Ecosystem consequences of non-native grass species removal in Puerto Rico

Background/Question/Methods

Management of introduced species is of primary concern for agriculture, habitat restoration, native species protection, and ecosystem function. However, invasive species are frequently present on the landscape due to past human practices and may be filling manmade niches that native species are not able to access successfully. Removal of invasive species, particularly at the local level (e.g. field, stand, or population) is frequently a management goal. However, removal may achieve some goals at the expense of others. For example, removal of invasive plant species may reduce weed control costs and competition against desirable native species or crop species, but this may cause increases in erosion or nutrient export from systems. The overall goal of this project is to assess trade offs that result from eradication of invasive plant species in Puerto Rico. We addressed this experimentally at the Laguna Cartagena Fish & Wildlife Reserve, Puerto Rico: a grass-dominated system that previously was forested and located adjacent to wetlands. We mechanically removed non-native grass by mowing or bulldozing in replicated plots. Bulldozing is the typical means of vegetation removal in PR. We measured biomass recovery, species richness, nutrient fluxes, and nitrogen fixation over the course of one year.

Results/Conclusions

Biomass recovered to pretreatment levels in both the mowed and bulldozed plots by 6 months, while the control plots showed no changes in biomass.  The dominant species in all plots were still non-native grasses, however, the bulldozed plots had higher species richness with 7 species compared to the 2 found in the control plots. The additional species were not native. Total plant available N was highest in the bulldozed treatment immediately after the non-native grass removal and quickly dropped to the levels found in the control and mowed treatments by 3 months.  The total N flux appears to be driven by the available nitrate in the system, and not ammonium.  Fluxes of available P were not detected in the bulldozed plots until 4 months after bulldozing. P fluxes in the control or mowed treatments were variable and showed no clear pattern over time.  Nitrogen fixation was found in all the plots. At least in the short term, vegetation removal did not result in higher native species richness while it provided a pulse of N that could increase N loads to adjacent wetlands or downstream systems.