Background
The authors begin their short Brevia report by noting that "large-scale numerical models that simulate the interactions between changing global climate and terrestrial vegetation predict substantial carbon loss from tropical ecosystems, including the drought-induced collapse of the Amazon forest and conversion to savanna."

What was done
Saleska et al. used Terra satellite data -- Enhanced Vegetation Index (EVI) derived from the Moderate Resolution Imaging Spectroradiometer (MODIS) -- to determine whether or not the widespread Amazon drought of 2005, which peaked during the dry season onset (July-September), did indeed reduce whole-canopy forest photosynthesis as predicted, which they say "should have been especially observable during this period, when anomalous interannual drought coincided with the already seasonally low precipitation."

What was learned
Quite to the contrary of model predictions, the four researchers found that observations of intact forest "greenness" in the region were, in their words, "dominated by a significant increase (P<0.0001), not a decline."

What it means
Saleska et al. note that the increased forest greenness they observed "is inconsistent with expectation if trees are limited by water, but follows from increased availability of sunlight (due to decreased cloudiness) when water is not limiting," which suggests that water was, in fact, not limiting during this period of significant drought.

One phenomenon they offer to explain this seemingly strange scenario is that the trees of the Amazon forest may be utilizing deep roots to "access and sustain" water availability during drought. Another possibility is that the historical increase in the air's CO2 content has significantly enhanced the trees' water use efficiency, enabling them to produce considerably more biomass per unit of water transpired and thereby conserve water.

Strong evidence for this latter phenomenon is provided by the work of Robock et al. (2005), who analyzed 45 years of gravimetrically-observed plant-available soil moisture in the top one meter of soil for 141 stations from fields with either winter or spring cereals in the Ukraine over the period 1958-2002. This work revealed, in their words, "a positive soil moisture trend for the entire period of observation," and they emphasize that "even though for the entire period there is a small upward trend in temperature and a downward trend in summer precipitation, the soil moisture still has an upward trend [our italics] for both winter and summer cereals."

Reference
Robock, A., Mu, M., Vinnikov, K., Trofimova, I.V. and Adamenko, T.I. 2005. Forty-five years of observed soil moisture in the Ukraine: No summer desiccation (yet). Geophysical Research Letters 32: 10.1029/2004GL021914.