Invasion of campo Cerrado by brachiaria (Urochloa decumbens): diversity losses and management techniques for restoration

Abstract

Invasion by African grasses has been one of the major challenges for conservation and restoration of the Brazilian savanna, particularly for open physiognomies. These invasive species cause biodiversity losses, modify the structure of plant communities,increase considerably the amount of fuel and, therefore,intensity and frequency of wild fires. Ecological processes related to invasion by brachiaria (Urochloa decumbens (Stapf.) RD Webster) are still poorly understood and this hinders the actions of management and conservation. The proposed research aims at: 1) characterize the factors and processes involved in biodiversity losses after invasion by in areas of cerrado grassland, 2) quantify the rate at which invasive species moves on natural ecosystems and especially 3) testing techniques to control the invasion which can be environmentally and economically viable. The study will be carried out at Santa Bárbara Ecological Station (Aguas de Santa Bárbara, state of Sao Paulo, Brazil). Losses in diversity resulting from the invasion will be assessed by comparing richness, density and ground cover of different vegetation components (woody plants, forbs, native and exoticgrasses) in invaded and non invaded plots. For this comparison, 20 pairs of plots of 2.0 mx 2.0 m will be sampled, each pair consisting of one invaded and another non-invaded plot, across five sectors of the Ecological Station. The microclimatic variables air temperature and relative humidity, and photosynthetically active radiation (taken 5 cm above the soil surface in each plot) will be investigated as factors possibly related to diversity losses. The rate of expansion of the invasive species will be quantified by monitoring over two years the expansion of 20 isolated invaded patches from their initial perimeter. That will permit estimating the increase of the invaded areas if invasion is not controlled. The eradication experiment will be installed in an area totally invaded, where nine treatments (five randomized replicates each), will be tested:i ) prescribed burning, ii ) herbicide, iii ) hand weeding, iv ) burn + herbicide, v) burn + hand weeding, vi ) burn + soil escarification + herbicide, vii) control (brachiaria with no treatment), viii ) reference (non invaded ), ix ) reference + burn. Prescribed burning will be applied just once in all treatments. Complementary techniques (herbicide and weeding) will be reapplied as often as necessary to control the invasive grass. Treatments will be compared on the basis of costs, richness, density and cover by native vegetation. Our expectation is that, after eradication of the invasive grass, the native plants remaining will spread over the treated plots, triggering their ecological restoration. Hopefully, the study will not only clarify the mechanisms through which invasive grass leads to loss of biodiversity in open cerrado physiognomies, but especially to provide subsidies for management, conservation and restoration of these ecosystems. (AU)