An initial assessment of drought sensitivity in Amazonian fish communities

Abstract

The Amazon River Basin encompasses the world’s largest remaining tropical rainforest, and the largest freshwater system with the highest fish species diversity on earth, but global climate change is predicted to cause the loss of 7–12% of fish species by 2070. The severe drought anomaly of 2005, caused by warming of Atlantic surface waters, provided a unique opportunity to examine the impact of a major climatic disturbance on a tropical fish assemblage. We monitored fish species diversity in six Brazilian floodplain lakes along the Solimoes River from 2004 to 2007 (before, during, and after drought). Statistical analysis revealed changes in species composition of these lakes following the drought, with both positive and negative responses observed. The response to drought was not uniform among species with regard to trophic guild or migratory behavior. SIMPER analysis showed that planktivores on the average increased in abundance in the years following the drought, carnivores and omnivores decreased, and herbivores and detritivores increased. Some of these changes were transitory, others persisted through monitoring. Migratory species disproportionately increased in abundance post-drought compared to non-migratory species. Interlake (β) diversity of fish declined during the drought year, indicating that lakes were becoming less heterogeneous in species composition, but showed a trend toward recovery of pre-drought level in the following years. According to both global climate change models and recent experience, the intensity and frequency of droughts in this region of the world is increasing. Given the sensitivity of resident fish species to the single, short-term, perturbation reported here, assessment of how tropical freshwater fish populations respond to drought will be crucial to understanding the consequences of this kind of perturbation to these communities and to the human inhabitants who depend upon this important protein source.