Plant species with affinity for harsh substrates often have well-defined edaphic (soil) niches and are ideal for exploring questions of community assembly. Vertic clay soils are chemically and physically challenging to plant establishment and productivity, and annual plant communities associated with these soils of the San Joaquin Desert (California, USA) form a distinctive mosaic pattern of species that reflects differences in soil properties across the landscape. We conducted field sampling and a pot study with 12 native annual forb species with an affinity for vertic clay soils to determine how heterogeneous soils at two sites in the San Joaquin Desert differed between realized niches of species, to test if species differed in their realized and fundamental edaphic niches, and to examine the competition effects of an invasive annual grass (Bromus rubens) on these species’ edaphic niches. From our field study, we found some differences in the vertic clay soils between the realized niches of species at both sites. In our pot study, we found species exhibited similar fundamental edaphic niche optima on our treatment soils, and that species’ competitive ability differed across the gradient of edaphic stress in our treatment soils. For some species, differences in competitive ability led to shifts in edaphic niche optima, likely contributing to more divergent realized niches. We found that the combination of competitive pressure and abiotic stress drove differences between the realized niche and fundamental niche for species in a novel, heterogeneous study system.

Data in "dryad_fieldsoilsweighted_sitesseparate.csv" and "dryad_fieldsoilsweighted.csv" was collected as follows":

Field-collected soil samples from the species patches for each study species (hereafter referred to as their “home soils”) were collected from sites at Cantua Creek, Carrizo Plain, or both, based on species distribution and verifiability of species patches during field visits. For each species, we sampled approximately 0.5 L of rhizosphere soil from a maximum depth of 25 cm within the centroid of a species patch dominated by that species. Replicate sampling (n ≥ 5) for a single species was done at a minimum distance of 15 m and efforts made to ensure patches sampled for a single species were separated by one or more patches dominated by other species to avoid pseudo-replication. All soil samples were air-dried and sieved to 2 mm prior to analysis.

Soil samples collected from field sites and soils used for the pot study were analyzed by A&L Western Laboratories (Modesto, California, USA) for saturated paste pH (Rhoades and Miyamoto, 1990), KCl (2 M) extractable NO₃^- (Keeney and Nelson, 1982), Weak Bray P1 extractable HPO₄^-2 (Bray and Kurtz, 1945), and bicarbonate extractable HPO₄^-2 (Olsen and Sommers, 1982), ammonium acetate (1M, neutral pH) extractable K⁺, Ca²⁺, Mg²⁺, Na⁺ (Thomas, 1982), Cation Exchange Capacity (CEC) calculated as the sum of extracted cations, and soil texture. Phosphorus content values used in data analysis were from either Weak Bray P1 or Olson/sodium bicarbonate results, as appropriate for each soil samples’ pH (Olsen et al., 1954; Frank et al., 1998). Molar ratios of Na:K and Ca:Mg were calculated from extractable elemental analysis.

Coefficient of linear extension (COLE) was used as a measure of shrink-swell potential following the method of Schafer and Singer (1976). For each sample, 100 g of sieved soil was wetted to a saturated paste and allowed to equilibrate. The soil paste for each sample was then extruded from a 25 cm³ syringe into 1-cm diameter rods, and length measured before being allowed to dry. Once dry, rod length was measured again, and COLE was calculated as: COLE = ((saturated rod length - dry rod length)/(dry rod length)).

Data in "dryad_biomass.csv" was collected as follows:

Seeds for all study species were provided by the Bureau of Land Management Central Coast Field Office, except for Layia munzii. All seed collections made by the Bureau of Land Management for species in this study were made from a minimum of 100 parent plants in collections of a minimum of 1000 seeds total that were homogenized post-collection. Seed collections were made from parent plants growing at or near one of our field sites (collection site was determined by species distributions). Parent plants for all species were growing on vertic clay soils. For L. munzii, we collected approximately 800 seeds from approximately 50 parent plants in July 2020 at Carrizo Plain.

The pot study was conducted outdoors in Fresno, California, from December 2020 to July 2021. The city of Fresno is located at the far eastern edge of the San Joaquin Desert and has a Köppen climate classification of BSk (cold semi-arid) to Csa (hot summer Mediterranean; Beck et al., 2023). Pots were set up outdoors in full sunlight, on a 1m high platform to ensure even and continuous sun exposure throughout daylight hours.

Three soil types were selected for use in this study to capture a gradient of soil chemical stress present at field sites (Table 3). Clay soils are susceptible to pore collapse when the soil moisture content exceeds the plastic limit (the lowest moisture content at which a soil can undergo deformation without cracking), and passive watering of potted clay soils is often ineffective. For this reason, soils were manually wetted to ensure homogenous wetting throughout the soil for each pot. Each soil type was mixed with water by hand to achieve a consistent, homogeneous level of moisture within a semi-solid state without exceeding the plasticity limit. Soils were wetted manually until stable, moist 6–8 mm crumb aggregates formed. Plastic tree pots with dimensions 12.7 x 12.7 x 30.5 cm were filled to the top with the moistened soil. Soil was settled into the pots by lifting pots from the top edges and lightly tapping the bottom of the pot on a solid surface. Additional moistened soil was added to all pots so that the soil surface was within 2 cm of top edge of the pot. The pots were then placed in 29.5 x 54.6 x 6.4 cm plastic subirrigation trays which were filled with water to a depth of approximately 5 cm. The potted soils were left for 24 hours to rest and allow the soil to further moisten to field capacity by capillary action. Due to low germination rates in germination trials, the pots were heavily seeded (50–100 seeds; seedlings were thinned to one per pot following germination) for all study species in December 2020. Seeds were left uncovered by soil and misted from above daily for four weeks, in addition to subirrigation.

Each of the 12 focal species was seeded in four replicate pots for every combination of soil treatment and competition treatment for n = 24 pots per species total (Figure 2). Pots were separated following seedling thinning into four blocks of replicates (one replicate of each species × soil treatment × competition treatment per block) and pots within each block were randomized in placement within the block. To ensure there was no possibility of dissolved minerals from one soil type leaching into the subirrigation tray’s water and affecting another soil type, each block was internally randomized into sub-blocks of one soil type within a tray.

Prior to seeding in pots, all seeds were cold-stratified at 3 °C for a 3-month period. For species with low germination rates during prior germination trials (Benitoa occidentalis, Caulanthus anceps, Deinandra halliana, Layia munzii, Monolopia major, M. stricta, and Phacelia ciliata), backups were seeded in potting soil in 3.8 cm x 3.8 cm-celled seedling trays on the same date all pots were seeded. Transplants for one or more pots were necessary for B. occidentalis, C, anceps, D. halliana, L. munzii, M. major, M. stricta, and P. ciliata, and were transplanted within five weeks of seeding in all cases. Following germination, study species seedlings were thinned to one individual per pot. Pots with the competition treatment were thinned to three individual B. rubens seedlings surrounding one focal species seedling, per pot.

 Overhead watering was conducted to keep the soil surface moist through seed germination and establishment. Following seedling establishment, overhead watering was reduced and applied as needed for all pots (watered in a uniform manner to all pots when applied). All overhead watering was tapered off after three months (late February 2021), and watering was limited to subirrigation for the remainder of the study. Beginning in May 2021, subirrigation watering was tapered down to reflect field conditions by gradually reducing the frequency of refilling the subirrigation trays with water. Beginning in June 2021, the volume of water used to refill subirrigation trays was also reduced. By July 2021, all plants had senesced, shortly before subirrigation tray refills were scheduled to be completely halted.

Plant biomass was harvested following senescence of each plant. For the purposes of this study, senescence was defined as having occurred when either the last flower had bloomed and withered, with no developing buds remaining, or over 50% of leaves on the plant had browned over at least one third of their surface area, regardless of fruit development on the last flowers. To reduce loss of reproductive parts that might alter final biomass, flowers were bagged with cinchable polyester organza exclusion bags enclosing the flower and top of the pedicel after petals began to wither or after the innermost disk flowers had been open for a minimum of 24 hours. For Asteraceae, bags enclosed the capitulum and top of the peduncle. If a plant had not yet met our definition of senescence but had mature fruits, the fruits and flower parts captured in the bag were removed and dried early to prevent loss of that biomass.

Following senescence, plant stems were cut at the soil surface and shoot biomass was collected. Roots were cleaned of soil immediately following shoot material collection. Root biomass was cleaned of soil residue using a method modified from Rechel (1993). For each plant, the entire soil mass from the pot was fully submerged and soaked in a dilute (≤ 0.25%) HCl solution for 15–45 minutes to loosen the soil by dissolution of carbonates, in combination with sonication at 70 KHz. Sonication was run for 15-minute intervals, as needed, to loosen the clay soil particles from the roots. Initial soaking and sonication were followed by further sonication and rinsing with water, as needed, in combination with manual agitation of the root mass over a 2 mm sieve. Controlled agitation and rinsing with water over the sieve were repeated as necessary to prevent roots from tearing or any substantial loss of root mass, until the roots were entirely free of soil particles. Roots of competition-treated plants were painstakingly and carefully separated from B. rubens roots by hand. All collected and cleaned plant materials (reproductive parts, shoots, and roots) were dried to constant mass at 41°C for 24–48 hours and the combined total biomass of all three was recorded.

Citations:

Beck, H. E., T. R. McVicar, N. Vergopolan, A. Berg, N. J. Lutsko, A. Dufour, A. Zeng, et al. 2023). High-resolution (1 km) Köppen-Geiger maps for 1901–2099 based on constrained CMIP6 projections. Scientific Data 10: 724.

Bray, R. H., and L. T. Kurtz. 1945. Determination of total, organic, and available forms of phosphorus in soils. Soil Science 59: 39–45.

Frank, K., D. Beegle, and J. Denning, J. 1998. Recommended phosphorus tests. In J. R. Brown [ed.], Recommended chemical soil test procedures for the North Central Region. Missouri Agricultural Experiment Station, Columbia, Missouri, USA.

Keeney, D. R., and D. W. Nelson. 1982. Nitrogen: inorganic forms. In A. L. Page, R. H. Miller, and D. R. Keeney [eds.], Methods of soil analysis. Part 2. Chemical and microbiological properties, 643–689. American Society of Agronomy, Madison, Wisconsin, USA.

Olsen, S. R., C. V. Cole, F. S. Watanabe, and L. A. Dean. 1954. Estimation of available phosphorus in soils by extraction with sodium bicarbonate. United States Department of Agriculture, Washington, District of Columbia, USA.

Olsen, S. R., and L. E. Sommers. 1982. Phosphorus. In A. Klute [ed.], Methods of soil analysis, part 2: Chemical and microbiological properties, 403–430. American Society of Agronomy, Madison, Wisconsin, USA.

Rechel, E. A. 1993. Using acetic acid to wash roots from calcareous soil. Communications in Soil Science and Plant Analysis 24: 1845–1848.

Rhoades, J. D., and S. Miyamoto. 1990. Testing soils for salinity and sodicity. In R. L. Westerman [ed.], Soil testing and plant analysis, 3^rd ed., 299–336. Soil Science Society of America, Madison, Wisconsin, USA.