African buffelgrass infestation in the Sonoran Desert of Arizona: A preliminary plant community assessment

Background/Question/Methods

Non-native invasive plants alter ecosystem functioning and threaten vulnerable species and their habitats. In the Sonoran Desert, relationships between plant invasion and increases in fire frequency and magnitude are a growing concern. African buffelgrass (Cenchrus ciliaris) is a relatively common invasive species in the Sonoran Desert that is capable of altering community structure, diversity, and disturbance regimes. For this study, we aimed to test preliminary methodologies for a large regional project seeking to develop empirically based, high-resolution models and maps of plant invasion. We visited sites in south-central Arizona to document native vs. non-native invasive plant communities and biomass, aiming to capture the invasion patterns of buffelgrass. Based on expert knowledge of known infestations, 16 plots were selected at which five subplots were established. At each subplot, we measured plant abundance using 25 regularly spaced point intercepts on a 25x25m grid. We recorded plant species and height at every point intercept and collected biomass of buffelgrass and native herbaceous plants within a 0.33 m2 circlet at up to eight point intercepts per plot. Meanwhile, we produced five habitat suitability models for buffelgrass based on known locations of buffelgrass occurrences and different combinations of environmental variables, including precipitation, topography, and substrate.

Results/Conclusions

Habitat suitability models indicated that our sampling sites represented medium to high suitability for potential buffelgrass habitats. Sampling sites comprised 62 native and four invasive species, with a mean 72% difference in species composition between paired sites (Sørensen’s dissimilarity index ranged from 0.3-1). Buffelgrass was present in 47% of the sites and 11% of total sampling points, whereas natives were present in all 77 sites and 46% of total sampling points. The percentage count for buffelgrass was 4-76% at each site in an average of 24%. Biomass relative to percentage count of buffelgrass was one to 22 times that of native herbaceous plants that occurred at the same site, with an average of nine times. A significantly negative relationship was found between percentage counts of buffelgrass vs. native plants. Linear regression indicated that both percentage count and biomass of buffelgrass were significantly related to summer precipitation and slope. Similarly, percentage count and biomass of buffelgrass showed significant relationships with habitat suitability models when summer precipitation was included as one of the predictor variables. This preliminary assessment demonstrates the magnitude of buffelgrass invasion and some of the environmental factors that are likely to influence buffelgrass infestation.