Age and growth of chum salmon from the Tumen River

Sample collection

Study specimens were broodstock from stock enhancement programs, captured near Fangchuan Village (located in lower Tumen River, Fig. 1), using three-layer drift gill nets (mesh size: 12.67 cm). Annual sampling was conducted in mid-October (2020–2023). Following capture, the live fish were transported to a fish hatchery and stocking center for temporary holding. During artificial propagation, biological measurements were performed, and 6–10 dorsal scales above the lateral line were collected and transported to the laboratory for examination and analysis. A total of 331 chum salmon were sampled, including 241 females and 90 males. All sampling procedures strictly adhered to the guidelines provided by Heilongjiang River Fisheries Research Institute of CAFS for Laboratory Animal Welfare and Ethical Review (No.: 20200920-001).

Materials handling and data analysis

Age determination

A professional researcher counted annuli on all collected scales. For each fish, the best-preserved scales with clear annuli were selected for analysis due to observed resorption in some samples. Age determination was initially based on the “calendar year method” (Seo et al., 2006), where one year was added to the annulus count (e.g., 2⁺ corresponding to 3 years old). However, the distance from the scale margin to the last annulus at capture primarily represented the pre-spawning spring-summer period, which did not constitute a full year. To improve accuracy, the “+” value was adjusted to the annual ring value plus 0.7, based on observations that chum salmon hatched from January to February and reached maturity around October (Wang, Tang & Liu, 2012; Wang et al., 2013). Consequently, 2⁺ and 3⁺ fish were recorded as 2.7 and 3.7 years old in this study, respectively.

Age structure

The age composition of Tumen River chum salmon population consisted of four groups: 2⁺, 3⁺, 4⁺, and 5⁺ (Table 1), with individuals of 2⁺ and 3⁺ age groups accounting for 96.37% of the total. Females were dominated by the 3⁺ age group (51.86% of females), exhibiting a mean age of 3.31 years and a T_50% of 2.00 years. Males were predominantly 2⁺ years old (56.67% of males), with a mean age of 3.14 years and a T_50% of 1.87 years. Comparative analysis with the Ussuri River and Suifen River populations (Table 2) showed that both the overall mean age and the female mean age of the Tumen River poupulation were lower than those of the other two populations. Conversely, the mean age of Tumen River males was lower than that of Ussuri River males but higher than that of Suifen River males. Additionally, the T_50% values for both sexes within the Tumen River population were lower than those within the Suifen River population. Across all populations, females consistently showed higher mean age and T_50% values than males. These results demonstrate distinct age differences among populations and between sexes.

Fork length and body weight composition

The FL of Tumen River chum salmon ranged from 51.0 to 74.6 cm, with a mean of 61.8 ± 4.28 cm. The BW ranged from 1,400 to 5,200 g, with a mean of 2,516.95 ± 572.35 g. Most individuals (72.21%) measured 55 to 65 cm in FL. Females exhibited FL from 53.9 to 72.2 cm (mean 62.4 ± 3.74 cm) and BW from 1,600 to 4,020 g (mean 2,561.08 ± 518.14 g), with 74.69% falling within 55 to 65 cm in FL. Males showed FL from 51.0 to 74.6 cm (mean 60.4 ± 5.23 cm) and BW from 1,400 to 5,200 g (mean 2,398.78 ± 686.50 g), with 65.56% in the 55 to 65 cm FL range (Fig. 2, Table 3). Overall, these results demonstrated that females exhibit larger FL and BW than males.

Growth analysis

The ANCOVA results revealed no significant difference in the FL-BW relationship between sexes of Tumen River chum salmon (n = 331, P = 0.905 > 0.05). Therefore, the pooled data yielded the power function: BW = 0.0159×FL^2.8969, where parameter b showed no significant different from 3 (t = 0.0336; Pauly, 1984) < t_{(0.05, 329)}, indicating isometric growth in chum salmon. Consequently, the VBGF was applied for modeling.

To investigate growth patterns across age groups, we fitted back-calculated FL (derived from scale radius) to age (Fig. 3). Figure 3 shows that the mean FLs at 2.7, 3.7, and 4.7 years were generally higher than those at 3, 4, and 5 years, suggesting distinct growth patterns among age groups (Wang, Liu & Tang, 2013). Due to limited samples in the 4⁺ and 5⁺ age groups, our analysis focused primarily on 2⁺ and 3⁺ age groups. Females exhibited greater FL and BW values across both age groups. Taking the 3⁺ age group as an example, ARSS analysis showed no significant growth difference between females and males (F = 0.3572 < F_(0.01,2,5) = 13.2739). Consequently, sexes within the same age group were combined for analysis. The VBGF parameters (Table 4) and growth curves (Fig. 4) demonstrated that the 2⁺ age group grew faster than the 3⁺ age group, with FL growth rates declining with increasing age. ARSS further confirmed significant growth differences between the 2⁺ and 3⁺ groups. (F = 33.50 > F_(0.01,1,4) = 21.198).

Fork length at first maturity

Based on the logistic formula (Fig. 5), the formulas for estimating the FL_50% of female and male Tumen River chum salmon were derived as follows:

For females: $P_i=\frac{1}{1+e^{-\left(-27.708+0.604l_i\right)}}$

For males: $P_i=\frac{1}{1+e^{-\left(-15.332+0.349l_i\right)}}$.

The calculated FL_50% for female and male were 45.87 cm and 43.93 cm, respectively, indicating that males attained sexual maturity earlier than females. ARSS analysis was applied to test the significance of differences in FL_50% between sexes. The results demonstrated a highly significant divergence in sexual maturation probabilities between sexes (F = 14.385 > F_(0.01,14,16) = 3.451).

Age structure analysis

The determination of age is critical for sustainable fisheries management and evaluating the recovery of endangered populations. Sexual maturation age directly influences fish fecundity and longevity, making it a fundamental biological metric widely studied by researchers. Studies on chum salmon populations have revealed temporal fluctuations in their sexual maturation ages. For example, Helle & Hoffman (1995) documented an aging trend in sexual maturity among North American chum salmon populations, a pattern that has also been observed in Asian populations (Kaeriyama, 1996; Kaeriyama & Edpalina, 2008; Seo, Kudo & Kaeriyama, 2009). In contrast to the aforementioned conclusions, recent studies have documented that spawning populations of chum salmon returning to China’s Ussuri River, Suifen River, and Heilongjiang River exhibit a trend toward younger sexual maturation ages (i.e., earlier maturation timing) (Wang et al., 2013; Wang, Liu & Tang, 2013; Wang et al., 2020).

The Tumen River chum salmon population, as one of the four major geographic populations within China (Chen et al., 2005), has previously been mainly studied in terms of fecundity (Li et al., 2021), population conservation (Kang & Wan, 2016; Wang et al., 2023), and genetic structure (Han et al., 2020; Yan et al., 2021). The absence of historical age data has led this study to focus solely on the current age composition of the population. Based on the fact that individuals aged 2⁺ and 3⁺ account for 96.37% of the total, this indicates that the current age structure is dominated by young individuals. Furthermore, the Ussuri River and Suifen River populations, used for comparing maturation age with the Tumen River population in this study, had their average maturation ages reported by Wang et al. (2013) and Wang et al. (2020) as historic lows during the study period. Therefore, the lower mean age of the Tumen River population and the lower T_50% values of both its female and male groups indicate that the current maturation age exhibits a young profile.

The interaction between the genetically influenced sexual maturation rate and abundance differences across adjacent generations jointly drives shifts in the age structure of spawning populations (Kaev, 2000). That is to say, the young age structure of the Tumen River chum salmon population is attributed not only to the high abundance of low-age fish returning to spawn but also potentially to the earlier sexual maturation of faster-growing individuals. This is corroborated by the number of returning groups at age 2⁺ and 3⁺, as well as the earlier maturation timing and faster growth rate of the 2⁺ age group compared to the 3⁺ age group observed in this study. Beyond genetic influences, anthropogenic factors such as overfishing may contribute to earlier maturation (Fukuwaka & Morita, 2008; Hard et al., 2008). Historical records show that the Tumen River chum salmon catch declined from over 100,000 individuals in the early 20th century to less than 1/10 of the 1940s level in recent years (Kang & Wan, 2016; Li et al., 2021). Overexploitation likely exerts selective pressure, prompting populations to mature earlier to ensure reproductive continuity.

Growth characterization

The Tumen River chum salmon population exhibited BW of 1,400–5,200 g and FL of 51.0–74.6 cm. Individuals with FL of 60–65 cm accounted for the largest proportion (45.52%), while those <60 cm and >65 cm constituted 31.72% and 22.66%, respectively. This distribution indicates a predominance of small-sized individuals in the population. The FL-BW relationship was described as BW=0.0159 × FL^2.8969, where the regression coefficient b showed no significant difference from 3, indicating isometric growth and suitability for simulation with VBGF, consistent with Ussuri and Suifen River populations (Wang et al., 2013; Wang et al., 2020). ARSS analysis revealed significant growth differences between the 2⁺ and 3⁺ age groups. The 2⁺ group displayed higher growth coefficient (k) and performance index (φ) values (Table 4), corresponding to faster FL growth rates and elevated growth curves (Fig. 4). This confirms that earlier-maturing individuals grow faster, with maturation age inversely related to FL growth rate, consistent with Morita et al. (2005). Similar to most fish, rapid growth in younger individuals enhances population survival (Duffy & Beauchamp, 2011). Earlier-maturing individuals exhibit greater environmental adaptability, but increased abundances intensify competition for food resources, suppressing growth in older individuals and progressively slowing their growth rates. Using φ, we further compared chum salmon growth rates among populations in the 2⁺ and 3⁺ age groups (Table 4). The results indicate that the Tumen River population exhibits an intermediate rate—faster than the Suifen River population but slower than the Ussuri River population.

Sexual maturity in fish is influenced by both age and body size. Given FL variability at maturity, we used FL_50% to represent Tumen River chum salmon maturation size, reflecting population growth and fecundity to prepare for assessing of reproductive potential and fisheries management. The FL_50% values were 45.87 cm for females and 43.93 cm for males, and the corresponding ages (T_50%) were 2.00 and 1.87 years, respectively, demonstrating earlier male maturation at smaller sizes. Females’ larger size at maturity aligns with natural selection pressures, as their reproductive success strongly correlates with body size (Wang et al., 2020; Li et al., 2021), while prolonged energy allocation for gonadal development further necessitates delayed maturation. Comparatively, Ussuri River populations exhibit FL_50% values of 48.36 cm (females) and 46.60 cm (males), while Suifen River populations show 51.53 cm (females) and 42.15 cm (males) (Wang et al., 2013; Wang et al., 2020). Tumen River females display smaller FL_50% values than both populations, whereas males exhibit marginally larger FL_50% than Suifen River males but are close in value. This indicates that the current Tumen River population exhibits characteristics of small body size at sexual maturity. Chum salmon exhibit high growth plasticity due to genetic and environmental influences. As an anadromous species, they demonstrate remarkable homing precision to natal spawning grounds, resulting in significant morphological and ecological divergence among geographic populations (Saito, 2017; Saito et al., 2020). After hatching, fry spend minimal time in freshwater before migrating to marine environments where they undergo most growth and development. Consequently, their body size is highly sensitive to marine temperature (Farley & Moss, 2009; Kasugai et al., 2012), climate change (Kitada & Kishino, 2019; Kuroda et al., 2020), and interspecific/intraspecific competition (Ishida et al., 1993; Helle & Hoffman, 1998; Bigler, Welch & Helle, 1996; Tadokoro et al., 1996). Given the Tumen River chum salmon’s significant scientific and economic value, small body sizes not only directly reduce fecundity and impair population recovery capacity, diminish marine-to-terrestrial nutrient transport, but also indirectly lead to decreased meat yield, increased processing costs, and lowered market prices, ultimately affecting fishermen’s income and regional economic stability. Therefore, minimum harvest size limits should exceed FL_50%, with sustained monitoring of this metric. Quantitative management of minimum sizes is critical for mitigating the miniaturization of returning offspring cohorts.