Assessing the risk to Aquaculture from harmful algal blooms (HABs)

As a new initiative in 2018, the Bermuda Institute of Ocean Sciences is funding research as part of XL Catlin’s Ocean Risk Scholarships to examine and quantify risks to ecosystems, businesses and people from the changes taking place in the ocean. The project will seek to improve HAB early warning skill for the benefit of the aquaculture industry.

Wild fish catches have been static for the last 25-years, and the growing shortfall in fish supply must be met by aquaculture. Hence, with increasing global demand for protein, the size of coastal aquaculture is increasing dramatically worldwide with an estimated value of 75-billion euros.

One of the largest, but currently unquantified risks to coastal aquaculture is that from Harmful Algal Blooms (HABs). Effects of these blooms are two-fold. 1) Mortalities of farmed fish, with associated financial consequences, 2) Human illness through biotoxins via shellfish consumption.
The potential large-scale impact of HABs was vividly illustrated by the blooms of Pseudochatonella that caused mass mortality of farmed fish in Chile in 2016. The blooms killed 39-million salmon mostly in Reloncavi Sound and Fjord reflecting a harvest weight of 100,000 metric-tons and a value of US$800 million (Anderson et al. 2017).

HABs are not a single entity. Harmful genera/species all have different ecology/life cycles and hence a single one size fits all risk assessment is not appropriate. Most HAB organisms are dinoflagellates but some are diatoms. Blooms of some organisms occur locally from excystment of resting cysts within the sediment. Other genera are thought to be holoplanktonic with blooms initiating offshore and being advectively transported to coast. Some HAB genera are autotrophic, gaining their energy from photosynthesis and their nutrition from mineral nutrients, others are mixotrophic, supplementing their nutrition by heterotrophically ingesting prey cells (Reguera et al. 2012).

Fortunately, the number of harmful organisms is limited. In the UK, the genera Alexandrium, Dinophysis and Pseudo-nitzschia are of most concern to shellfish aquaculture through production of saxitoxin (paralytic shellfish poisoning), okadaic acid and derivatives (diarrhetic shellfish poisoning) and domoic acid (amnesic shellfish poisoning).

UK farmed fish mortality most frequently results from the species Karenia mikimotoi and spiney diatoms of the genus Chaetoceros.

With the exception of a few locations where HABs can be linked to anthropogenic nutrient loadings (e.g. waste water/sewage effluent), blooms are spatially and temporally variable and are thought to be dependent on a number of interacting environmental drivers (Gowen et al. 2012). Most HABs, including in Scotland, occur naturally and are not directly anthropogenically generated. Prevention of HAB events in aquaculture areas is therefore not possible. Mitigation of their impact is thus necessary, and best achieved through understanding of when/where events will occur and early warning that allows remedial action to be undertaken (moving fish cages, early harvesting of shellfish/delay in harvesting until toxins depurate).

Two forms of improved HAB risk assessment are needed:
1) Better understanding of the long term and likely changes in key HAB species/genera and their link to environmental conditions in different regions
2) Methods to model the short-term development of HABs at aquaculture sites to allow action to be taken in response to these events.

While funding constraints has increasingly limited the collection of multi-year high-resolution data sets, we are fortunate that phytoplankton and HABs in particular have bucked this trend with a number of data sources being available. These include:

1) Continuous Plankton Recorder data collected from plankton tows over a large spatial scale for many years.
2) HAB Regulatory monitoring data. E.g., 1996 in Scottish waters harmful phytoplankton have been monitored at aquaculture sites. Since 2005 this monitoring has been on a weekly basis at ~ 40 sites. Similar monitoring programmes are operated in EU countries and elsewhere.
3) The harmful algal event database HAEDAT records harmful algal events since 2003 in ICES countries. (http://envlit.ifremer.fr/var/envlit/storage/documents/parammaps/haedat/)

These and other HAB data provide an excellent data resource to study the temporal and spatial variability in the key harmful algal genera of concern to aquaculture. Their analysis in conjunction with measured or modelled environmental information related to e.g. chlorophyll, water-temperature, salinity, currents, rainfall, stratification etc. provide the potential to evaluate how different HAB genera or species will respond to climate change and to develop models of these processes.