Woody plant invasion of grasslands: establishment of honey mesquite (Prosopis glandulosa var.glandulosa) on sites differing in herbaceous biomass and grazing history

Emergence and survival of honey mesquite (Prosopis glandulosa var.glandulosa Torr.) seedlings was quantified on sites with contrasting grazing histories: long-term continuous grazing (LTG) and long-term protection (LTP) from grazing by cattle. On each site, different levels of heroaceous defoliation were imposed at monthly intervals (no defoliation=ND, moderate=MD and heavy=HD). The two weeks following seed dissemination appeared to be the most critical toProsopis establishment on LTP-ND plots. Openings in the herbaceous layer created by moderate defoliation of grasses on the LTP site increased germination and/or survival 7-to 8-fold during this period. However, increasing the degree of defoliation from moderate to heavy did not stimulate additional emergence on either the LTP or LTG site. Emergence from scarified seed placed in cattle dung (17 to 30%) was lower than that of bare seed placements in various microhabitats (43-60%). However, deposition of scarifiedProsopis seed in dung in conjunction with graminoid defoliation may be the most likely combination of events when livestock are present. Emergence from seeds transported into grasslands by other fauna likely would be low, unless seeds were deposited in areas where grasses had been defoliated.Prosopis survival was comparably high in dung and bare seed placements after one growing season. survival of seedlings present two weeks after seed dissemination ranged from 74 to 97% at the end of the second growing season. Seedling survival and shoot development (biomass, leaf area and height) were similar on LTP and LTG sites, regardless of the level of herbaceous defoliation or seed placement. In addition, the magnitude and patterns of net photosynthesis, stomatal conductance and xylem water potential were comparable among one-year-old seedtings on ND, MD and HD plots, even though differences in herbaceous species composition and above- and below-ground biomass between these treatments were substantial. Such data suggest competition for soil resources between grasses andProsopis may be minimal early in the life cycle ofProsopis. High rates ofProsopis emergence and establishment on LTP-MD plots are counter to the widespread assumption that long-term and/or heavy grazing is requisite forProsopis encroachment into grasslands. Results are discussed with regard to factors contributing to the recent, widespread invasion of this woody legume into grasslands of southwestern North America.