Abstract

across the landscape favorable for specific species groups. We also tested two multivariate methods for quantifying associations among species groups, and found that Mantel tests using traditional distance measures were inappropriate because of the double-zero problem of species absences, whereas canonical correlation modeled species group associations consistent with species distributions among sites. This study is among the first to develop ecological species groups in the southern United States, and the species group approach was useful for explaining vegetation-environment relationships, identifying groups of ground-flora and tree species that varied together across the landscape, and for determining the environmental gradients most strongly associated with species distributions.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call