Eight generations of native seed cultivation reduces plant fitness relative to the wild progenitor population.

Native seed for restoration is in high demand, but widespread habitat degradation will likely prevent enough seed from being sustainably harvested from wild populations to meet this need. While propagation of native species has emerged in recent decades to address this resource gap, few studies have tested whether the processes of sampling from wild populations, followed by generations of farm cultivation, reduce plant fitness tolerance to stress over time. To test this, we grew the eighth generation of farm-propagated Pursh (Onagraceae) alongside seeds from two of the three original wild source populations that established the native seed farm. To detect differences in stress tolerance, half of plants were subjected to a low-water treatment in the greenhouse. At the outset, farmed seeds were 4.1% heavier and had 4% greater germination compared to wild-collected seed. At maturity, farmed plants were 22% taller and had 20% larger stigmatic surfaces, even after accounting for differences in initial seed size. Importantly, the mortality of farmed plants was extremely high (75%), especially in the low-water treatment (80%). Moreover, farmed plants under the high-water treatment had 90% lower relative fitness than wild plants due to the 1.3 times greater weekly mortality and a 3-fold reduction in flowering likelihood. Together, these data suggest that bottlenecks during initial sampling and/or unconscious selection during propagation severely reduced genetic diversity and promoted inbreeding. This may undermine restoration success, especially under stressful conditions. These results indicate that more data must be collected on the effects of cultivation to determine whether it is a suitable source of restoration seed.