VU

Critical Zone Hydrology Group

Home

Research

Global change impacts

Permafrost hydrology
Flood forecasting
Cloud forest hydrology
Tropical deforestation
Reforestation

Sustainable aquifer use

Teaching

People

Valorisation and services

Photo gallery

Links


Resilience of Amazonian ecosystems to reduced rainfall in relation to small-scale topographic variation

Researcher: Fabricio B. Zanchi
Affiliation: Universidade Federal do Amazonas (UFAM) - Instituto Agricultura e Ambiente. Humaitá, Amazonas State, Brasil

Supervisors:

Collaborators:

Duration: Field phases 2006-2008, analysis 2008-2012
Funding: Dutch Government NWO WOTRO Integrated Programme grant

Project outline

In Central Amazonia, the climate does not show much seasonality. The landscape generally consists of a mosaic of well-drained dissected plateaus, separated by steep slopes and broad, swampy river valleys. Permanently wet soil status in the valley is maintained by a continuous supply of groundwater from the plateau and slope areas. Whilst oxisols and ultisols on the slopes and plateau have a significant clay fraction, valley soils generally consist of strongly leached quartz sands (spodosols) with little capacity to retain water or nutrients. Nutrients may be retained in the valley soil by the soil carbon fraction (humus), which also may buffer the pH to some extent, allowing tall forests to grow on these white sands. Under the present climate conditions the vegetation in the swampy valleys are not likely to experience significant drought and may even benefit from a slightly dryer climate as root water uptake, photosynthesis and decomposition may be reduced in saturated soils. However, if reduced recharge on the plateaus and slopes would result in a significant lowering of the valley groundwater table over extended periods, soil carbon could decompose and oxidise, leaving an extremely poor and potentially toxic soil. This might trigger a change from the existing tall valley forest to increasingly stunted Caatinga forest with more sclerophyllus leaves, locally called Campinarana or Campina forest. The local presence of Campinarana forest between the tall forest in the valley bottom and that on the slope is an indication that a relatively small change in the valley groundwater level may be sufficient to cause an extension of the Campinarana zone towards the river. Although these campina forests are quite common in the tropics, little is known about their evapotranspiration rates. Depending on the reduction in rainfall and the response of different forest types we can postulate three stages of forest development:

  1. Reduced rainfall caused by global change and local deforestation does not affect the functioning of the rainforest on the plateau and slope. However, recharge of plateau groundwater that keeps the valleys wet is reduced, resulting in a lowering of the valley groundwater table and reduced river runoff. The ensuing oxidation of organic matter in the spodosols and associated change in nitrogen cycling and pH triggers a change in the forest to a more Campinarana - Campina type of vegetation.
  2. The rainforest on the plateau and slope cannot maintain it's functioning and degrades to a vegetation cover with reduced water uptake. Groundwater recharge on the plateau and slope is maintained and the valleys remain wet. The reduction of evaporation on the plateau and slope may partly be compensated by increased evapotranspiration by the valley vegetation (advection) if this has not degraded to a campina forest in the first stage.
  3. The reduction in rainfall is such that the rainforest on the plateau changes to savannah, with a much lower evapotranspiration. Groundwater recharge is increased but the valley vegetation cannot compensate for the reduced evapotranspiration on the slopes and plateau, leading to an increase in runoff. This suggests that the valley vegetation may retain or even increase its production and evaptranspiration, or collapse and change to a campina type of forest with presumably much reduced production and evaporation rates.

The second possibility would mitigate the effects of a dying off of the forests on the plateau on the regional climate, whereas the first and third option represents a positive feedback where rainfall becomes even further reduced.

Drought sensitivity of the landscape and vegetation has been investigated by looking at soil respiration and decomposition of organic matter, canopy photosynthesis and transpiration processes (micro-meteorology and sapflow) in different natural forest (terra-firme and valley forests, campinarana, campina) and soil types (waterlogged sands, dry sands, unsaturated clayey soil). The investigation focussed on the response of the ecosystem to short-term (diurnal), medium-term (seasonal) and short distance (along topographic gradients) variation in water availability and drought. A controlled experiment lowering the water table was included to assess the impact on organic matter in the saturated valley soils.

Research questions

Project publications


Fabricio Berton Zanchi
Last modified: Wed Dec 30 13:20:58 CET 2015