A study published in Science magazine contains new population assessments for thousands of fisheries around the globe, providing insight on the health of data-poor fisheries that account for more than 80 percent of the world's catch. The research confirms suspicions that these fisheries are in decline, but it also highlights hope for the future: most of these fisheries have not yet collapsed. If we act quickly to prevent overfishing and allow depleted stocks to recover to sustainable levels, they could provide more seafood over the long-term. This could increase the amount of fish brought to shore by 8-40 percent on average - and more than double it in some areas - compared to yields predicted if we continue current fishing trends.
But on a global scale the odds of better policies seem slim to me. The political map of Africa especially seems to work against good policies. Too many different countries have borders on the ocean. Fish that don't get fished for off of one country will get fished for in the next country. That contrasts with the United States, Canada, and Russia which have very long ocean borders. Coordination of restraints on fishing are far easier to achieve if the country that refrains from overfishing in its own waters will reap most of benefits.
Even if the yields could be increased the potential increase is a fixed max which is not huge. Most people aren't going to be able eat wild fish. The future of fish production growth has got to be from farmed fish. The wilds can't produce enough to feed the world. The human population has become too large for that.