Low forest-loss thresholds threaten Amazonian fish and macroinvertebrate assemblage integrity

2021 
Abstract Deforestation is a major threat globally, but especially in tropical regions because they are biodiversity strongholds and carbon storehouses. Some studies have reported changes in species richness and composition in lotic ecosystems with increased forest-loss in their catchment, presumably resulting from the replacement of sensitive taxa by more resistant or tolerant taxa. Also, sensitive taxa respond to deforestation in a non-linear manner and fish and macroinvertebrates have different sensitivities to landscape pressures. Therefore, it is useful to determine the effects of forest-loss on widespread sensitive or threshold taxa in aquatic ecosystems. We used Threshold Indicator Taxa Analysis (TITAN) to assess forest-loss and land use history impacts in 92 eastern Amazonian stream sites. We determined TITAN peak-change thresholds for fish at 1% and 6% of forest-loss at total-catchment and local-riparian spatial extents, respectively, and at 2% and 40% of land-use intensity change at total-catchment and local-riparian spatial extents, respectively. For macroinvertebrates, TITAN peak-change thresholds were 1% and 11% of forest loss at total-catchment and local-riparian spatial extents, respectively, and at 3% of land-use intensity change for both total-catchment and local-riparian spatial extents. Because of these thresholds, inherent ecoregional variability and key literature, we have three major recommendations. 1) Logging should be prohibited in riparian reserves that are at least 100-m wide on each side of headwater streams and in a network of catchments across all biomes and as many landscape types as possible. 2) An ecologically and statistically rigorous monitoring program with standard methods should be implemented to assess and regulate land uses better. 3) Conservation planning areas should consider aquatic biota as well as terrestrial biota.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    105
    References
    6
    Citations
    NaN
    KQI
    []