Applying ecological treatments to boost yields among restoration target species of seed production areas

Introduction

To perform ecological restorations, many practitioners use revegetation to assist ecosystem recovery, particularly using seeds due their efficiency to revegetate areas (Brudvig 2011). Seed production areas (SPAs) are critical to supply these seeds for ecological restorations (Broadhurst et al. 2016). SPAs are plantings of wild species which are grown using agricultural and horticultural practices to maximize seed production (Broadhurst et al. 2016). They are considered a viable method to yield seed for restorations when wild seed sources are unavailable.

Despite the role of SPAs in ecological restoration, stakeholders in the native plant industry face severe business challenges. SPA managers have described rigorous labor requirements, poor managerial knowledge, and unreliable seed outputs of their native crops as major obstacles to efficient production (Jones and Young 2005). Likewise, institutional support for SPA managers is low and research regarding SPA management is scarce (Merrit and Dixon 2011). These obstacles can reverberate to other stakeholders. For example, prairie restorationists frequently claim that they are limited by inadequate seed availability and expensive seed prices (Rowe 2010). Restorations are also becoming larger in size and provoking seed shortages in restorations (Broadhurst et al. 2016). Hence, there is a substantial need to improve SPA management and yields (Merrit and Dixon 2011; Broadhurst et al. 2016).

Inoculation with indigenous arbuscular mycorrhizal fungi (AMF) is a promising management strategy to improve SPA yields. AMF are a broad group of soil fungi which penetrate plant roots and generally provide nutrients and water in exchange for an energy source (Berruti et al. 2016). The AMF relationship can benefit a plant by enhancing the plant’s resource acquisition, buffering it from stresses (e.g., drought and herbivory), and even defending against soil pathogens (Berruti et al. 2016). Recent evidence suggests that many restoration target species have substantial growth benefits when they associate with AMF inocula that are indigenous to prairies in experimental settings (Koziol and Bever 2017; Bauer et al. 2017). AMF inoculum was found to improve survival of planted plugs and promote early flowering of prairie restoration target species (Koziol and Bever 2017). Although the positive potential of AMF inoculation to SPAs has been identified, its value in commercial settings has yet to be evaluated.

Alternatively, SPA managers could apply chemical fertilizers to boost SPA yields, which is often done in the industry (Dunne and Dunne 2003). However, fertilizers may be a poor management choice in SPAs. Fertilizer may not be conducive to the ecology of many native target crops. Many SPA crops are slow growing and have growth traits less suited for the quick acquisition of labile soil nutrients (Bauer et al. 2017), in contrast to traditional agricultural species.

For SARE GNC19-293, I created experimental SPAs which tested the effects of AMF inoculation and fertilizer addition as management strategies. This report presents final results of the project, including descriptions of the project’s methods, outreach, and future.

Methods

Field site descriptions

Three sites were used for the project, representing different stakeholders for SPAs (Fig. 1): Barefoot Nursery & Restoration (Springfield, IL, and hereafter, “Barefoot”), Heritage (Champaign, IL, and hereafter, “Heritage”), and Lake of the Woods Forest Preserve (Mahomet, IL, and hereafter, “LOTW”). These locations represent a private enterprise, a local organization, and a regional conservation organization, respectively.

Figure 1: Locations of the three field sites of experimental SPAs in Illinois.

Barefoot is a private nursery which specializes in native plants, including existing plots for seed production. The site consists of small-scale infrastructure for growing and producing seed mixes. The SPA at this site was established on a 1-acre section of a corn-soy farm field, which was adjacent to the nursery. Soils at the site range from sandy loam to pure sand due to historical deposition from the nearby Sangamon River. Thus, some sections in the SPA appear unusually nutrient-poor, as well as prone to heating and desiccation. The Heritage location is former turf grass managed by the Champaign Park District. The site is marginal property for the Park District, with low pedestrian and recreational traffic. The site is flanked by a prairie restoration to the east, and Copper Slough to the west. The soils are Drummer silty clay loam and have a low weed seed bank. Growing conditions at Heritage seem the least stressful (e.g., high quality soil, high water retention). LOTW is a nature preserve managed by the Champaign County Forest Preserve District (CCFPD), consisting of a matrix of woodland, prairie, and savanna habitats. The SPA was established on an acre of old field adjacent to the popular Buffalo Trace Bike Trail. Previously, the site was dominated by Bromus inermis (smooth brome), Trifolium pratense (red clover), and ruderal species. Soils at the SPA are Dana silt loam and appear to be of intermediate growing quality when compared to Barefoot and Heritage.

SPA preparation and experimental design

SPAs were established in early April 2020 by using 4’ Dewitt Sunbelt weed barrier fabric to smother existing vegetation (Fig. 2).

Figure 2: The dimensions and experimental design of the SPAs at the three sites. A) A visual representation of the dimensions of each block and its components. B) An example of the experimental design, showing randomized treatments within two example blocks. Each block has randomized positioning of the study species within each row.

Experimental SPAs were created using a randomized block design, with a 2x2 factorial treatment combination, with AMF inoculation and fertilizer addition as treatments (Fig. 2). Each block was 32 ft x 43 ft and consisted of 4 25-ft rows. To prevent treatment spillover, buffers between each row were 6 ft; buffers between rows within blocks were 7 ft (Fig. 2A). Buffers were seeded with turfgrass at Barefoot and LOTW due to the lack of suitable vegetation at these sites.

Study species descriptions, cultivation, and field transfer

I chose four study species: Asclepias tuberosa (butterfly weed), Echinacea pallida (pale purple coneflower), Eryngium yuccifolium (rattlesnake master), and Parthenium integrifolium (wild quinine). These species were selected because 1) they are herbaceous flowering target species in restorations and are valuable for use in USDA programs such as the Conservation Reserve Program and CP-42 pollinator plantings (National Conservation Practice Standards 2015), 2) they were also previously shown to have positive growth responses to indigenous AMF treatments in experimental settings (Koziol et al. 2017).

I purchased seed of the four study species in January 2020, from Prairie Moon Nursery (Winona, MN). Prairie Moon Nursery sells sourced seed (i.e., local ecotype) from a network of producers throughout the Midwest. All seed was cold moist stratified in double-sterilized sand for 60 days.

I applied the nurse plant method for inoculation. This method pre-infects seedlings or plugs prior to field establishment, which can boost inoculum survival and have infected plants act as nuclei to nearby plants (Koziol et al. 2017). In late March 2020, seeds were sown into flats containing a mixture of 50-50 sand and general-purpose soil mix (1:1:1 soil/peat/perlite), which was double sterilized using steaming. There were one to two rest days in between each sterilization event. After germination and the appearance of cotyledons, seedlings were transferred to 72-cell growing trays. Thereafter, for all the study species, half of seedlings were put into -AMF growing trays (60-40 double-sterilized sand/potting soil mixture), while the other half were transferred to +AMF growing trays. +AMF growing trays consisted of 45-45 double sterilized sand/potting soil mixture, and 10% by volume of MycoBloom AMF. 10% of AMF mixture by volume is considered a liberal volume by MycoBloom instructions. I used the 50-50 sand/potting soil mixture as the base in growing trays in order to increase nutrient stress to promote +AMF seedling inoculation. Because AMF infection occurs over 2 weeks, seedlings were grown for 4-6 weeks before field transfer to promote root infection. Seedlings were grown in temperature-controlled conditions (24°C) at the University of Illinois Agricultural and Consumer Sciences Plant Care Facility (40° 6' 7" N, 88° 13' 26" W).

Most plugs were planted in-field between late April through June in Heritage and LOTW. Plugs were planted in late June and early July at Barefoot to allow the grass seed in the buffers to stabilize the soil. At each site, a particular species was always planted on a single day to minimize priority effects. Dead seedlings were replaced within 2 weeks of the initial establishment to have more complete rows, while avoiding substantial time elapsed to lead to priority effects. In each row, 6 plugs of each species were planted per row, totaling 24 plants per row across the four study species. Plugs were planted by staggering at a 1.4-ft distance (Fig. 2A); plants were staggered to avoid overcutting the center of the fabric. The fabric was cut by a 15-cm “X” in the plastic to transfer the plugs.   

Field management

Throughout the growing season from 2020-2022, buffers within the blocks were mown biweekly. Rows were weeded as needed, generally monthly for Heritage and Barefoot, and once every three weeks for LOTW. During each growing season from 2020-2022, I conducted monthly fertilization treatments across all +nutrient rows in the three sites. The fertilizer used was Peters Excel 15-5-15 (NPK) fertilizer; a high N:P ratio was sought out because P fertilization is less important with respect to seed production (Ken Fromm, personal communication). Fertilization treatments consisted of adding 350 mL at 125 ppm (N) concentration to each individual plant within the +fertilizer.

Data collection and analysis

At each site, I determined each individual plant’s survival, flowering status, and relative size (Fig. 3). Size was assessed by measuring a plant’s height and its estimated leaf area. For each plant, leaves were subjectively counted as “small,” “medium,” or “large” based on the species. For each species, I estimated the area of each size class by averaging twenty samples of each species-leaf size combination. Leaf size class counts were then multiplied by the calculated average, and then summed for each plant. Apart from A. tuberosa, the seed was hand-harvested by row. A. tuberosa seed was not collected during the experiment because, across all three years and the three sites, there were nearly no viable seed pods through the duration of the experiment. For plant size estimators and seed yield, I averaged the values across species and within rows, yielding mean per capita leaf area and seed mass produced for each row, per block.

Figure 3: Data collection at Heritage during fall, 2020.

I used linear mixed effects models to assess treatment effects on mean leaf area and seed production. Seed production was log+1 transformed, and leaf area (cm²) was log-transformed. Fixed effects were nutrient and AMF treatment, sampling year, and the interaction of nutrient and AMF treatments. Random effects were site and block nested within site. All data analyses were conducted using R v. 4.0.2 (R Core Team 2020). Linear mixed effects models were constructed using the packages lme4 (Bates et al. 2015). All combinations of fixed effects were tested using the MuMIn package using the dredge() function from MuMIn (Bartoń 2022). I compared models that were fit with maximum likelihood using AICc. I used this procedure for each species individually for per capita seed yield (apart from A. tuberosa) and estimated plant size. For seed yields, I also excluded rows that did not produce seed (i.e., log+1 per capita mass = 0) because those data led to unsuitable model fit.