Local Populus fremontii outperformed by transplanted genotypes: Re-assessing the home-site hypothesis in restoration

Background/Question/Methods

We examined the hypotheses that individual plant genotype and source population of Fremont cottonwood, Populus fremontii, would have a significant effect on the survival of trees in a 6,000 tree restoration site at Cibola National Wildlife Refuge, AZ. Experimental examination of genetic effects is important because standard restoration practices advocate the use of local genotypes, but tests are rarely conducted to quantify genetic effects and to determine if genotypes derived from local populations actually perform better than genotypes from more distant source populations.

Results/Conclusions

Our results show: 1) Genetic effects were large; mean mortality rate among genotypes ranged from 9.84% to 88.89%; 2) Environmental and/or maternal (i.e., planting micro-location, sapling diameter, or sapling height) effects were not significant. Genotype was a significant factor in determining mortality (p < 0.001), explaining 11.26 % of variation among genotypes; 3) Surprisingly, local genotypes performed significantly worse than genotypes from much farther away (up to 900km). Significant relationships were found between mortality and geographic distance from site of origin (r² = 0.33), elevation of site of origin (r² = 0.33), mean annual precipitation of site of origin (r² = 0.28), and mean annual temperature of site of origin (r² = 0.29). In combination, these results show that mortality rates generally decrease with increasing distance from the restoration site. They have significant implications for restoration efforts, as they do not support the ‘home-site advantage’ hypothesis (Montalvo and Ellstrand 2000). Our results suggest that climate change and other anthropogenic effects may shift local environments so much that local genotypes are no longer best adapted to local conditions. By selecting genotypes for specific locations, it may be possible to lower mortality rates, thus increasing restoration efficiency. We view our findings as a major conundrum for restoration. Although these data should be cautiously interpreted, they may be an important warning that anthropogenic-based environmental changes may so alter environments that in restoration efforts “home” may no longer select for local genotypes.