Population genomics of the orange roughy

The orange roughy is a deepwater fish found along continental slopes and on seamounts between depths of 450 and 1800m in the Atlantic, Indian and Pacific Oceans. Major fisheries off Australia, New Zealand, Namibia, Chile and in the North Atlantic were initially productive, but then needed to be reduced substantially to prevent collapse of the stocks. At one stage there were 50,000 tonnes per year being taken in the fishery off New Zealand. Part of their vulnerability relates to their behaviour of gathering in huge schools on seamounts to breed, where they can be easily taken by fisheries. They are also very slow to mature and can live to more than 150 years old. These characteristics make population recovery slow. Therefore there is some urgency to understanding how fisheries may be impacting natural populations. In the North Atlantic we had undertaken an earlier study that showed little evidence for differentiation among regional populations, however the resolution was relatively low. In the study facilitated by our grant from the Leverhulme Trust, we applied genomic technologies to permit a much more detailed assessment. An ongoing study based in Oceania (with colleagues Phillip England & Andres Gonçalves da Silva at CSIRO, Tasmania) was extended to include samples from the North and South Atlantic.

roughy jpg
Images by National Institute of Water and Atmospheric Research Ltd, with funding from the New Zealand government through Ministry of Fisheries and Land Information New Zealand Ocean Survey 20/20.


Instead of the 14 genetic markers applied previously, we now had 5600 to work with. Now we could see structure among regional populations that we hadn’t seen before. There was evidence for differentiation between coastal margin populations and those at seamounts in the mid-Atlantic, but also for coastal margin populations either side of a large marine bank (the Porcupine Bank – a submarine mound off Ireland). The differences are small, but real. As we’d seen earlier, there was a comparatively strong difference between the populations in the Northern and Southern Hemispheres, as may be expected. This method also permits something quite new compared to the earlier work. Some proportion of the genomic markers (in this case about 5%), show evidence for being under natural selection. When we focus on these markers in particular, we get quite a different pattern of population structure. This time one of the seamount populations (Faraday Seamount) stands out as quite distinct, and the other populations in the North Atlantic cluster together. Our next task in this ongoing project is to try to find out what types of genes may be involved, and why that population may be showing signs of local adaptation.

As this study demonstrates, population structure revealed by ‘neutral’ genes can be quite distinct from that seen for functional genes under natural selection, and local adaptation can happen even when there is no apparent structure for neutral genes. For the development of effective conservation strategies we need to be increasingly aware of this type of structure, as the future survival of species in changing environments may depend on it.

Professor Alan Hoelzel
University of Durham

Alan was awarded an Emeritus Fellowship grant in 2011.