This study investigates the dynamics between propagule pressure and environmental resistance in plant invasion using Oxalis stricta L. (common yellow woodsorrel) as a model species. Oxalis stricta is a North American forb that can become invasive in native and novel habitats. A simulated invasion was conducted in a controlled glasshouse environment with propagule pressure from proximate reproducing plants and environmental resistance from planted turfgrass communities (biotic) and environmental heterogeneity in moisture and fertility (abiotic). Whereas O. stricta successfully established and proliferated in bare soil (essentially disturbed) environments with high propagule pressure, its invasion into established plant communities – and subsequent reproduction – was largely inhibited. Additionally, O. stricta reproduction was highest at intermediate levels of O. stricta cover, suggesting that negative density dependence somewhat limited reproductive output. Still, O. stricta was able to establish and persist in some trays with minimal propagule input, which may be enough for self-pollinating plants that may thrive in small populations. Such findings might challenge the propagule pressure versus environmental resistance dichotomy as environmental resistance may outweigh propagule pressure in determining invasion success, particularly in environments with well-established plant communities, but the ability to self-pollinate with high germination rates might allow even low-density invaders to establish and persist. The research underscores the importance of considering both propagule pressure and environmental resistance, as well as the reproductive strategies of the invader, in understanding invasion dynamics.

The experiment was conducted in a climate-controlled glasshouse bay (3.5 x 6.5 m) November 2020-March 2021. The glasshouse bay walls and floor were thoroughly cleaned before initiating the experiment, and all outside ventilation was kept closed. Eight mature O. stricta plants were collected from plants that naturally colonized plant containers in the glasshouse and placed (2 each) in 4, 27.8 x 54.5 x 6.2 cm trays filled with commercial potting soil. The four O. stricta source trays were placed on growing tables at one end of the glasshouse bay (Fig. 1, S1). The plants were producing ballistic dispersing seeds at the beginning of the experiment. Thirty-two experimental trays were filled with commercial planting soil and half (16) were planted with an equal mix (30 cm³ seeds per species) of the field/turf graminoids and herbaceous plants (hereafter, “plant resistance trays”). Eight experimental trays were each placed on tables at 1, 2, 3 and 4 m distance from the O. stricta source trays. Each plant resistance and bare soil tray was then randomly assigned a treatment so that each group got the same four treatments: control, supplemental watering, supplemental fertilizer and supplement watering + supplemental fertilizer (n = 8 replicates each treatment). All trays were watered thoroughly (until water drained through tray bottom) once per week; the high moisture trays were watered three times per week. Each tray was lightly fertilized (6-5-4 NPK) every 4 weeks (2 g · l^-1); the high fertility (“fertilized”) trays were fertilized weekly. All plants were grown in ambient light augmented with full spectrum LED lights (~1200-1500 max. PAR) to create 12 hours of photosynthetic light per day. Glasshouse temperatures ranged from 13-30° C. To track O. stricta establishment in the experimental trays, O. stricta cover and seedpod production were measured weekly (visually assessed). Four additional trays were filled with the same commercial soil as used in the experiment and maintained in similar conditions in a separate glasshouse bay to account for potential contamination seeds – none appeared.

To track O. stricta seed input that was ballistically dispersed from the source trays, two 25 x 30 cm white Styrofoam seed traps (S1) were placed at each row (n = 8 total). Oxalis stricta seeds are sticky, and they were easy to observe and count against the white background (hereafter, “seed input”). The seeds were rinsed off the traps with each weekly count.