No abstract

Large areas of the Brazilian savanna (cerrado) have been converted into farmland in recent years; however, little attention has been paid to the consequences of this land use and land cover change on groundwater recharge. Here, we assessed groundwater recharge in different physiognomies of the cerrado located in an outcrop area of the Guarani Aquifer System. Water table fluctuations were measured from October 2011 through August 2013, by 58 monitoring wells distributed on four physiognomies of the undisturbed cerrado. We used multiple monitoring wells located in “campo limpo” (cerrado grassland), “campo sujo” (shrub cerrado), “campo cerrado” (open wooded cerrado), and “cerrado sensu stricto” (wooded cerrado), cover types. Recharge rates were computed for each well using the water table fluctuation method. The measured precipitation for hydrological years 2011–2012 and 2012–2013 were 1247 and 1194 mm, respectively. We found values of average annual recharge of 363 ± 87 mm, 354 ± 85 mm, 324 ± 78 mm, and 315 ± 76 mm for “campo limpo,” “campo sujo,” “campo cerrado,” and “cerrado sensu stricto,” respectively. Our results suggest that recharge tends to decrease with the increase in the density of vegetation (grassland to woodland). These results indicate that water table depth may have an influence on the cerrado physiognomies or vice versa. Furthermore, replacement of undisturbed dense cerrado with croplands will likely alter recharge dynamics. Therefore, sound management of land use and land cover is needed to ensure future groundwater quantity and quality.

No abstract

No abstract

The impacts of soil properties and fire regime on Neotropical savannas are wellknown, but the importance of hydrological regime for plant species assembly has received less attention. Here we assessed changes in diversity patterns of herbaceous and woody communities along a water table gradient in a fire-excluded Neotropical savanna. We found that increased waterlogging of soils was associated with declines in both herbaceous and woody species richness. Woody species richness decreased once the water table depth is less than 4 m and no woody species occurred once water table depth was less than 23 cm. Herbaceous communities remained species rich until the shallowest water table depth, where there is flooding at some point in the year, and even there, over a dozen species occurred. Woody species that occurred in areas with shallower water tables were a nested subset of those in areas with deeper water tables. In contrast, herbaceous communities showed turnover over the hydrological gradient, with distinct species specialised for different water table levels. However, we found that those specialists are restricted to few evolutionary lineages, evidenced by increased phylogenetic clustering over the water table gradient in herbaceous communities. We suggest that evolutionarily conserved hydrological niches define the herbaceous layer over the hydrological gradient, whereas only generalist woody species persist under high water tables. Our findings show that the effect of soil waterlogging differs between the herbaceous and woody layer of savannas, indicating that these communities will respond differently to shifts in the hydrological regime under future environmental change.

Communities subject to stress, including those with low invasibility, may be dominated by exotic generalist species. African grasses are aggressive invasive species in Neotropical savannas, where their response to abiotic stress remains unknown. We assessed the role of waterlogging and canopy closure on the presence, abundance and reproductive tillering of African and native grasses in a Neotropical savanna in southeastern Brazil. We obtained abundance and reproductive tillering data of exotic (Melinis minutiflora, Melinis repens and Urochloa decumbens) and common native grasses in 20 sites. We also determined the groundwater depth, soil surface water potential and canopy cover at these sites. The grass species generally had a low frequency and performed poorly where soil remained waterlogged throughout the year, except for two native species. Most native species were exclusive to either well-drained savannas or better drained wet grasslands. However, two species (Loudetiopsis chrysothrix and Trachypogon spicatus) occurred in both vegetation types. Two exotic species (M. minutiflora and M. repens) were less common but demonstrated reasonable performance in wet grasslands, possibly due to their root system plasticity. Furthermore, U. decumbens had a lower occurrence, density and reproductive tillering at these sites, but was successful at sites where the groundwater level was slightly deeper. Although the favourable water regime in the savannas increases their invasibility in general, resistance to invasion by African grasses may be greater at microsites with high canopy closure, where these species showed lower performance and did not affect the abundance of co-occurring native grasses. In summary, the Brazilian savanna becomes more susceptible to the spread of African grasses when disturbances decrease canopy closure or lower rainfall associated with climate change reduces the average groundwater depth and consequently releases invasive species from soil waterlogging in grasslands.

No abstract

No abstract