery used in 2006. Thereafter an adjustable hand-driven corn grinder was used to thresh seed.

^c Clipper seed cleaner.

^d Wintersteiger Nurserymaster small-plot combine with alfalfa seed concave. For the Sphaeralcea spp., flailing in the fall of 2007 resulted in more compact growth and one combine harvest in 2008, 2009, and 2010.

^e Hard seed pods were broken by rubbing against a ribbed rubber mat.

^f Harvested by hand in 2007 and 2009 due to poor seed set.

Results and Discussion

The soil volumetric water content in the various species in 2010 responded to the irrigation treatments on each specie (Figs. 3-7) and remained fairly moist due to the late distribution of precipitation in 2010 (Fig. 2).

Emergence for the two prairie clover (Dalea spp.) species in the spring of 2007 was again poor. Emergence for Penstemon deustus and for basalt milkvetch (Astragalus filipes) was extremely poor; A. filipes produced negligible amounts of seed in 2007.

Flowering and Seed Set

Penstemon acuminatus and P. speciosus had poor seed set in 2007, partly due to a heavy lygus bug infestation that was not adequately controlled by the applied insecticides. In the Treasure Valley, the first hatch of lygus bugs occurs when 250 degree-days (52°F base) are accumulated. Data collected by an AgriMet weather station adjacent to the field indicated that the first lygus bug hatch occurred on May 14, 2006; May 1, 2007; May 18, 2008; May 19, 2009; and May 29, 2010. The average (1995-2010) lygus bug hatch date was May 18. Penstemon acuminatus and P. speciosus start flowering in early May. The earlier lygus bug hatch in 2007 probably resulted in harmful levels of lygus bugs present during a larger part of the Penstemon spp. flowering period than normal. Poor seed set for P. acuminatus and P. speciosus in 2007 also was related to poor vegetative growth compared to 2006 and 2008. In 2009, all plots of P. acuminatus and P. speciosus again showed poor vegetative growth and seed set. Root rot affected all plots of P. acuminatus in 2009, killing all plants in two of the four plots of the wettest treatment (2 inches per irrigation).

The three Sphaeralcea species (globemallow) showed a long flowering period (early May through September) in 2007. Multiple manual harvests were necessary because the seed falls out of the capsules once they are mature. The flailing of the three Sphaeralcea species starting in the fall of 2007 was done annually to induce a more concentrated flowering, allowing only one mechanical harvest. Precipitation in June of 2009 (2.27 inches) and 2010 (1.95 inches) was substantially higher than average (0.76 inches). Rust (Puccinia sherardiana) infected all three Sphaeralcea species in June of 2009 and 2010, causing substantial leaf loss and reduced vegetative growth.

Seed Yields

Eriogonum umbellatum

In 2006, seed yield of Eriogonum umbellatum increased with increasing water application, up to 8 inches, the highest amount tested (Table 5, Fig. 8). In 2007-2009 seed yield showed a quadratic response to irrigation rate (Tables 5 and 6). Seed yields were maximized by 8.1 inches, 7.2 inches, and 6.9 inches of water applied in 2007, 2008, and 2009, respectively. In 2010, there was no significant difference in yield between treatments. The 2010 season had unusually cool (Table 4, Fig. 1) and wet weather (Fig. 2). The accumulated precipitation in April through June of 2010 was the highest over the years of the trial (Table 4). The relatively high seed yield (252 lb/acre) of E. umbellatum in the nonirrigated treatment in 2010 seemed to be related to the high April-June precipitation in 2010. Averaged over 5 years, seed yield of E. umbellatum increased with increasing water applied up to 8 inches, the highest amount tested (Fig. 3). The quadratic seed yield responses most years suggests that additional irrigation above 8 inches would not be beneficial.

Penstemon acuminatus

There was no significant difference in seed yield between irrigation treatments for Penstemon acuminatus in 2006 (Table 5). Precipitation from March through June was 6.4 inches in 2006. The 64-year-average precipitation from March through June is 3.6 inches. The wet weather in 2006 could have attenuated the effects of the irrigation treatments. In 2007 and 2008, seed yield showed a quadratic response to irrigation rate (Fig. 9). Seed yields were maximized by 4.0 and 8.5 inches of water applied in 2007 and 2008, respectively. In 2009, there was no significant difference in seed yield between treatments (Table 6). However, due to root rot affecting all plots in 2009, the seed yield results were compromised. By 2010, substantial lengths of row contained only dead plants. Measurements in each plot showed that plant death increased with increasing irrigation rate. The percent of stand loss was 51.3, 63.9, and 88.5 for the 0-, 4-, and 8-inch irrigation treatments, respectively. The trial area was disked out in 2010. Following the 2005 planting, seed yields were substantial in 2006 and moderate in 2008. P. acuminatus is a short-lived perennial.

Penstemon speciosus

In 2006-2009 seed yield of Penstemon speciosus showed a quadratic response to irrigation rate (Tables 5 and 6). Seed yields were maximized by 4.3, 4.2, 5.0, and 4.3 inches of water applied in 2006, 2007, 2008, and 2009, respectively. In 2010, there was no difference in seed yield between treatments. Seed yield was low in 2007 due to lygus bug damage, as discussed previously. Seed yield in 2009 was low due to stand loss from root rot. The plant stand recovered somewhat in 2010, due in part to natural reseeding, especially in the nonirrigated plots. The average of the seed yield for 2006 and 2008 was estimated to be maximized by 4.7 inches of water applied (Fig. 10).

Penstemon deustus

There was no significant difference in seed yield between irrigation treatments for P. deustus in 2006 or 2007. Both the replanting of the low stand areas in October 2005 and the replanting of the whole area in October 2006 resulted in very poor emergence and plots with very low and uneven stands. The production of P. deustus seed will depend on learning how to establish plant stands.

Lomatium triternatum

Lomatium triternatum showed a trend for increasing seed yield with increasing irrigation rate in 2007 (Table 5). The highest irrigation rate resulted in significantly higher seed yield than the nonirrigated check treatments. Seed yields of L. triternatum were substantially higher in 2008-2010 (Tables 5 and 6). In 2008, 2009, and 2010, seed yields of L. triternatum showed a quadratic response to irrigation rate (Fig. 11). Seed yields were estimated to be maximized by 8.4, 5.4, and 7.8 inches of water applied in 2008, 2009, and 2010, respectively. Averaged over 4 years, seed yield of L. triternatum was estimated to be maximized by 7.2 inches of applied water.

Lomatium grayi

Lomatium grayi showed a trend for increasing seed yield with increasing irrigation rate in 2007 (Table 5). The highest irrigation rate resulted in significantly higher seed yield than the nonirrigated check. Seed yields of L. grayi were substantially higher in 2008 and 2009. In 2008, seed yields of L. grayi showed a quadratic response to irrigation rate (Fig. 12). Seed yields were estimated to be maximized by 6.9 inches of water applied in 2008. In 2009, seed yield showed a linear response to irrigation rate. Seed yield with the 4-inch irrigation rate was significantly higher than in the nonirrigated check, but the 8-inch irrigation rate did not result in a significant increase above the 4-inch rate. In 2010, seed yield was not responsive to irrigation. The unusually wet spring of 2010 could have caused the lack of response to irrigation. A further complicating factor in 2010 that compromised seed yields was rodent damage. Extensive rodent (vole) damage occurred over the 2009-2010 winter. The affected areas were transplanted with 3-year old L. grayi plants from an adjacent area in the spring of 2010. To reduce their attractiveness to voles, the plants were mowed after becoming dormant in early fall of 2010. Averaged over 4 years, seed yield of L. grayi was estimated to be maximized by 5.6 inches of applied water.