This project used simulation modelling to explore potential bias in snapper SNA 8 stock assessments when there has been a change in stock productivity (i.e., regime-shift), as may be expected under climate change.

Three SNA 8 stock productivity-change scenarios were investigated: the first assumed an upward shift in productivity after 2000; the second assumed a downward productivity shift after 2000; the third had no productivity shift.

Various SNA 8 stock assessment models were run under these productivity shift scenarios including one that explicitly allowed for a post-2000 productivity shift. Assessment bias was investigated specific to two important management metrics: current-stock-biomass; current-stock-status (being the ratio of current-stock-biomass to stock virgin (unexploited) biomass).

All assessment models produced unbiased estimates of current-stock-biomass under the no-regime-shift scenario. Only the post-2000 productivity shift model produced unbiased current-stock-biomass estimates under increasing and decreasing productivity scenarios.

All model current-stock-status estimates were biased under increasing and decreasing productivity scenarios. Although the post-2000 productivity shift model current-stock-status estimates were markedly less biased that those of the other models. An important finding from the study was all models were substantively less biased in their estimates of current-stock-biomass than current-stock-status.

An important conclusion from the simulation work was that we should not be using model predicted stock-status ratios as stock assessment measures when it is suspected that stock productivity is likely to have changed. Instead, we should be placing more ‘faith’ in assessment model predictions of current-stock-biomass and therefore be measuring sustainability solely against these estimates.

This report presents results of the 2022 inshore trawl survey of the west coast North Island (WCNI), the 9th in a time series starting in 1989, but with a 19-year gap between 1999 and 2018 surveys.

The survey extends from Scott Point on Ninety Mile Beach to Mana Island covering a depth range from 10–200 m north of Cape Egmont and 10–100 m to the south. Since 2018, there has been no sampling within 2–4 nautical miles of the coast between Maunganui Bluff and the Waiwhakaiho River, New Plymouth, a no-trawl area established to protect the Māui dolphin.

Everything that is caught in the trawl is sorted, identified, and weighed, and length and maturity data are collected for selected species and otoliths (fish ear stones) for ageing the four main species of interest: snapper, red gurnard, John dory, and tarakihi. The trawl survey provides time series of relative biomass estimates and age, length, and maturity stage information used for stock assessments and fisheries management advice for key inshore species.

In 2023, 95 phase one stations were successfully completed followed by four phase two stations completed to improve the coefficient of variation for tarakihi.

There were 72 species recorded in total, with snapper by far the most abundant. Biomass estimates (in tonnes) for the key species across the whole survey were: snapper, 8396.3 t ; red gurnard, 1160.5 t; tarakihi, 447.6 t ; John dory, 305.3 t.

The 2022 snapper biomass estimate was lower than that from the 2019 and 2020 surveys, but still significantly higher than the historical surveys in the 1990s. There were high numbers of juvenile snapper under 5 years old but a lower number of adult fish. The variability in adult snapper biomass in this survey may be due to fish moving inshore into shallow water to spawn, or may reflect the highly patchy distribution of snapper at this time of year.

The biomass estimate for adult red gurnard was relatively stable, but the biomass of smaller fish was substantially below historical estimates.

Snapper are the most important recreational fish species in New Zealand and are often released back to the sea after capture.
Little is known about the survival of fish after they are released.
NIWA conducted a study using volunteer fishers to catch 960 snapper at different depths and with different hook placements.
The captured snapper were kept in holding nets and monitored by NIWA divers over several days.
Fish hooked in the lip had a low chance of dying if caught at shallow depths, but the chance of dying increased as depth increased.
Fish hooked elsewhere on the body had a higher chance of dying, with those hooked deep in the gut having the highest chance of dying.
This study suggests that fishing practices can impact fish survival, but there are ways to potentially reduce post-release mortality.
Understanding how fishing affects fish survival is therefore an important consideration for catch and release fisheries and when setting catch limit regulations.

This study estimated the post-release survival of inshore finfish with current commercial minimum legal sizes—blue cod, blue moki, butterfish, kingfish, red moki, red cod, sand flounder, snapper, tarakihi, trevally, and yellowbelly flounder—and those currently allowed to be returned under disposal code X— kingfish, rig, sand flounder, school shark, rough skate, smooth skate, and spiny dogfish.

A questionnaire was developed and circulated to fishers, fishery observers, and scientists with knowledge of each species to obtain their estimates of at-release survival (i.e., the probability the fish/shark was alive when put back into water) and post-release survival (the probability an individual was both alive at release and survived following release). Estimates were obtained for each gear type as well as fishing categories within each gear type, e.g., duration, depth, and bag size. For some species, estimates of post- release survival were informed by literature on the survival of same or similar species.

These data were used with fishery characterisations to model the survival for each species. For species with a minimum legal size, both at-release and post-release survival estimates were used, whereas for those species released under disposal code X, which may only be released if alive and likely to survive, only the post-release survival estimates were used.

The post-release survival from longlines for snapper, kingfish, and rough skate (assuming they are released alive) was “medium-high”, i.e., the lower bootstrapped 90% confidence range was lower than 0.50, but greater than 0.25, and upper 90% confidence range exceeded 0.75. The same survival range was estimated for snapper caught in pots. For snapper, this result was based on expert knowledge and incorporated literature-based mean values based on empirical studies for this species in New Zealand.

However, for rough skate, the result was based on the informed opinion of 2 science experts only (at-vessel survival was assumed to be 100%) and without the benefit of literature-based empirical estimates as none exist for this species. If at-release survival estimates are included for kingfish, the range decreases to “medium”. For blue cod, and other sharks, skates, and rays, survival from capture on bottom longline was “medium-low”. Red cod survival was “low”, and an “uncertain” outcome was applied to smooth skate and blue moki due to the lack of available knowledge. For species such as spiny dogfish and school shark, the lower range of perceived survival was at least partially a result of the wide range of depths where these species are discarded, which includes deepwater fisheries with larger vessels and potentially different handling practices.

Post-release survival of most species from trawl gear was perceived to be “medium-low” at best, with 90% confidence range either spanning 0.25 up to but less than 0.75, or else between 0.25 and 0.5. Blue cod, red cod, flatfish, and tarakihi were considered in the “low” range, where the 90% confidence range did not exceed 0.25. Survival of both rough and smooth skates in trawl gear was “uncertain”, based on the lack of empirical data for these species and the wide range of estimates for related species in overseas fisheries. Where bottom trawl with a Modular Harvest System cod-end was considered as a separate gear, the lack of available data on this gear type meant survival was considered “uncertain”. For set net, the perceived survival of all species where this gear was considered an important method, was “medium-low”.

The species-method survival confidence ranges presented in this report are based on the best currently available expert knowledge and thorough reviews of the current survival literature; as such, these ranges are unlikely to be improved upon without further investment in release survival research.

This report presents the results from the 16th inshore trawl survey in a time series started in 1992 along the west coast of the South Island, from Farewell Spit to the Haast River mouth, and in Tasman Bay and Golden Bay.

The survey covers depths from 20 to 400 m (core strata) and surveys many species but is mainly focused on giant stargazer, red cod, red gurnard, spiny dogfish, and tarakihi. Since 2017, two additional strata have been surveyed in 10–20 m in Tasman Bay and Golden Bay to cover the full distribution of snapper in the geographic area.

Data collected include length, weight, and maturity data for selected species, and collection of otoliths (fish ear stones) of the key species for ageing. The trawl survey provides time series of relative biomass estimates and age, length, and maturity stage information used for stock assessments and fisheries management advice for key inshore species.

In 2023, 58 phase one stations were successfully completed in the core strata and another six were carried out in strata 20 and 21. Four phase two stations were completed to reduce the coefficient of variation for spiny dogfish and snapper.

Biomass estimates (in tonnes) for the target species in the core strata were: giant stargazer, 915 t; red gurnard, 1498 t; red cod, 69 t; snapper, 3633 t; spiny dogfish, 3043 t; and tarakihi, 493 t.

The snapper biomass (core strata plus the 10–20 m strata) was the highest ever in the time series and nearly triple that from the previous survey in 2021, with most fish 20 years or younger. Juvenile snapper were caught mostly in the 10–20 m strata. These strata provide important information on future recruitment and contain a variable proportion of the adult population.

A catchability analysis of the survey indicates that the survey can be considered representative of the time series.

A spatial analysis of catch and effort data from the SNA 1 bottom trawl fishery was used to refine the boundary of the East Northland and Hauraki Gulf snapper stocks. For the two stock units, bottom trawl CPUE indices were derived for the spring-summer period. The two sets of CPUE indices were included in the 2023 stock assessment of SNA 1.

Three stratified random trawl surveys completed between 2018–2020 off the west coast North Island using RV Kaharoa provided a comparison with historical surveys last carried out in 1999. The time series reveals a substantial increase in snapper biomass has occurred since 1999, but a decline in pre-recruit red gurnard was apparent. John dory biomass also trended downwards. CVs for these species were all within target. For tarakihi, CVs were higher and trends were less clear.

Random age frequency sampling in 2019–20 gave catch-at-age estimates for SNA 1 longline and SNA 2 North trawl fisheries and a length frequency and age-length key approach allowed estimates of catch-at-length and -age for SNA 1 catches. Mostly, target sample sizes for numbers of landings were achieved and spatio-temporal comparisons revealed good sample representativeness. Catch-at-age distributions remain relatively broad with good representation across all recruited age classes up to 20 years.

This report details the results of a bottom trawl survey of the west coast South Island in 20–400 m depth and in Tasman Bay and Golden Bay in 10–70 m depth carried out using RV Kaharoa (KAH2103) in March–April 2021. This was the fifteenth survey since 1992 optimised to estimate abundance of red gurnard, red cod, giant stargazer, spiny dogfish, and tarakihi in core strata (20–400 m) and the third survey since 2017 optimised to estimate snapper only in 10–400 m.

The estimated year class strength (YCS) of 1+ and 2+ snapper from a November 2020 (Hauraki Gulf) and February 2021 (Bay of Plenty) trawl survey were all well above the long-term average. Coefficients of variation (CV) for pre-recruit snapper were reasonable, except for 1+ snapper in the Bay of Plenty (CV = 38.5%). Surveys used a stratified two-phase design optimised for pre-recruit snapper. These surveys may also have potential to monitor abundance of recruit snapper and other inshore species.