Remote Forest Refugia for Fijian Wildlife

Abstract:  On Pacific islands non-native rats and mongooses threaten many native species. In Fiji we compared visitation rates of rats and mongooses at bait stations and measured biomass of leaf-litter invertebrates to assess the relative predation pressure from these species in forest areas at different distances from the forest edge. Forest areas over 5 km from the forest edge had significantly fewer baits encountered by rats or mongooses than did natural forest areas nearer agricultural and urban habitats. Remote forest areas may function as a last refuge for island species threatened by predation from non-native rats and mongooses. The biomass of leaf-litter invertebrates in remote forest areas was higher, indicating a refuge effect for some taxa targeted by rats and mongooses. Protection of the few remaining large blocks of natural forests on Pacific islands may be the most cost-effective approach for conserving many island endemics threatened by rats and mongooses. Logging roads can compromise this refuge effect by acting as dispersal routes for rats into natural forests.     

The Value of Rehabilitating Logged Rainforest for Birds

Abstract: The recent advent of carbon crediting has led to a rapid rise in biosequestration projects that seek to remove carbon from the atmosphere through afforestation and forest rehabilitation. Such projects also present an important potential opportunity to reverse biodiversity losses resulting from deforestation and forest degradation, but the biodiversity benefits of different forms of biosequestration have not been considered adequately. We captured birds in mist nets to examine the effects of rehabilitation of logged forest on birds in Sabah, Borneo, and to test the hypothesis that rehabilitation restores avian assemblages within regenerating forest to a condition closer to that seen in unlogged forest. Species richness and diversity were similar in unlogged and rehabilitated forest, but significantly lower in naturally regenerating forest. Rehabilitation resulted in a relatively rapid recovery of populations of insectivores within logged forest, especially those species that forage by sallying, but had a marked adverse effect on frugivores and possibly reduced the overall abundance of birds within regenerating forest. In view of these results, we advocate increased management for heterogeneity within rehabilitated forests, but we strongly urge an increased role for forest rehabilitation in the design and implementation of a biodiversity‐friendly carbon‐offsetting market.     

Indirect Effects of Biological Control on Plant Diversity Vary across Sites in Montana Grasslands

Abstract:  Biological control with specialist, nonindigenous, herbivorous insects is an important option for controlling invasive exotic plants in wildlands and nature reserves. It is assumed that biological control agents will reduce the dominance of the target weed, thereby increasing the native diversity of the associated plant community. However, this hypothesis has rarely been tested. We introduced Aphthona nigriscutis into grassland sites infested with the invasive exotic species Euphorbia esula L. on a nature reserve in Montana (U.S.A.). Two sites with better soil had been treated previously with herbicide, whereas two other sites had not. We measured the density and biomass of Euphorbia vegetative and flowering stems and number of native and exotic shrubs, grass-like plants, and forbs in 48 microplots in Aphthona release and control macroplots at each site. After 5 years, Aphthona release was associated with a 33–39% decline in Euphorbia aboveground biomass compared with controls at all sites. Other effects of the biocontrol depended on the site. Biocontrol slowed the recovery of species diversity at the sites previously treated with herbicide but slowed the loss of diversity at sites without a history of herbicide. Biocontrol introduction was not associated with a disproportionate increase in nontarget exotic species. Release of Aphthona caused a decline in the biomass of flowering stems relative to controls at good-soil, previous-herbicide sites but was associated with a relative increase in flower stem mass at poor-soil, no-herbicide sites. Our results suggest that biocontrol reductions in weed dominance will not always be associated with increased species diversity. More emphasis should be placed on conserving desirable communities and less on simple weed control. Monitoring of community-level effects should accompany biocontrol introductions on nature reserves.     

Armageddon Versus Extinction

Note from the Editor : Some readers may feel that topics such as politics and religion are inappropriate for a scientific journal. Our "Conservation in Context" column, however, is intended to cover the entire spectrum of the context within which the science of conservation biology must operate. Thus, we freely and openly explore topics in this column that ordinarily are not addressed in scientific journals, to better understand the real-world constraints and opportunities that define how conservation science is applied in a complex world.     

Correlations among Extinction Risks Assessed by Different Systems of Threatened Species Categorization

Abstract:  Many different systems are used to assess levels of threat faced by species. Prominent ones are those used by the World Conservation Union, NatureServe, and the Florida Game and Freshwater Fish Commission (now the Florida Fish and Wildlife Conservation Commission). These systems assign taxa a threat ranking by assessing their demographic and ecological characteristics. These threat rankings support the legislative protection of species and guide the placement of conservation programs in order of priority. It is not known, however, whether these assessment systems rank species in a similar order. To resolve this issue, we assessed 55 mainly vertebrate taxa with widely differing life histories under each of these systems and determined the rank correlations among them. Moderate, significant positive correlations were seen among the threat rankings provided by the three systems (correlations 0.58–0.69). Further, the threat rankings for taxa obtained using these systems were significantly correlated to their rankings based on predicted probability of extinction within 100 years as determined by population viability analysis (correlations 0.28–0.37). The different categorization systems, then, yield related but not identical threat rankings, and these rankings are associated with predicted extinction risk.     

Trends in Sex Ratios of Turtles in the United States: Implications of Road Mortality

Abstract:  Road mortality has been implicated as a significant demographic force in turtles, particularly for females, which are killed disproportionately on overland nesting movements. Moreover, the United States' road network has expanded dramatically over the last century. We therefore predicted that historical trends in sex ratios of turtle populations would be male biased. To test this prediction, we synthesized published estimates of population-level sex ratios in freshwater and terrestrial turtles in the United States (165 estimates for 36 species, published 1928–2003). Our analysis suggests that the proportion of males in populations has increased linearly ( p = 0.001); the trend in male bias is synchronized with the expansion of the surfaced portion of the road network since 1930; sex ratios became more male biased in states with higher densities of roads; and populations have become more male biased in aquatic species, in which movement differentials between males and females are greatest, and are least biased in semiaquatic and terrestrial species, in which overland movements are more comparable between sexes. Our results suggest an ongoing depletion of breeding females from wild turtle populations over the last century because of many factors, including, and perhaps chiefly, road mortality.