Citation
Summary
The JMA 3 Quota Management Area (QMA) includes the southern part of the New Zealand EEZ. Unlike the JMA 1 and JMA 7 QMAs, that had established target fisheries, JMA 3 appears to have been established to include the small jack mackerel catches that occurred around southern New Zealand. These catches were initially bycatch in a range of middle depths and inshore trawl fisheries, together with target purse seine catch taken off the Kaikoura coast.
A target midwater trawl fishery for jack mackerels in JMA 3 became established in the late 1990s, primarily in response to the arrival of Chilean jack mackerel (Trachurus murphyi) in New Zealand waters. T. murphyi was first identified from eastern and southern parts of JMA 3 and subsequently spread west into JMA 7 and north into JMA 1. While it is assumed that some T. murphyi, initially present in JMA 3, subsequently emigrated to the other QMAs, sufficient remained that T. murphyi was the key species taken in the JMA 3 fisheries until the early 2020s.
Observer sampling of the JMA 3 fishery has been much sparser than in the JMA 7 fishery. As a result, it was necessary to develop species composition modelling to estimate the catch by species. The modelling assumed that yellowtail jack mackerel (T. novaezelandiae) was largely absent from JMA 3, at least in the areas of JMA 3 accessible to the midwater fleet, and therefore focussed on separating the catches into T. murphyi, and greenback jack mackerel, T. declivis.
The species composition modelling used Bayesian generalised linear mixed models. Temporal and spatial variables explained much of the variation in species proportions, with much of the remaining variation attributable to a random effect for the observer trip that collected the data. It is unknown whether this reflects small scale spatial variation in catch composition, or may be related to observers misidentifying the jack mackerel species. However, the consequence is that there is considerable variation in the predicted catches by species. A separate model was developed to characterise the predictive skill of the species composition modelling, and this model was used to assign weights to individual records used in catch per unit effort (CPUE) standardisation.
The CPUE models suggested differing trends in abundance of the two species in JMA 3, with T. murphyi abundance increasing rapidly in the 1990s, but then declining rapidly to a low level during the 2000s, with stable abundance since around 2010. Abundance of T. declivis also declined in the 2000s, but has trended upwards from the mid 2000s to 2024. However, the scale of the change, and considerable interannual variability, suggested that the CPUE for T. declivis was unlikely to represent abundance, at least of an independent JMD 3 stock.
Size data from observer sampling, together with a single year of catch at age sampling in 2013, suggested that T. declivis in JMA 3 comprised larger, and older fish, that had originated in another area. For stock assessment purposes, it was assumed that T. declivis in JMA 3 (JMD 3) were part of the JMD 7 stock. On the basis of the observer-data based CPUE index for T. declivis in JMA 7, the status of the stock in both areas was assessed as Likely to be at or above the target (the geometric mean JMD 7 CPUE for the period 2004–2011). Catches of T. declivis in both JMA 3 and JMA 7 were used in assessing relative exploitation rate for the combined stock; this had been below the overfishing threshold since 2011.
More extensive species composition sampling for jack mackerels in JMA 3, both from target tows and when jack mackerels are taken as bycatch in other midwater trawl fishing, is required to better monitor the JMA 3 fishery going forward.