Striped marlin (Kajikia audax) supports a popular and valuable recreational fishery in New Zealand. They are occasionally caught by the surface-longline fleet, but retention of commercial catches has been prohibited since 1987. Striped marlin prefer warm waters between 20◦C and 24◦C and primarily occupy the mixed layer near the surface. They are found in New Zealand during austral summer months especially, with large, mature individuals moving from spawning grounds in the Coral Sea to feed in New Zealand and neighbouring waters.

This study assessed bycatch of striped marlin by commercial surface longliners in New Zealand waters. The characterisation focused on data for fishing years from 2003–04 to 2018–19, with striped marlin bycatch records prior to this period considered to be less reliable. In addition to the characterisation, a bycatch prediction model was also developed to assess the influence of environmental and operational covariates on striped marlin bycatch rates.

Overall, striped marlin bycatch levels were stable from 2003–04 to 2018–19, with some year-to-year variability. Both striped marlin captures and catch-per-unit-effort declined in the last three years of the assessment period (2016–17 to 2018–19), following a peak in 2015–16. The three key surface-longline target species with striped marlin bycatch were bigeye tuna (Thunnus obesus), swordfish (Xiphias gladius), and southern bluefin tuna (Thunnus maccoyii). Striped marlin bycatch was common in sets targeting bigeye tuna and swordfish, with occasional bycatch in sets targeting southern bluefin tuna.

Trends in striped marlin bycatch have mirrored changes in the surface-longline fishery over time. Early in the reporting period, striped marlin bycatch mostly occurred in surface-longline sets targeting bigeye tuna. However, fishing effort targeting swordfish has increased since 2003–04 and striped marlin bycatch rates for those sets are high compared with other target species. As such, about half of the recent striped marlin bycatch was in sets targeting swordfish. In contrast, although fishing effort targeting southern bluefin tuna has also steadily increased since 2003–04, this increase did not result in increased striped marlin bycatch levels; striped marlin captures rates for this fleet were consistently low.

There were distinct spatial patterns in the distribution of striped marlin bycatch in New Zealand: most capture events occurred in North Island waters, from Bay of Plenty to Northland, and there were few captures (and low catch-per-unit-effort, CPUE) on the east coast south of Hawke Bay. This pattern appeared to correspond with patterns of spatial occupancy and not catchability given similar fishing fleets were active in both of these areas and over the same seasons. In parallel, increases in sea surface temperatures may have impacted the distribution of striped marlin in New Zealand waters as the boundary of striped marlin bycatch has expanded southward along the South Island west coast.

Striped marlin’s preference for warm waters was reflected in seasonal trends in the bycatch records. Capture rates were particularly high in warm months between January and March. This period coincided with the fishing season for swordfish, as well as bigeye tuna to some extent. For these two target species, striped marlin captures were highest in February and March. In contrast, for southern bluefin tuna, striped marlin bycatch was highest in the winter months between June and August, when effort targeting this species shifted from the North Island southern east coast towards Bay of Plenty and Northland.

The bycatch prediction model confirmed a distinct signal in striped marlin capture rates across target species, with the probability of striped marlin bycatch highest in swordfish-targeted sets. However, the operational covariates considered in this analysis did not explain a high proportion of variation in the capture rates across sets. Instead, oceanographic covariates, mostly sea surface temperature (SST) and also surface water chorophyll-a, had the strongest predicted influence on bycatch rates. Moon illumination was also predicted to impact capture rates, with a predicted increase in the probability of bycatch for sets during periods approaching the full moon. Overall the proportion of bycatch explained by covariates was low (about 20%), which may indicate high variability in striped marlin capture rates that is due to chance, or the importance of other significant covariates that were not considered in this analysis.

When the dataset for the prediction model was extended to examine the effect of the El Niño Southern Oscillation (ENSO) on striped marlin bycatch and CPUE, two prominent trends were identified. First, there was a distinct increase in bycatch and CPUE during or immediately following the two strong El Niño events of 1998 and 2015; however, there was no consistent pattern in catch or CPUE in the period in-between these two years when weaker El Niño or La Niña events occurred. Second, there was a gradual decline in CPUE over three years following the peaks associated with the two strong El Niño events in 1998 and 2015. It is unclear whether these patterns were due to changes in local abundance, catchability, or recruitment of striped marlin, or a combination of these factors. Given the main covariate explaining striped marlin bycatch in New Zealand waters was SST, both ENSO and long-term climate change are expected to influence bycatch rates and the distribution of captures of this species in the medium- to long-term.