The waters surrounding Japan boast exceptionally high species diversity, even on a global scale (Fujikura et al. 2010). Various fishing methods have been employed along Japan’s coast since ancient times. These methods, including set nets, seine nets, trawl nets, and gillnets, capture a diverse array of fish species. In coastal fisheries, the primary target species for catch are those with high commercial value and large landings. However, due to the wide variety of species targeted for catch, unintentional bycatch of other species is common.

Bycatch refers to catches that are not specifically targeted, utilized, or managed catches. It is reported to constitute 40.4% of the global ocean catch (Hall 1996; Davies et al. 2009). This unintentional bycatch threatens the stock status of the affected fish species. The lack of information for stock assessment poses a significant challenge (Hall 1996). Accumulating detailed information on biological characteristics, in addition to species, size, and weight, is essential to consider appropriate bycatch management.

However, studies on bycatch of non-target fish species (i.e., often underutilized), excluding sea turtles, marine mammals, and seabirds, usually provide broad, country-specific, or fishery-specific examples (Alverson et al. 1994; Kelleher 2005; Finkbeiner et al. 2011; Oliver et al. 2015). These studies lack biological data (Kindong et al. 2020; Armelloni et al. 2021; Heimann et al. 2023). Information on underutilized fish is seldom recorded during fish surveys (Stohs and Harmon 2022), hindering assessments of their distribution, abundance, and population structure. Fish surveys often record bycatch of non-fish vertebrates and large crustaceans, which can result in undesirable mortality of both target and non-target species.

Length–weight relations (LWR) are crucial ecological data for understanding underutilized fish species. LWR provides valuable insights into fish population dynamics, aiding in stock assessments, size limit determinations, and fishing quota settings (Rodríguez-García et al. 2023). Understanding the LWR of each fish species enables tracking changes in body mass index and length composition, estimating growth patterns, and body condition of sampled fish specimens (Shubhadeep et al. 2022). Establishing LWRs for newly recorded fish species is vital for estimating biological weights and observing changes over time (Kodeeswaran et al. 2020). Thus, summarizing LWRs of bycatch species is crucial for enhancing our comprehension of fish populations and providing valuable insights for future research and management endeavors.

The set-net fishery is one of Japan’s primary fisheries, contributing approximately 40% of the total production in Japan’s coastal fisheries (Akiyama 2007; Anonymous 2024). It represents a passive fishing method, leading to the capture of various species due to low selectivity yet raising concerns about bycatch (Cheng and Chen 1997; Akiyama 2007). This study aimed to estimate the length–weight relations (LWR) of underutilized bycatch fish species collected from set nets in Nishiki, Oki-cho, Mie Prefecture, facing the Pacific Ocean.

The study was conducted on large set nets operating in the northern coastal area of the Kumano-nada Sea. This set net was positioned off the coast of Nishiki, Taiki-cho, Watarai-gun, Mie Prefecture (34°11′38′′N, 136°24′31′′E) (Fig. 1) and comprised a total length of 230 m with an installed depth of 50 m. The set net consisted of two final chambers with the top covered. Surveys were conducted 26 times during the 2018 fishing season (28 Nov. 2018–17 Jun. 2019), 94 times during the 2019 fishing season (2 Dec. 2019–21 Jul. 2020), and 131 times during the 2020 fishing season (2 Dec. 2020–18 Jun. 2021), totaling 251 operations.

During the study, researchers boarded a set-net fishing boat to collect underutilized bycatch species that the fishermen sorted after the catch was hauled on board. The remainder of the catch was landed at Nishiki Port (34°12′54′′N, 136°23′40′′E). The landed catch was further sorted on a sorting table, and underutilized fish species that were discarded in the process were also collected. The collected unutilized bycatch species were measured for length (fork length) and wet weight. Fork length was measured in 1 cm increments, weight less than 1 kg in 1 g increments, and weight greater than 1 kg in 10 g increments. Species identification was based on standard literature (Nakabo 2013).

In the analysis, simple regression analysis was performed on the logarithm of fork length (L) and weight (W) (log L, log W) for the bony fish species that could be discriminated to species. The calculated Cook’s distance was rounded to the second decimal place, and values greater than 1 were excluded as outliers (Cook and Weisberg 1982). After outliers were excluded, species with ten or more observations were considered for analysis. For each analyzed fish species, the relation between L and W was applied to the allometric equation

W = aL^b

to calculate the initial growth coefficient a, relative growth coefficient b, and determination coefficient R². The statistical analyses were performed with R 4.2.2 (R Core Team 2022).

A large-scale language models, Claude 3 (https://www.anthropic.com/claude), were used for the English review of this manuscript.

The survey revealed that a total of 679.4 tonnes of total catch weight was landed during 251 operations during the 2018–2020 fishing season. A total of 208 species with a total weight of 5.9 tonnes of underutilized bycatch species were recovered, representing 1% of the total catch weight landed. Bony fish species comprised 152 species and 1227 kg, or 31% of the total. The analyzed bony fishes encompassed 14 species from 13 orders and 14 families. Five of these species were removed from the final results because three species of them had either significantly low R² values such as Abudefduf vaigiensis (Quoy et Gaimard, 1825); Lagocephalus sceleratus (Gmelin, 1789); and Pempheris schwenkii Bleeker, 1855 or were deviated significantly from the a or b values of existing reports in FishBase (Froese and Pauly 2024), such as Pterocaesio trilineata Carpenter, 1987, and Spratelloides gracilis (Temminck et Schlegel, 1846). Commonly exploited species like Engraulis japonicus Temminck et Schlegel, 1846 and Sardinops melanosticta (Temminck et Schlegel, 1846) were also caught as bycatch at the study site but were excluded from the analysis in this study. The presently reported study focused on a total of 909 fish individuals representing the following nine fish species: Hyperoglyphe japonica (Döderlein, 1884); Labracoglossa argenteiventris Peters, 1866; Macroramphosus sagifue Jordan et Starks, 1902; Microcanthus strigatus (Cuvier, 1831); Ostorhinchus semilineatus (Temminck et Schlegel, 1843); Parapriacanthus ransonneti Steindachner, 1870; Rexea prometheoides (Bleeker, 1856); Siganus fuscescens (Houttuyn, 1782); Upeneus japonicus (Houttuyn, 1782). Those fishes belonged to eight orders and nine families (Table 1).

The relation between weight (W) and fork length (L) was plotted for 9 species (Fig. 2), revealing distinct patterns for each fish species. The number of samples, along with length, weight, and estimated values of a, b, and R² for each fish species, are presented in Table 1. The a values ranged from 0.003 (Rexea prometheoides) to 0.019 (Parapriacanthus ransonneti; Upeneus japonicus), while the b values ranged from 2.862 (Macroramphosus sagifue) to 3.311 (Hyperoglyphe japonica). High coefficients of determination (R² values) ranging from 0.806 (Parapriacanthus ransonneti) to 0.994 (Hyperoglyphe japonica) were obtained for all species.

The survey results indicated that underutilized fish accounted for only 1% of the total catch weight, a significantly low value compared to other regions (Akiyama 2007). Among the estimated length–weight relations (LWR) of fish species reported in this study, five species (Hyperoglyphe japonica, Labracoglossa argenteiventris, Microcanthus strigatus, Parapriacanthus ransonneti, and Rexea prometheoides) were not listed or credibly described in FishBase (Froese and Pauly 2024). Additionally, four novel species (Macroramphosus sagifue, Ostorhinchus semilineatus, Siganus fuscescens, and Upeneus japonicus) for which no credible LWRs had been reported for the Japanese coast were newly reported. Therefore, this study’s major significance lies in providing new biological information on these fish species.

Nine of the estimated LWR equations demonstrated high accuracy, with coefficient of determination (R²) greater than 0.8 and b values within a realistic range of 2.8623.311 (Froese 2006). On the other hand, all specimens used for Hyperoglyphe japonica considered to be juveniles (Kawano and Shigenaga 2011) and further information is required. Therefore, improving the accuracy of these parameters requires ensuring sufficient populations over a broader size range (Froese et al. 2011).

Alverson et al. (1994) classified discard problems in the northwest Atlantic were classified into four groups:

Regulatory approaches and management actions to address these problems are also discussed. In Japanese coastal fisheries, juvenile individuals of the target fish, belonging to category 1 in the Alverson et al. (1994) classification, are often caught as bycatch, but biological information is limited if the species themselves are not commercially utilized. None of the 9 species discussed in this study were utilized along the Mie Prefecture coast, and no biological information was available for any of them. Therefore, identifying LWR in this study provides fundamental and important information for assessing these fish species’ stock status and future stock management. However, for many underutilized fish species in coastal waters, the actual bycatch and even the existence of biological data are unknown (Alverson et al. 1994). Considering that enhancing such primary data is essential for proper resource management and biodiversity conservation (Pauly et al. 2013), bycatch assessments in various fishing modes in different regions should be pursued, along with the comprehensive collection and accumulation of biological information on underutilized fish species in the future (Zeller et al. 2017).