Appendix S1. the Covariance Model Used for Repeated Measures Analyses of Different Response

Appendix S1. The covariance model used for repeated measures analyses of different response variables. We used the methods of Littell et al. (2002)a to select the appropriate covariance model. Data collected over several time periods were analyzed with repeated measures analysis using PROC MIXED, whereas data collected at just one time were analyzed with PROC GLM.

Response variable / Covariance model
Sown non-native forb % coverage / Compound symmetry
Sown C3 grass % coverage / Compound symmetry
Sown C4 grass % coverage / Compound symmetry
Soil moisture / First-order ante dependence
Light penetration / Compound symmetry

a Littell RC, Stroup WW, Freund RJ (2002) SAS for linear models, 4th edn. SAS Institute Inc., Cary, NC, USA.

Appendix S2. Response variables and how they were transformed and calculated. Our primary response variable was target C4 grass abundance, but we were also interested in the effects of treatments on 15N, soil moisture, light penetration, productivity, change in species richness, and change in species diversity. We examined the abundance of non-native forbs and C3 grasses to describe how well our treatments established. Additional lighting was provided in the greenhouse on cloudy days and between 5:00-8:30 am and 5:00-8:30 pm, but we did not measure light levels when the greenhouse lights were on. The null expectation for the change in richness and diversity was that no change would occur in diversity between the beginning and the end of the experiment. Simpson’s diversity was calculated as (1 ÷  pi2).

Response variable / Transformation / Calculation
Sown non-native
forb % coverage / natural log
Sown non-native
forb productivity / none
Sown C3 grass
% coverage / natural log
Sown C3 grass
productivity / none
Sown C4 grass
% coverage / natural log
Sown C4 grass
productivity / natural log
Aboveground
productivity / none
Root productivity / natural log
Leached 15N / none / (Total 15N – soil 15N – root 15N – aboveground 15N) ÷ total 15N
Soil 15N / none / Soil 15N ÷ total 15N
Root 15N / natural log / Root 15N ÷ total 15N
Aboveground 15N / none / Aboveground 15N ÷ total 15N
Total soil N / none
Total aboveground
and root N / natural log
Soil moisture / none / (weight of water in soil ÷ weight of wet soil)
Light penetration / natural log / (below canopy light level ÷ above canopy light level)
Change in richness / none / [ln (live aboveground biomass richness ÷ initial
seeding biomass richness)]
Change in diversity / none / [ln (live aboveground biomass diversity ÷ initial
seeding biomass diversity)]

Appendix S3. Mean seedling emergence per tub for target C4 prairie grasses, sown non-native forbs, and sown C3 grasses.

Target C4 prairie grasses
Bouteloua
Curtipendula / 20
Other C4 grasses / 24
Sown C3 grasses
No sown C3 grasses / 0
Bromus inermis / 148
seeded
Pascopyrum smithii / 74
transplanted
Pascopyrum smithii / 20
Sown non-native forbs / 0x density / 0.1x density / 1x density
Melilotus officinalis / 0 / 9 / 86
Verbascum thapsus / 0 / 4 / 23
Lespedeza cuneata / 0 / 2 / 14