Provisional methods to guide species‐specific seed transfer in ecological restoration

Abstract

AbstractTransferring plant material during ecological restoration has inherent risk. The use of seed transfer guidelines minimizes the possibility of introducing maladapted genotypes. We delineated biogeographic regions relevant to the distribution of Cercocarpus montanus for the purpose of creating provisional seed transfer zones for ecological restoration. We also modeled seed transfer guidelines using quantitative estimates of environmental tolerance and thresholds. Analyses identified broadscale environmental patterns relevant for seed transfer success. First, a species distribution model was used to identify the distribution of C. montanus. Next, we used non‐metric multidimensional scaling to investigate the structure of environmental data, and hierarchical cluster analysis to delineate biogeographic regions (i.e., environmental discontinuities) using species distribution data. Finally, we calculated measures of environmental tolerance and thresholds for C. montanus to model the probability of seed transfer success with multiple logistic regression. Biogeographic regionalization of C. montanus resulted in four major clusters, which agreed with ordination methods. Logistic regression was implemented using estimates of environmental tolerance and threshold data to model seed transfer success. We compared our species‐specific seed transfer zones and guidelines with other provisional seed transfer zone methods and found that our species‐specific methods performed better at explaining phenotypic variation of C. montanus in four out of six cases. Seed transfer zones are useful for restoration planning; however, zonal models fail to reflect much of the environmental heterogeneity present across the range of C. montanus. Continuous models for seed transfer success using environmental tolerance and thresholds enhance the development and use of seed transfer guidelines because they reflect landscape heterogeneity at a fine scale, and the results are relative to restoration sites of interest. Herein, we describe a methodology to construct provisional seed transfer zones and continuous seed transfer guidelines using species‐specific distribution models and multivariate analyses.