Orange roughy

Stock assessments previous to 2008 are no longer considered to be reliable due to conclusions about stock structure that consider East and South Chatham Rise as a single biological stock (joining sub-areas that were previously assessed separately).

The 2014 stock assessment used an age-structured CASAL population model with Bayesian estimation of posterior distributions, fitted to acoustic-survey estimates of spawning biomass – from the Old-plume (2002–2013), Rekohu (2011–2013) and the Crack (2011, 2013), trawl-survey biomass indices (from the Spawning box, 1984-1994), age frequencies from spawning aggregations (2012 and 2013), and length frequencies from trawl surveys and commercial fisheries. Catch overruns at 5% were included (MPI, 2014a). Input data are considered of “high quality”; “low” and “medium” quality data (e.g. catch per unit effort index) were not used. Besides, the assessment meets the New Zealand’s Science and Research Information Standard for New Zealand Fisheries in place since 2011 (MPI, 2014b). Natural mortality (M) was assumed as a constant at 0.045 and a Beverton-Holt stock-recruitment relationship. Uncertainties relate to the proportion of the spawning stock that is indexed by the acoustic survey in each year, the timing of appearance of the Rekohu spawning plume which is unknown and is important for the determination of the stock status, patterns of year class strengths that are based only on 2 years of age composition data (MPI, 2014a,b) as well as in the estimation of BMSY, due to uncertainty in steepness and in the form of the stock-recruitment relationship (Cordue, 2014; MPI, 2014c).

Stock assessments are conducted by independent researchers contracted by the Ministry for Primary Industries (MPI) and the results are peer reviewed by the Deepwater Fisheries Assessment Working Group (DFAWG) which is composed by scientists of the National Institute of Water and Atmospheric Research (NIWA), the Ministry and the industry. The DFAWG, under MPI, evaluates the state of the fisheries and stocks and analyze management scenarios, not being responsible by management recommendations or decisions which are taken by the MPI. A final decision is discussed in a multi-stakeholder meeting.

According to the stock status and based on five years projections, two scenarios were considered: the current catch limit, at 3.100 tons, and the current estimated yield (assuming fishing intensity U35% B0), at 6.400 tons. Both are expected to lead to a steady increase in SSB in the next 5 years and the probability of reduction below Soft and Hard limits is close to zero (MPI, 2014a).

B_MSY was reassessed in 2014 at 26% of B₀ considering a Beverton-Holt stock-recruitment relationship (DWG, 2014c), which for this stock corresponds to 83,200 tons.

The management target range was also revised from 30-40% of B₀ to 30-50% of B₀ = 96,000–160,000 tons in order to cope with variability in recruitment and its effects on biomass (DWG, 2014c).

The fishing intensity target range was revised to U_{30%B0–U50%B0} and F_mid, defined at 4.5% of the current vulnerable biomass, set as the target at 40% B₀ (DWG, 2014c; Cordue, 2014).

The reference point previously called the “soft limit”: 20% of B₀ = 64,000 tons, has now been explicitly defined as B_lim, “the point at which catch limits should be significantly reduced to prevent serious impairment of recruitment (DWG, 2014c).

The previous “hard limit”, a ‘biomass level below which a stock is deemed to be “collapsed” where fishery closures should be considered in order to rebuild a stock at the fastest possible rate’, and defined as 10% of B₀ = 32,000 tons, had been considered by the MSC pre-assessment report to be B_lim and was considered to be inadequate for the management of this fishery (Punt et al., 2013).

B₀ is estimated at 320.000 tons. B_current has been increasing in the last years and in 2014 is at 30% of B₀= 95.000 tons (MPI, 2014c); is Unlikely (< 40%) to be below the previous “soft limit”, Very Unlikely (< 10%) to be below the previous “hard limit” and About as Likely as Not (40–60%) to be at or above the lower end of the management target range (MPI, 2014a,d). Overfishing is Very Unlikely (< 10%) to be occurring, fishing intensity is estimated at U_52%B0, meaning that fishing at intensity of 52 will cause SSB to reach deterministic equilibrium at 52% of B₀ (MPI, 2014a).

The spawning stock attained the maximum values in the 1980s and decreased in the following decade to levels around 30% of B₀ which were maintained until now. In the last 5 years a slight increase can be noted.
The sub-stock is managed separately since 2008/2009 so catch limits cannot be compared. Last year, the increase of the TAC in the overall QMA 3B was only due to the increase in this sub-stock, that rose from 1.950 tons to 3.100 tons, after the discovery of the newly Rekohu plume (MPI, 2014c). Catches have been decreasing over time in parallel to catch limits’ reduction. Fishing intensity oscillated above the target during the mid-1980s and 1990s, and around the target in the late 1990s; there is a decrease since 2000 to below the target (MPI, 2014a).

All deepwater coral species are protected under Schedule 7A of the Wildlife Act 1953 (MPI, 2014a). All seabirds except Southern black-backed gull Larus dominicanus are protected as well as all marine mammals and reptiles, some sharks (e.g. great white shark Carcharodon carcharias, basking shark Cetorhinus maximus), rays (e.g. giant manta ray Manta birostris) and groupers (e.g. spotted black grouper Epinephelus daemelii). 

Effects of the orange roughy fishery on ETP species (fish, seabirds, sharks, and marine mammals) are reportedly minimal but the direct and indirect impacts on coral are not yet well understood (Punt et al.,  2013). Solenosmilia variabilis, Madrepora oculata, Enallopsammia rostrate and Goniocorella dumosa occur at the trawling depth range (Clark and Anderson 2013) and are being assessed. Coral bycatch has been observed to occur in 10% of trawls in orange roughy fisheries. In northern waters, a higher proportion of tows included coral (MPI, 2014a). Between 2007-2009, observers reported over 50 tonnes of corals being taken in orange roughy target trawls (MPI, 2010).

Boyd (2013), an industry commissioned white paper, considered the  risks of serious or irreversible harm as ‘low to moderate’ on Scleractinian and also gorgonian, golden and black corals (Punt et al., 2013) but some of the assumptions of this ecological risk analysis have been called into question -. the geographical scale of QMA used for the habitat assessment; the definition of ‘habitat’ and habitat elements considered for the structure and function analysis; identification of limited data in some of the ecological categories assumed. An assumption that no more than 25% of underwater topographic features (UTF) are impacted was adopted to determine the impacts of each of the fisheries on the habitat, and the accuracy of this assumption was recommended to be checked after the workshop, but no further news on this could be found. A risk analysis developed by Campbell and Gallagher (2007) for the orange roughy fishery in New Zealand found that the fishery represents a “likely” likelihood, “significant” consequence and “extreme” risk to the “Habitat” and “Protected species (corals)”. The same authors considered the results as merely indicative and as an important basis to recommend further scientific research but note that the study “does not represent a true outcome of a risk analysis” or “the current state of NZ fisheries effects on the environment”.

The fishery interacts with sharks which, depending on life cycle characteristics, may be vulnerable to overfishing. In 2015, a formal quantitative analysis will be available for most important species. There is an effort to properly identify ‘shark species’ caught (MPI, 2014b). The risk of interactions at the species level with commercial fisheries and technical measures to minimize the interaction with sharks and seabirds are summarized in the New Zealand National Plan of Action for the Conservation and Management of Sharks (NPOA-Sharks; MPI, 2013c), in the National Plan of Action to reduce the incidental catch of seabirds in New Zealand Fisheries and in the report New Zealand marine mammals and commercial fisheries (Berkenbusch et al., 2013; MPI, 2013d; MPI, 2014d). The Conservation Services Programme for 2014/2015 outlines goals, timelines and projects to determine the impact of NZ commercial fisheries (DoC, 2014). The interaction with other species will be monitored. Catch of marine mammals and seabirds is not expected to increase taking into account the occasional interaction (MPI, 2014b).

Several mitigation measures are in place to avoid seabird interactions: warp mitigation is mandatory since 2006, Brady bird bafflers, warp deflectors, offal management (MPI, 2012) and Vessel Management Plans (VMP) setting onboard practices (MFA, 2011). There are research projects underway to improve the efficiency of the mandatory measures (MPI, 2014b).

Based on Anderson (2011), a report about ‘fish and invertebrate bycatch and discards in orange roughy and oreo fisheries from 1990–91 until 2008–09’, the MPI (2014a) considers that the target species reached around 84% of the total catch. Bycatch decreased from 27,000 tons to 4,000 tons (2005/2006 and 2008/2009) and non-commercial species are 5-10% of the catch. Discards have also been decreasing from <3,400 tons (1990/1991) to around 300 tons (2007/2008) and are mostly composed of non-QMS species – rattails (Macrouridae Family), birdbeak dogfish (shovelnose dogfish) Deania calcia and chimaeras (Chimaeridae and Rhinochimaeridae). Squids, corals, echinoderms and crustaceans encompass the portion of invertebrates. Bycatch represents 25% of the total catch (MPI, 2014c).

In the 2012/2013 fishing year, the most recent for which data could be found, orange roughy represented 94.36% of the catch in East and South Chatham Rise (based on observers’ onboard reports). The main non-target and QMS species (required to be landed by law) were smooth oreo Pseudocyttus maculatus (1.87%), common mora (ribaldo) Mora moro (0.40%) and black oreo Allocyttus niger (0.36%) which are managed under the QMS. The non-QMS species comprise New Zealand lanternshark Etmopterus baxteri (0.08%), birdbeak dogfish (0.06%) and Sharks & Dogfish nei (Selachii Order) (0.05%) (DWG, 2014b).

The MSC pre-assessment (Punt et al., 2013), based on the ‘Assessment of ecological effects of four New Zealand orange roughy fisheries’ (Boyd, 2013) and other reports, concluded that the risks of serious or irreversible harm to bycatch species identified are ‘low to moderate’. All retained species analyzed are under the QMS, not all are subject to regular stock assessment, but Boyd (2013) considers the risks of serious or irreversible harm as ‘negligible or very low’. The risks of serious or irreversible harm to discarded species are at ‘low to moderate’ (Punt et al., 2013).

Trawling for orange roughy, like trawling for other species, is likely to have effects on benthic community structure and function and there may be consequences for benthic productivity (MPI, 2014a). The vulnerability (slow rate regeneration) of the seabed ecosystems and protected coral species is being studied in result of poor current knowledge (Ramm, 2012). According to Black et al. (2013), there have been no studies investigating whether current trawling frequencies have had adverse effects on the structure and function of benthic communities, or on the productivity of the associated fisheries.

However, according to Baird et al., (2013) the orange roughy fishery was found to be one of the fisheries posing the greatest risk to coral habitats in New Zealand. The annual footprint of the fishery (area of seafloor contacted by trawling operations) has been decreasing and in 2009/2010 was a third of the maximum peak of 1998/1999, however some overlap is still recognized, which poses the risk of biogenic impact (Black et al, 2013).

The MSC pre-assessment raised concerns about the expansion of trawling operations into new areas, the poor knowledge of corals in unfished areas and the lack of monitoring or management of the interaction of the fishery with the benthic ecosystem, especially in closed areas already established (Punt et al., 2013).

Habitats of particular significance for fisheries management (HPSFM) is a policy definition in development regarding the impact of the fishing gears and sensitive coral species (MPI, 2014a). Impacts are minimized if the same towlines are used for fishing (MPI, 2014d) but although this has been a general practice by the fishery it is not a requirement (Punt et al., 2013).

Since 2007, 30% (1,200 Km2) of the Economic Exclusive Zone (EEZ) has been closed to bottom trawling (and dredging) to protect the seafloor, through the implementation of 17 benthic protection areas (BPA) (MPI, 2014a). However, 82% of the protected areas were implemented in waters too deep to fish, while only 3%–15% of marine habitats less than 1500 m deep were protected (Rieser et al. 2013). With respect to the orange roughy species range, only 16% of habitat has been estimated to fall under the protection of BPAs (MPI, 2010). Some portions of the Chatham Rise are recognized to harbor vulnerable marine ecosystems (Rieser et al., 2013). Ten of the seamount closures established in 2001 and twelve of the BPAs are within QMA 3B (15% is closed to bottom trawling), namely the Arrow Plateau area and the ‘Morgue’ seamount (MPI, 2014c).

A network of Marine Protected Areas (MPA) is established and includes 38 marine reserves (territorial waters) and 6 marine mammal sanctuaries besides other levels of protection (e.g. marine parks, closures, protection zones). The MPA policy was published in 2006 and the first phase concerned the territorial sea where impacts were more important and considering that BPAs were already providing protection to seabed ecosystems within the EEZ. Implementation of offshore MPAs was postponed to 2013 until when all protected areas and levels, readjustments and proposals were being analyzed (MPI, 2009).

According to Leathwick et al. (2008), the current system of marine protected areas is not very representative and does not provides protection for a large number of fish species.

The current BPAs were to be reviewed after 2013 and if existing BPAs were found not to be representative then further closures were to be considered (MPI, 2010).