Graduation date: 2008
In ecological restoration, species that are sown to increase the native plant
diversity range in establishment ability. Some species readily establish, while others
rarely do. This study set out to investigate some of the potential processes influencing
species establishment, as well as the traits that are associated with the success of
species in restoration. Twenty‐eight species native to upland prairies of the Willamette
Valley of Oregon were sown in different seed mixtures in field plots in a former
agricultural field. These species were divided into three a priori functional groups,
annual forbs, perennial forbs, and grasses, to determine whether interactions among
functional groups influenced the performance of functional groups and other measures
of restoration success, including native species richness, cover, and biomass. There
was no evidence of inter‐group competition; rather, competition was greater within
functional groups, particularly within annual forbs. Native cover and biomass increased significantly with the number of functional groups sown; however, the amount of variation explained by functional group diversity was less than 10%. Nonnative plant abundance was found to influence native performance much more than functional group richness. Sown native richness was not strongly influenced by either functional group richness or non‐native abundance.
To look for correlations between species traits and performance, eleven different traits of each species were measured from both laboratory and field‐grown plants. These were related to measures of field performance, including cover (%) and
frequency of establishment using step‐wise regression techniques. Models relating
traits to measures of performance were strong, with traits explaining up to 56% of
variation in cover, and 49% of establishment frequency. The relationship between traits
and performance varied depending on functional group, and intergroup interactions
among annual forbs also influenced cover within this functional group.
If these results were to be put into practice, a functionally diverse seed mix for
greater native abundance would be recommended for greater native cover. The regression models should be tested using different species or at a different site to determine their predictive ability. The results presented here should be useful to land
managers and from a general ecological sense as well.