COS 159-1 - Climate-adjusted provenancing in the Great Basin: seedling performances under drought differ by seed source

Abstract: Rapid climate change poses a fundamental challenge to traditional approaches to seed sourcing in restoration. For example, using the “local” seed source that is locally adapted to the restoration site is the gold standard for restoration, but these seeds may not survive or persist in variable climate conditions in the future. This is especially of a high concern in arid rangelands of the West where the drought frequency and severity are forecasted to increase. An emerging idea called climate-adjusted provenancing proposes that the use of plant material (seed sources) pre-adapted to future climate change conditions could brace restored communities for future average climate conditions and extremes, such as drought. We tested this strategy in the Great Basin rangelands where increasingly vast areas are seeded for restoration after wildfires. We hypothesized that seed sources from warmer and drier average climate zones may increase seedling success in drought conditions. Here, we conducted a common garden experiment that crossed rainfall and seeding treatments to assess how seed provenance affects seedling performance under three different drought scenarios (ambient, moderate drought, severe drought). We manipulated rainfall with shelters that shielded 50% and 80% of the rain year-round, and sowed seeds sourced from six locations along an aridity gradient. We used a native perennial grass, Bottlebrush squirreltail (Elymus elymoides Raf.) as a focal species. We found that apparent seedling growth at the end of the second growing season was similar across seed sources but the strategies to cope with drought differed. Specifically, seeds from cooler/wetter sites had higher emergence rates (0.15 ± 0.023) but also higher mortality rates (0.13 ± 0.016) in drought conditions compared to seeds from warmer/drier sites. Conversely, seeds from warmer/drier sites exhibited higher water use efficiencies (∆¹³C; -28.5 ± 0.2) and higher survival rates (0.005 ± 0.002) in moderate drought than seeds from cooler/wetter sites. The trade-off in emergence and survival suggests that no one seed source is a silver bullet for seeding success, and mixed-provenance seed mixes are recommended. We provide an empirical proof of concept to the climate-adjusted provenancing strategy that could potentially improve restoration outcomes in drought-prone rangelands.