Potential fishery benefits of Marine Protected Areas (MPAs) are widely acknowledged, yet seldom demonstrated, as fishery data series that straddle MPA establishment are seldom available. Here we postulate, based on a 15-year time series of nation-wide, spatially referenced catch and effort data, that the establishment of the Goukamma MPA (18 km alongshore; 40 km 2 ) benefited the adjacent fishery for roman (Chrysoblephus laticeps), a South African endemic seabream. Roman-directed catch-per-unit-effort (CPUE) in the vicinity of the new MPA immediately increased, contradicting trends across this species' distribution. The increase continued after 5 years, the time lag expected for larval export, effectively doubling the pre-MPA CPUE after 10 years. We find no indication that establishing the MPA caused a systematic drop in total catch or increased travel distances for the fleet. Our results provide rare empirical evidence of rapidly increasing catch rates after MPA implementation without measurable disadvantages for fishers.

A ntarctic krill Euphausia superba, a key species in Southern Ocean food webs 1 , plays a central role in ecosystem processes and community dynamics of apex predators, and is the target of a commercial fishery 2 . Krill spawn in late spring and their larvae develop during summer, autumn and under the ice in winter to emerge as juveniles in the following spring. The newly spawned eggs sink from the surface to up to 1,000 m depth where they hatch and the developing larvae actively swim upwards to feed in the upper water column 1 . Larval Antarctic krill have low lipid reserves, which are insufficient to support long periods of starvation. Therefore, winter, when primary production is minimal, is assumed to be a critical bottleneck for larval krill development and hence recruitment to the adult population 3,4 . Present hypotheses suggest that high algal biomass in winter sea ice enhances larval krill winter-feeding conditions and growth [5][6][7][8] . This implies that larvae have access to this high algal biomass within the sea ice. However, recent observations 9,10 indicate that the linkage between sea ice and krill recruitment success is not as direct as has been suggested. The timing of ice-edge advance and annual ice-season duration is highly variable, and does not necessarily show a clear link to krill recruitment in the following year ( Supplementary Figs. 1 and 2). Along the Antarctic Peninsula, adult krill have a five to six year population cycle with oscillations in biomass exceeding an order of magnitude 9 . According to bioenergetics models, part of the variability is due to interannual variation in reproductive output 11 as well as autumn blooms that may govern the possible overwinter survival rate of larvae 11,12 . In the Bransfield Strait, three krill winter surveys have shown that krill abundance is an order of magnitude higher than in summer, regardless of concurrent sea-ice conditions 10 A dominant Antarctic ecological paradigm suggests that winter sea ice is generally the main feeding ground for krill larvae. Observations from our winter cruise to the southwest Atlantic sector of the Southern Ocean contradict this view and present the first evidence that the pack-ice zone is a food-poor habitat for larval development. In contrast, the more open marginal ice zone provides a more favourable food environment for high larval krill growth rates. We found that complex under-ice habitats are, however, vital for larval krill when water column productivity is limited by light, by providing structures that offer protection from predators and to collect organic material released from the ice. The larvae feed on this sparse ice-associated food during the day. After sunset, they migrate into the water below the ice (upper 20 m) and drift away from the ice areas where they have previously fed. Model analyses indicate that this behaviour increases both food uptake in a patchy food environment and the likelihood of overwinter transport to areas where feeding conditions are more favourable in spring.

No abstract