Woody plant establishment and spatial heterogeneity in grasslands

Root exclusion experiments demonstrated the importance of belowground competition between grasses and Prosopis glandulosa (honey mesquite) during the critical seeding establishment phase of the woody plant life cycle. Belowground available volume accounted for 67% and 79% of the variance in first- and second-year Prosopis seedling growth and survival, respectively. Available volume in the vertical dimension was more important than that in the horizontal dimension. Trials spanned years with contrasting annual precipitation, suggesting that root competition occurs in years of near-average as well as below-average annual rainfall. Spatial heterogeneity in canopy gaps and belowground biomass was also quantified in a Schizachayriuim-Paspalum grassland matrix and evaluated with respect to Prosopis seedling establishment. Of the 100 grid points encountered in four grass stands, 62% were unoccupied; and 50% of these exceeded 80 cm² (∼10 cm diameter). Gaps ≥ 10 cm in diameter were sufficient for successful Prosopis germination and survival after one (40%) and two (15%) growing seasons. Herbaceous root biomass was statistically comparable among stands, but point-specific biomass varied three orders of magnitude (<50 g/m² to >3000 g/m²). In addition, root biomass was temporally variable, ranging from a mean (± 1 SE; g/m²) of 768 (86) in a year of below-average annual rainfall (1996; 721 mm) to 1108 (104) in a year of roughly average annual rainfall (1995; 1032 mm). It is often assumed that grasslands dominated by productive, late seral species will be resistant to woody plant encroachment; and that grazing, by reducing the ability of grasses to competitively exclude woody seedlings, makes grasslands susceptible to tree/shrub invasion. However, given the substantial annual variation in belowground biomass observed in this study, it seems reasonable that ungrazed grasslands may be more susceptible to woody plant encroachment in some years and more resistant in others. Furthermore, given the substantial spatial variation in aboveground gap area and belowground biomass, there may be numerous low-competition microsites for woody plant seedlings within ungrazed grasslands. A high degree of temporal and spatial variability in gap area and belowground biomass may therefore help explain successful establishment of woody seedlings in ungrazed or lightly grazed, late seral grasslands in the absence of fire. We found no correlation between aboveground structure and belowground biomass at scales of 1-10 m². Thus, readily quantifiable attributes such as grass basal area or gap area could not be used to infer site susceptibility to woody plant seedling establishment.