Search

1. Selective logging is the most widespread driver of land-use change in biodiverse and carbon-rich tropical forests. However, the effects of selective logging on bio-diversity are less than those associated with other drivers of forest degradation. A suite of recent research has shown that reduced-impact logging (RIL) results in few or no changes to biological assemblages. But because this logging technique is relatively new, most studies have only considered short-term impacts. 2. We address this research gap by quantifying changes in biodiversity assemblage as a result of RIL over the longer term. We comprehensively sampled bird and bat assemblages pre-logged, 1 year after, and 10 years after RIL in Guyana, using a before-after control-impact (BACI) sampling design. We compared bird and bat assemblages in each timeframe, and additionally appraised the impact of time since logging, and the number of trees harvested across the suite of species which we further divided between different feeding guilds, disturbance sensitivity and vertical stratification of forest use. 3. We found that 1 year after logging only minor changes could be detected, but 10 years later richness had slightly declined in some groups, while others had shown complete recovery. Nectivorous and insectivorous birds, and carnivorous bats declined in richness, while carnivorous birds, showed a clear recovery to a state akin to pre-logging. This indicates that for some niches a subtle, but long-term relaxation effect may be occurring, whereby extinction debts are realized long after the initial disturbance, while other groups have either recovered or not changed after logging. 4. Assemblage changes were also predicted by vertical stratification of forest use, with avian species using the understorey and mid–upper levels of the forest being most affected. 5. Synthesis and applications: Our study demonstrates how best practice forestry and logging can maintain healthy vertebrate populations over the lo

Selective logging is the most widespread use of tropical forests. Building logging roads facilitates access to previously remote rainforests, and so proper management is essential for ensuring biodiversity retention in logged landscapes. Terrestrial mammals often directly use logging roads (via movement corridors, hunting or foraging), making them vulnerable to poorly managed roads. Here we explore how the presence, arrangement and use of logging roads influence terrestrial mammal occupancy and detection within a Forest Stewardship Council (FSC) certified logged forest in Guyana. We compared camera trap data from20 natural ‘game’ trails in an unlogged area, with camera trap data from 23 sites set near to or on logging roads within the Iwokrama forest. Our findings showed high occupancy within logged areas with no statistically significant difference to unlogged areas. Higher detections were noted along secondary and feeder roads compared to skid trails and the natural trails in control areas. Additionally, our data showed a negative correlation between occupancy and distance to village for a scatter hoarding rodent, most likely driven by subsistence hunting by local communities. Our results indicate that proper road management geared towards the monitoring and guarded access of logging roads, can have a positive effect on terrestrial mammal occurrence within responsibly managed rainforests.

Many vertebrate species undergo population fluctuations that may be random or regularly cyclic in nature. Vertebrate population cycles in northern latitudes are driven by both endogenous and exogenous factors. Suggested causes of mysterious disappearances documented for populations of the Neotropical, herd-forming, white-lipped peccary (Tayassu pecari, henceforth "WLP") include large-scale movements, overhunting, extreme floods, or disease outbreaks. By analyzing 43 disappearance events across the Neotropics and 88 years of commercial and subsistence harvest data for the Amazon, we show that WLP disappearances are widespread and occur regularly and at large spatiotemporal scales throughout the species' range. We present evidence that the disappearances represent 7-12-year troughs in 20-30-year WLP population cycles occurring synchronously at regional and perhaps continent-wide spatial scales as large as 10,000-5 million km2. This may represent the first documented case of natural population cyclicity in a Neotropical mammal. Because WLP populations often increase dramatically prior to a disappearance, we posit that their population cycles result from over-compensatory, density-dependent mortality. Our data also suggest that the increase phase of a WLP cycle is partly dependent on recolonization from proximal, unfragmented and undisturbed forests. This highlights the importance of very large, continuous natural areas that enable source-sink population dynamics and ensure re-colonization and local population persistence in time and space.