A Stand-Level, Water-Addition Experiment

To evaluate hypotheses relating the biology of tropical forest species to seasonal variation in moisture availability, we eliminated seasonal drought in the soil and understory atmosphere by adding 114,750 m3 of water to two 2.25-ha plots of old?growth forest over five consecutive dry seasons. Irrigation prevented the shrink-swell cycle of the soil and increased soil pH and fertility, understory plant survival, and egg survival and recruitment for a lizard. Irrigation did not alter insect herbivory; the sea?sonality, abundance, or species composition of flying insects; nor the phenology and productivity of canopy trees. These negative results might be related, with insects track?ing the canopy plants that dominate productivity. Irrigation did not affect clouds, atmo?spheric transmissivity, and solar irradiance, and, for taller plants, atmospheric drought in the canopy. We hypothesized, and it has since been confirmed, that the seasonality of incoming solar radiation limits productivity and provides proximate phenological cues for many canopy species. Recent increases in dry-season rainfall observed across the tropics will alter soils and plant and animal populations through direct effects on soil moisture in humid tropical forest.