Patch expansion and community characteristics associated with Bromus inermis stands in mixed-grass prairie

Background/Question/Methods

Invasion of mixed-grass prairie by perennial cool-season grasses is a conservation concern in natural areas. At Theodore Roosevelt National Park (North Dakota), smooth brome (Bromus inermis) patches in undisturbed prairie are often elliptical in shape, possibly representing vegetative expansion of a single plant. In 2000, I paired nine purposively selected brome patches (‘brome clones’) with nearby, similarly-shaped plots in native C3 grass-dominated prairie (‘native plots’). I measured clone size and recorded percent cover of vascular plants in 0.25-m² quadrats at the centers of brome clones and native plots. I repeated the survey on the same plots in 2011. In 2011 I also recorded percent cover of plant species in nine, 0.0625-m² quadrats spaced from center to edge of the clone or plot (‘spaced quadrats’).  If brome clones originated as germinating seeds at their centers and grew radially, then these quadrats represent a space-for-time substitution in the community trajectory following brome invasion. I used ADONIS (a nonparametric MANOVA analogue) to test for differences and canonical discriminant analysis to characterize differences in plant communities between brome clones and native plots. 

Results/Conclusions

Brome clones were 1.1 to 3.9 (median = 1.7) times larger in 2011 than in 2000. Split-plot, repeated measures ANOVA indicated that most brome effects were consistent between 2000 and 2011. Plant diversity was lower in clones than in native plots. Native plant diversity, native forb cover, and cover of introduced annual grasses did not differ between brome clones and native plots, but introduced forb cover was lower in brome clones compared to native plots. There was more bare ground, and in 2011 only (time x brome interaction) live plant cover was greater in brome clones than in native plots. Split-plot regressions of the 2011 spaced-quadrat data showed that diversity decreased from edge to center of clones, as did cover of native plants. Native diversity, introduced cover, live cover, bare ground and litter did not vary significantly within clones. ADONIS (excluding perennial C3 grasses) indicated that the plant community within brome clones differed significantly from the plant community in native plots. Twelve taxa were positively correlated with native-plot canonical scores, compared to just three taxa with brome-clone canonical scores. Smooth brome appears to be homogenizing mixed-grass prairie by reducing the abundance of native and non-native species, and altering community characteristics such as cover of litter and live biomass. Further study with objectively selected brome areas is warranted.