Search the SPREP Catalogue

Three mangrove restoration methods were tested at Nu'uuli, Tutuila island, American Samoa. Since cleaning 27 years ago converted the mangrove into a mudflat, the ecosystem showed no natural regrowth despite an ample supply of propagules.

Coral reefs are hotspots of marine biodiversity, and their global decline is a threat to our natural heritage.

This paper develops a model of invasive species control when the species’ population size is unknown. In the face of an uncertain population size, a resource manager’s species-control efforts provide two potential bene ts: (1) a direct bene t of possibly reducing the population of invasive species, and (2) an indirect bene t of information acquisition (due to learning about the population size, which reduces uncertainty). We provide a methodology that takes into account both of these bene ts, and show how optimal management decisions are altered in the presence of the indirect bene t of learning. We then apply this methodology to the case of controlling the Brown Treesnake (Boiga irregularis) on the island of Saipan. We nd that the indirect bene t—the value of information to reduce uncertainty—is likely to be quite large.

Five decades after a series of nuclear tests began, we provide evidence that 70% of the Bikini Atoll zooxanthellate coral assemblage is resilient to large-scale anthropogenic disturbance. Species composition in 2002 was assessed and compared to that seen prior to nuclear testing. A total of 183 scleractinian coral species was recorded, compared to 126 species recorded in the previous study (excluding synonomies, 148 including synonomies). We found that 42 coral species may be locally extinct at Bikini. Fourteen of these losses may be pseudo-losses due to inconsistent taxonomy between the two studies or insufficient sampling in the second study, however 28 species appear to represent genuine losses. Of these losses, 16 species are obligate lagoonal specialists and 12 have wider habitat compatibility. Twelve species are recorded from Bikini for the first time. We suggest the highly diverse Rongelap Atoll to the east of Bikini may have contributed larval propagules to facilitate the partial resilience of coral biodiversity in the absence of additional anthropogenic threats.

A group of researchers, managers and tribal leaders with extensive local knowledge from Papua New Guinea, Solomon Islands, Vanuatu and Papua, Indonesia, provided new information on the status of leatherback nesting populations in the western Pacific Ocean. Twenty-eight nesting sites were identified, of which 21 were previously unknown or poorly described. Although data are still incomplete, we estimate a total of ca. 5000-9200 nests currently laid at 4 sites along the northwest coast (Bird's Head Peninsula) of Papua, Indonesia. Genetic analysis by using mitochondrial deoxyribonucleic acid sequences identified a total of 6 haplotypes among the 106 samples analysed for Solomon Islands, Papua and Papua New Guinea, including a unique common haplotype that is only found in these western Pacific populations. There was no significant difference in haplotype that is only found in these western Pacific populations. There was no significant difference in haplotype frequencies among these rookeries, which suggests that they represent a metapopulation composed of a single genetic rock. Further work is needed to define the demographic structure within this metapopulation.

Internesting and migratory movements of female leatherback turtles (Dermochelys coriacea) from Kamiali Wildlife Management Area were tracked by satellite telemetry during the 2001-2002 and 2002-2003 nesting seasons, and an aerial survey of nearly 2800 km of the north Papua New Guinea coastline and New Britain Island was conducted during January 2004. Nesting occurred during November-March at the Mamiali Wildlife Management Area monitoring beach with peak activity during December-January. Telemetry results indicated that nesting females used Huon Gulf waters adjacent to the nesting beach and renested inside and outside the protected area within the Kamiali Wildlife Management Area. After nesting, the turtles migrated over deep waters in a southeastern direction to high latitude waters of the South Pacific Ocean.

In October 2005, my Google protected areas alert showed several links to the creation of new protected areas in Papua New Guinea (PNG). As I clicked on the links, I was already wondering if this issue could be my master's research topic in Summer 2006. I had no idea that the timing of this press release would be the focus of my efforts to understand who was actually achieving conservation in PNG.

This paper is based on presentations and discussions held during a marine managed areas (MMAs)2 session organised by NOAA, SPREP and Conservation International (CI) as part of the “Our Seas of Islands” Regional Forum for Oceania on MMAs convened by NOAA, UNESCO and partners held in Hawaii in January 2007. The overall objective of the session was to promote a shared understanding of the extent and diversity of MPA/MMA approaches and key lessons learned in their design, and to identify practical and strategic future actions to further apply these tools in the Oceania marine realm.

With the wide acceptance of forest?protection policies in the developing world comes a requirement for clear demonstrations of how deforestation may erode human well?being and economies. For centuries, it has been believed that forests provide protection against flooding. However, such claims have given rise to a heated polemic, and broad?scale quantitative evidence of the possible role of forests in flood protection has not been forthcoming. Using data collected from 1990 to 2000 from 56 developing countries, we show using generalized linear and mixed?effects models contrasted with information?theoretic measures of parsimony that flood frequency is negatively correlated with the amount of remaining natural forest and positively correlated with natural forest area loss (after controlling for rainfall, slope and degraded landscape area). The most parsimonious models accounted for over 65% of the variation in flood frequency, of which nearly 14% was due to forest cover variables alone. During the decade investigated, nearly 100?000 people were killed and 320 million people were displaced by floods, with total reported economic damages exceeding US$1151 billion. Extracted measures of flood severity (flood duration, people killed and displaced, and total damage) showed some weaker, albeit detectable correlations to natural forest cover and loss. Based on an arbitrary decrease in natural forest area of 10%, the model?averaged prediction of flood frequency increased between 4% and 28% among the countries modeled. Using the same hypothetical decline in natural forest area resulted in a 4–8% increase in total flood duration. These correlations suggest that global?scale patterns in mean forest trends across countries are meaningful with respect to flood dynamics. Unabated loss of forests may increase or exacerbate the number of flood?related disasters, negatively impact millions of poor people, and inflict trillions of dollars in damage in disadvantaged economies over the coming decades. This first global?scale empirical demonstration that forests are correlated with flood risk and severity in developing countries reinforces the imperative for large?scale forest protection to protect human welfare, and suggests that reforestation may help to reduce the frequency and severity of flood?related catastrophes.