Transplant Experiment and Data Collection
All four species were transplanted into each plot from our regional pool of seeds. At the corner of each plot, we hammered a 1 x 2.5 cm (diameter x length) polypipe tube into the ground with a 0.5 cm lip above the soil surface (Germain et al. 2022). We left 15 cm around each tube to prevent transplants from interacting directly, as it is a typical neighborhood size for small annual plants (Mayfield & Stouffer 2017). In each tube we planted two seeds of the same species to increase chances that at least one would germinate (7,200 seeds total). Using tubes prevents seeds from washing away and allows us to distinguish our transplanted individuals from individuals that naturally occur in plots (Germainet al. 2022); the tubes are short enough that they are unlikely to impede root growth or competition belowground. If we encountered a rock in locations designated to receive a transplant, we forgoed planting the seed and marked germination and seed production as ‘zero’ to reflect the unsuitability of this microsite.
At the end of the growing season (May 2020), we returned to quantify occupancy and persistence. First, to quantify occupancy, we photographed every block for image analysis. We recorded if our four focal species occurred naturally (binary), as well as a rough measure of abundance (i.e., four categories: zero individuals, 1-10 individuals, 11-100 individuals, and >100 individuals per plot). Second, to quantify persistence, we scored whether transplanted individuals in each tube germinated (binary), survived to reproduce (binary), and collected all seed and biomass. We also recorded instances of obvious herbivory (e.g., lopped off seed head) or if a tube went missing (total of 393 instances, mostly due to birds stealing tubes). We counted every seed produced by each transplanted individual, as well as noted whether or not one or both of the two transplanted seeds per tube survived to produce seed so we could account for it in our analyses. Given that two seeds were planted into each tube, population growth rates were calculated per capita (λ ) and were used to create a binary variable of whether each population was persistent (λ ≥1) or not (λ <1). Together, these data allow us to identify how populations experienced the suitability of the landscape along a productivity gradient.