The lack of progress in establishing ambitious and legally binding global mitigation targets means that the need for locally based climate change adaptation will increase in vulnerable localities such as Africa. Within this context, "ecosystem-based adaptation" (EBA) is being promoted as a cost-effective and sustainable approach to improving adaptive capacity. Experience with the ongoing development of Durban's Municipal Climate Protection Programme indicates that achieving EBA in cities means moving beyond the conceptualization of a uniform, one-size-fits-all layer of street trees and parks to a more detailed understanding of the complex ecology of indigenous ecosystems and their resilience under climate change conditions. It also means engaging with the role that this "bio-infrastructure" plays in improving the quality of life and socioeconomic opportunities of the most vulnerable human communities. Despite the long-term sustainability gains of this approach, implementation in Durban has been shown to be both technically challenging and resource intensive. The close association between human and ecological systems in addressing climate change adaptation has also led to the development of the concept of "community ecosystem-based adaptation".KEYWORDS bio-infrastructure / community ecosystem-based adaptation / Durban / ecosystem-based adaptation / green economy / local government

This paper reflects on the progress made in climate change adaptation in the city of Durban since the launch of the Municipal Climate Protection Programme in 2004. This includes the initial difficulties in getting the attention of key sectors within municipal government, and how this was addressed and also served by the more detailed understanding of the range of adaptation options and their cost-benefits. There is also a better understanding of the potentials and constraints on community-based adaptation and the opposition from some landowners to measures to protect and enhance ecosystem services. The paper ends with lessons learnt that contradict some common assumptions – for instance, what approaches best build support for climate change adaptation within local governments, what measures work and from where lessons can be drawn. It also describes the perhaps unexpected linkages between local action and international influence and highlights the need for international climate change negotiations to recognize the key roles of urban governments in developing locally rooted adaptation and resilience.

The annual movement of South African sardine Sardinops sagax up the east coast of South Africa, known as the 'sardine run', was investigated using data from aerial surveys for the period 1988-2005 and compared with remotely sensed sea surface temperature (SST) and chlorophyll a data. Sardine sighting rates were highest within the Waterfall Bluff Bight off the Eastern Cape Coast, where conditions appeared to be most favourable. Sardine and predator sightings decreased significantly northwards of Mdoni on the KwaZulu-Natal (KZN) coast, whereas the proportion of nearshore sightings increased. The causal mechanism for this inshore concentration is suggested to be the influx of warm Agulhas Current water from the Durban Eddy that forces sardine shoreward. Cape gannet Morus capensis, common dolphin Delphinus capensis and sardine distributions were associated, and there was an association between SST and sardine and predator distributions. There was a marked increase in bottlenose dolphin Tursiops aduncus sightings upon commencement of the sardine run, with these dolphins being considered to be a 'migratory' stock that enters KZN waters every winter.

This paper describes the design and application of a benefit-cost model to the city of Durban's (South Africa) climate change adaptation options. The approach addresses the inability to compile an accurate damage-cost function for economic prioritizations at the local level. It proposes that uncertainty over climate impacts and the efficacy of adaptation responses, in conjunction with the lack of economic data, high levels of economic informality and inequality make it difficult to link adaptation efforts to positive GDP impact in Durban. Instead, the research based its calculations of "benefits" on the number of people impacted and the extent of the welfare benefits imparted by the respective adaptation efforts. It also took into account the uncertainty over future events, capacity constraints, priorities of decision makers and the risk of maladaptation. The results were reported as benefit-cost ratios for 16 clusters of interventions (many of which were primarily the responsibility of one municipal department or agency) in each of four future scenarios (defined by minor or major climate change and weak or strong socio-institutional capacity). The paper presents and discusses the benefit-cost ratios and total benefits for each of the intervention clusters in each of the future scenarios. It emphasizes how these are influenced by choices of time frames. It also highlights how the most efficient interventions across all futures and time frames tend to be socio-institutional − for instance the creation of a cross-sectoral disaster management forum, sea level rise preparedness and early warning system, and creating climate change adaptation capacity within the water services unit. Ecosystem-based adaptation measures had moderate benefit-cost ratios, probably because in Durban the land that needs to be purchased for this is relatively expensive. Infrastructure-based clusters generally had the lowest benefit-cost ratios.

The nearshore presence of sardine Sardinops sagax on the KwaZulu-Natal (KZN) coast was investigated using sightings data collected by the KZN Sharks Board from 1997 to 2007. The spatio-temporal distribution of sardine was described in relation to that of their predators and to environmental conditions, and subjected to generalised linear model (GLM) and generalised additive model (GAM) analyses. Variables describing spatio-temporal conditions performed best in the models (r 2 = 0.52) with seasonal effects, specifically June and July, making the greatest contribution towards sardine presence. The contribution of the years 2003, 2006 and 2007, and the KZN North Coast, was significantly lower. The predator variables were highly significant (r 2 = 0.48) with Cape gannets Morus capensis, followed by the sharks/gamefish and common dolphins Delphinus capensis, being most closely associated with sardine presence. Environmental variables were not as influential in the GLM models (r 2 = 0.23), but some variables were useful in describing conditions favouring sardine presence, namely calm current conditions, light north-westerly land breezes and stable atmospheric conditions. Increasing sea surface temperature (SST), moderate north to south currents, large swells and turbid water had a negative impact upon sardine presence. North-easterly and north-westerly winds and north to south currents had a cooling effect upon nearshore SSTs, whereas south-easterly winds and increasing air temperatures caused nearshore warming. Results are discussed in the context of developing an understanding of the mechanisms that govern fine-scale movements of sardine shoals during the KZN sardine run, with a view to predicting such movements.