Comparative Demography of Five Holocentridae Species from American Samoa

Abstract

This study provides the first insights into the age-based life histories of five Holocentridae species (soldierfish and squirrelfish) from American Samoa: Myripristis amaena, M. berndti, M. murdjan, Sargocentron spiniferum, and S. tiere. Examination of sagittal otoliths revealed that all five species exhibited long lifespans with maximum estimated ages of 17 to 40 years. The Holocentridae complex exhibited a consistent slow-turnover strategy characterized by long lifespans, asymptotic growth profiles, and low rates of instantaneous total mortality (Z ≤ 0.20 yr⁻¹). Reproductive information, derived from histological examination of gonads, indicated that the onset of maturity for all five species occurred later than is typical for many demersal reef fishes. This pattern was exemplified by female S. tiere, which reached 50% maturity at age 6.5 years. The size at maturity occurred between 65% and 91% of asymptotic length for all species. All species demonstrated a primary spawning season during the austral summer (October–February).

1. Introduction

The Holocentridae family, comprised of two sub-families: squirrelfishes (subfamily Holocentrinae) and soldierfishes (subfamily Myripristinae), are common nocturnal reef-associated fishes. The family is comprised of 90 species worldwide and are found in tropical to warm temperate waters across the Indian, Pacific, and Atlantic Oceans [1]. Holocentridae species are recognized for their distinctive red coloration and large eyes. Squirrelfish are further distinguished by a long, sharp spine at the corner of the preopercle, which can sometimes be venomous [2]. During the day, holocentrids inhabit shallow reefs, hiding within caves or beneath ledges and emerging at night to forage [3,4]. As prominent members of the nocturnal reef community, they serve as predators feeding on a variety of crabs, shrimp, and fish [5,6,7].

Holocentrids are not only ecologically important as nocturnal reef predators but are also commonly targeted in small-scale artisanal fisheries across the tropics. This is particularly true in Pacific Island communities where they support both subsistence and commercial harvests [8]. Within the nearshore fishery of American Samoa, species belonging to the Holocentridae family, locally referred to as malau, consistently rank among the top five most frequently landed families, with commercial landings exceeding 4000 pounds in 2024 [2,9]. They are primarily caught during nighttime spear operations conducted from both boat and shore.

Species within the Holocentridae family are well-adapted to their nocturnal lifestyle. Their large eyes are specialized for hunting in low-light environments, and research into their visual systems has shown a strong adaptation for scotopic (low-light) vision as they mature [10,11,12]. While their visual and feeding ecologies are relatively well-documented, there remains a surprising lack of fundamental information regarding their basic life-history traits, such as longevity, growth rates, and reproductive strategies.

Existing age estimates for the family vary and warrant scrutiny, as they were derived using indirect methods like back-calculations and extrapolation based on daily increments, which can limit reliability. Initial estimates for Sargocentron suggested longevities of less than 10 years for Sargocentron spiniferum and Sargocentorn rubrum from the Indian Ocean and Red Sea [13,14], while data from Hawaiian waters suggest slightly longer lifespans for Myripristis amaena of up to 14 years, with slow growth, and maturation occurring at a late age (approximately 6 years) [7,15]. To ensure the long-term sustainability of these important species for nearshore fisheries, effective management is necessary. This management is directly dependent upon acquiring basic life-history traits, including accurate determination of growth rates, longevity, age and size at maturity, and spawning cycles.

This study addresses these knowledge gaps by determining estimates of longevity, growth, size and age at maturity, spawning seasons, and mortality from fishery-dependent market sampling of five commonly targeted Holocentridae species from American Samoa. The focal species were selected in collaboration with the Division of Marine and Wildlife Resources (DMWR) to support local management goals and include brick soldierfish Myripristis amaena, blotcheye soldierfish Myripristis berndti, pinecone soldierfish Myripristis murdjan, sabre squirrelfish Sargocentron spiniferum, and blue lined squirrelfish Sargocentron tiere. By providing this critical life-history data from a long-term, fishery-dependent market sampling program, this research establishes the scientific basis for the future management of the malau fishery in American Samoa.

2. Materials and Methods

2.1. Study Site and Sampling

All samples were processed from archived samples collected through the NOAA Pacific Island Fisheries Science Center (PIFSC) commercial fisheries bio-sampling program [16]. Samples were collected from commercial catches procured in Pago Pago, American Samoa (14°16′ S and 170°42′ W). The American Samoa bio-sampling program was established to collect reef fish life-history data from commercial spear fishers whose catch were sold at the marketplace in Pago Pago or delivered directly to the DMWR laboratory across the street. This study used a subset of samples from March 2011 through April 2015 of the larger bio-sampling database. These samples are considered to still accurately reflect the American Samoa fishery, since there have not been any significant changes in regulations or fishing pressure since the time of collection.

Most of the sampling occurred two times a week on Wednesdays and Saturdays from 5:30 a.m. to 8:00 a.m., either at the marketplace or the DMWR laboratory [17]. Fishermen were paid USD 0.25 for each fish measured as an incentive to participate in the bio-sampling program. Fish subsampled for life-history studies were purchased opportunistically and processed at the DMWR lab [17].

For all specimens, the following morphometric data were recorded upon collection: capture date, fork length (FL) to the nearest 0.1 cm, and total body weight in grams. Paired sagittal otoliths were extracted and stored dry, while gonads were excised, weighed to the nearest 0.001 g, and preserved in 10% buffered formalin for histological analysis.

2.2. Age and Growth

Ages were estimated from a subset of sagittal otoliths from a total of 119 M. amaena, 67 M. berndti, 98 M. murdjan, 158 S. spiniferum, and 163 S. tiere. A single sagittal otolith from each specimen was weighed and prepared following the methods of Taylor et al. (2017) [18]. Otoliths were attached to a glass slide with thermoplastic adhesive (Crystalbond 509^®) and ground to the core using a 600-grit diamond lap wheel with continuous water flow. The otolith was then re-mounted and ground a second time to produce a thin transverse section (≤220 μm) encompassing the core. Finally, a thin layer of Cyrstalbond adhesive was applied to improve clarity for reading.

Alternating translucent and opaque bands representing annual growth (annuli) were counted independently along the face of the otolith section between two and three times by the same reader using a stereo microscope with transmitted light. A final age was assigned when two counts agreed. If the counts differed by one annuli (e.g., 11, 13, 12), then the middle value (e.g., 12) was assigned. If counts differed outside of these criteria, then the second otolith was sectioned and the process repeated. The relationship between otolith weight and age was first measured using a linear regression to confirm otolith weight as a reliable indicator of age.

The von Bertalanffy growth function (VBGF) was used to model growth from length-at-age data:

$$L_t ={ L}_{\infty }-{[L}_{\infty }-L_0]e^{-Kt}$$

L_t is the mean FL (cm) at age t (years), $L_{\infty }$ is the mean asymptotic FL, K is the growth coefficient toward $L_{\infty }$, and L₀ is the FL at settlement. Since most samples were collected from a commercial fishery and lacked recently settled individuals, the y-intercept of the fitted curve (L₀) was constrained to 4 cm based on the average settlement size for Holocentridae [19].

2.3. Reproduction

The sexual identity of a subsample of specimens was confirmed through histological analysis of 134 M. amaena (57 females), 86 M. berndti (49 females), 84 M. murdjan (46 females), 71 S. spiniferum (33 females), and 144 S. tiere (82 females). Gonadal tissue from each species was processed at the University of Hawaii John Burns School of Medicine. Samples were embedded in paraffin wax, sectioned transversely, and stained with hematoxylin and eosin [20]. A compound microscope was used to classify specimens as male or female based on the presence of various stages of male or female reproductive tissue. Females were classified as immature, developing, spawning capable, actively spawning, regressing, and regenerating following the modified terminology of Brown-Peterson et al. (2011) [21], with the onset of vitellogenesis marking maturity. Males were classified as either immature or mature based on the presence/absence of spermatozoa.

Length at 50% maturity (L₅₀) and age at 50% maturity (A₅₀) were estimated by fitting a binomial logistic curve to the proportion of mature individuals in 2 cm length bins. The logistic curve was defined as:

$$P_L = {\left[1+e^{-ln\left(19\right)(L-L_{50})(L_{95}-L_{50})}\right]}^{-1}$$

where P_L is the estimated proportion of mature individuals at length L, and L₅₀ and L₉₅ represent the fork lengths at which 50% and 95% of the population reach maturity, respectively. Corresponding 95% confidence intervals (CI) were determined using bootstrap resampling (1000 iterations). The same process was used to estimate age at 50% maturity (A₅₀).

Spawning season was assessed by plotting the gonadosomatic index (GSI) of mature females across the calendar year. GSI data were pooled by calendar month due to limited interannual sample numbers.

2.4. Mortality

Total instantaneous mortality (Z) was estimated from the commercial fishery age-frequency distribution using an age-based multinomial catch curve model. The model assumed knife-edged selectivity and constant mortality for ages fully recruited to the fishery. The age of full recruitment (t_rec) was set at the age of peak catch frequency. The expected per-recruit survival (S_t) for fish at or above t_rec was calculated as:

$$S_t = e^{-Z(t-t_{rec})}$$

The expected proportion of fully recruited fish at age t ($\widehat{P_t}$) was then calculated by standardizing the expected survivors over the total expected survivors up to the maximum observed age (t_max):

$$\widehat{P_t} = {\displaystyle \frac{S_t}{\sum _{t_{rec}}^{t_{max}}{S}_t}}$$

The estimate of Z was obtained by maximizing the multinomial log-likelihood associated with the observed age-proportions (P_t) and the model’s expected age-proportions ($\widehat{P_t}$).

3. Results

Over 2300 Holocentridae specimens were measured from America Samoa’s commercial fishery through the bio-sampling program. Length-frequency distributions from the commercial catch are displayed in Figure 1.

3.1. Age and Growth

Transverse sections of sagittal otoliths for all species consistently displayed clear alternating opaque and translucent zones, indicative of annuli (Figure 2). Due to time and effort constraints, we did not perform annual validation in this study and acknowledge that the absence of direct validation may introduce a potential source of uncertainty. Despite this limitation, our annual interpretation is strongly supported by several lines of evidence. Daily increment validation using tetracycline and acetazolamide has previously been performed for M. amaena in Hawaiian waters [15]. Furthermore, we proceeded with an annual interpretation based on the scientific consensus that, while incremental bands in tropical otoliths may sometimes be difficult to distinguish, there is virtually no evidence for increment formation that is not annual in fishes of this type [22]. Finally, the strong correlation between otolith weight and estimated age (Figure 3) provides additional confidence in the age estimates derived from zone counts across all species examined. Future studies focusing on formal annual validation for Holocentridae species would be valuable to definitively confirm these age estimates.

All five species exhibited moderate to long lifespans, ranging from 17 to 40 years, with both males and females reaching similar maximum ages for each species. Two Myripristis species, M. amaena and M. berndti, had the shortest lifespans, with maximum ages of 18 and 17 years, respectively. Myripristis murdjan and Sargocentron tiere showed slightly longer maximum ages of 23 and 25 years. Sargocentron spiniferum was the longest-lived species, reaching a maximum age of 40 years for females and 29 years for males.

Overall, females and males exhibited similar growth patterns across all species, as depicted by the von Bertalanffy growth function (VBGF) curves (Figure 4;

Table 1. Summary of sex-specific life-history traits for commercially harvested Holocentridae species from American Samoa. Associated 95% confidence intervals are presented in parentheses where appropriate. LWa and LWb are length weight parameters; N_aged number of specimens used in age analysis; t_max maximum age observed; L_∞ asymptotic length; K growth coefficient; L₅₀ length at 50% sexual maturity; A₅₀ age at 50% sexual maturity; Z instantaneous total mortality estimate from the multinomial catch curve.

Species	Sex	LWa	LWb	Naged	tmax	L∞	K (Year−1)	L50	A50	Z (Year−1)
Myripristis amaena	Female			49	18	17.3 (16.9–17.7)	1.01 (0.78–1.65)	14.9 (11.2–15.7)	3.3 (2.9–3.7)
	Male			70	18	17.6 (17.4–17.9)	1.95 (1.46–3.05)
	Combined	1.5 × 10−1	2.40	119	18	17.4 (17.2–17.7)	1.75 (1.32–2.96)			0.13 (0.08–0.18)
Myripristis berndti	Female			36	16	19.2 (18.7–19.9)	0.56 (0.45–0.73)	17.5 (17.2–17.8)
	Male			29	17	20.1 (19.4–21.4)	0.38 (0.28–0.51)	17.4 (16.7–18.3)
	Combined	1.3 × 10−1	2.44	67	17	19.7 (19.2–20.2)	0.46 (0.38–0.55)			0.20 (0.11–0.30)
Myripristis murdjan	Female			48	23	16.1 (15.8–16.5)	1.03 (0.74–2.30)	13.9 (12.1–14.3)
	Male			49	22	16.9 (16.4–17.5)	0.67 (0.53–0.94)	13.2 (11.2–14.0)	2.4 (1.5–3.0)
	Combined	2.1 × 10−1	2.27	98	23	16.5 (16.2–16.8)	0.76 (0.62–1.06)			0.16 (0.12–0.20)
Sargocentron spiniferum	Female			51	40	29.1 (28.2–30.2)	0.28 (0.24–0.35)	21.5 (20.0–22.6)
	Male			107	29	29.6 (28.8–30.5)	0.32 (0.30–0.35)	23.3 (20.3–25.4)	4.4 (1.9–6.8)
	Combined	1.1 × 10−1	2.53	158	40	29.3 (28.6–29.9)	0.32 (0.30–0.35)			0.06 (0.03–0.09)
Sargocentron tiere	Female			81	25	17.8 (17.4–18.2)	0.41 (0.33–0.53)	15.8 (15.5–16.2)	6.5 (4.9–7.4)
	Male			78	20	19.1 (18.5–20.0)	0.37 (0.28–0.57)	15.0 (14.0–15.6)	5.0 (2.1–5.7)
	Combined	7.6 × 10−2	2.59	163	25	18.3 (18.0–18.7)	0.42 (0.34–0.54)			0.20 (0.16–0.24)

). Sargocentron spiniferum had the largest asymptotic length (L_∞ = 29.3 cm) and the lowest growth coefficient (K = 0.32 yr⁻¹), implying the slowest approach to terminal size. The other species attained similar asymptotic lengths with 16.1 cm for M. murdjan, 17.4 cm for M. amaena, 19.7 cm for M. berndti, and 18.3 cm for S. tiere. The growth coefficient for M. berndti and S. tiere were similar at 0.46 yr⁻¹ and 0.42 yr⁻¹, respectively. M. murdjan had a slightly higher growth coefficient at 0.76 yr⁻¹, while M. amaena had the fastest estimated growth rate of 1.75 yr⁻¹, though this value may be upwardly biased due to the underrepresentation of smaller, younger samples in the dataset. Sex-specific VBGF values and associated confidence limits for all five species are presented in Table 1.

3.2. Reproduction

All five species reached maturity between 65% and 91% of L_∞. Size at maturity was similar between males and females and ranged from 13.9 cm to 23.3 cm (Table 1, Figure 5). However, the female L₅₀ estimate for M. murdjan should be interpreted with caution, as the available gonad samples included only two immature females. All five species had a limited size distribution with no samples smaller than 14 cm for any of the species, limiting the number of immature specimens, and resulting in near-complete overlap of immature and mature individuals for the smallest size classes for most of the species. This complete overlap of immature and mature individuals did not allow us to calculate male L₅₀ for M. amaena.

Unfortunately, the sample size was limited in the number of young individuals which hindered our ability to determine A₅₀. We could only determine female A₅₀ for M. amaena and S. tiere, which were 3.3 and 6.5 years, respectively, and male A₅₀ for M. murdjan (2.4 years), S. spiniferum (4.4 years), and S. tiere (5 years) (Figure 6). These results suggest a difference in life-history strategies, with members of the Holocentrinae subfamily maturing later than those of the Myripristinae subfamily.

The gonadosomatic index (GSI) data for all five species of Holocentridae indicate that the primary spawning season occurs from October to February during the summer months in American Samoa (Figure 7). Histological staging showed that all five species had females that were either spawning or spawning capable from November to February. Myripristis murdjan and S. tiere showed a strong peak in GSI during October-December, while S. spiniferum and M. amaena exhibited a high GSI in January and February. Myripristis berndti showed very little variability in GSI throughout the year, suggesting a more protracted spawning pattern.

3.3. Mortality

Total instantaneous mortality (Z) estimates were derived from the pooled catch-at-age-frequency distribution for all five species (Figure 8). The age of full recruitment (t_rec), defined as the peak age frequency, ranged from 3 to 8 years across the assessed species. Specifically, M. berndti (t_rec = 7 years) and S. tiere (t_rec = 8 years) recruited to the commercial fishery the latest, while the remaining three species were fully recruited between 3 and 4 years of age.

All five species exhibited very low instantaneous total mortality rates, consistent with a long-lived life-history strategy and high longevity (Table 1; Figure 8). Total mortality estimates were low and tightly clustered, ranging from a minimum of Z = 0.06 yr⁻¹ for S. spiniferum to a maximum of Z = 0.20 yr⁻¹ for M. berndti and S. tiere (Figure 8). The intermediate Z estimates were 0.13 yr⁻¹ for M. amaena and 0.16 yr⁻¹ for M. murdjan.

4. Discussion

This study represents the first application of robust direct-aging techniques using annual increment counts from sagittal otoliths to calculate age and growth parameters for five species within the family Holocentridae. This direct-aging approach revealed stark differences in the reported lifespans of the genus Sargocentron compared to studies that utilized indirect methods, such as back-calculations or the ELEFAN size-based methodology, which yielded maximum ages of seven years and younger [13,14]. The vast difference in maximum ages between studies emphasizes the value of labor-intensive, age-based methods like otolith annuli counts, which are essential for accurately determining the longevity of fish species.

The only other life-history study for Holocentridae counted daily increments and then extrapolated maximum age based on body and otolith size for M. amaena in Hawaii [15]. That approach estimated M. amaena could live up to 14 years with a growth coefficient of 0.22 yr⁻¹. This maximum age is only slightly younger than the estimates from the present study. However, this comparison must be viewed through an ecological lens. The sea surface temperature (SST) in Hawaii is generally around 3–4 °C cooler than in American Samoa. As SST has been found to be a significant factor in determining lifespan, with warmer waters generally yielding populations of shorter-lived fish, Hawaii’s M. amaena should theoretically be longer lived than the American Samoa population [23,24,25]. That indirect estimate from the cooler Hawaiian waters is shorter than our direct estimate from the warmer American Samoan waters, further highlighting the potential underestimation of longevity in earlier studies.

All five Holocentridae species exhibited long lifespans, reaching maximum ages of 17 years and older, with Sargocentron spiniferum estimated at up to 40 years. This longevity, combined with their relatively small adult body size, positions Holocentridae as a unique assemblage within the demersal coral reef fish life-history spectrum. Most other long-lived Indo-Pacific species, such as Snappers and Groupers, reliably reach a much larger maximum size. For example, Lutjanus xanthochilus and Lutjanus gibbus from American Samoa live over 25 years but have asymptotic lengths of 40 cm and 32 cm, respectively [26], significantly exceeding the size of Holocentridae. Conversely, while Surgeonfish (Acanthuridae) also exhibit long lifespans with small body size, they generally show faster growth rates. Species like Naso lituratus and Acanthurus lineatus from American Samoa live past 18 years, but their growth coefficients are notably higher (K ≈ 0.70–0.77 yr⁻¹) and fewer older individuals are observed within the population [27,28], suggesting a different demographic strategy than the slow-growing, long-lived Holocentridae.

The maturity estimates reinforce the slow-growth life-history strategy of these species. Species reached maturity (L₅₀) between 65% and 91% of L_∞, which is consistent with Myripristis amaena in the central Pacific [7]. Additionally, all five Holocentrids had a comparatively low investment in instantaneous reproductive output, with a mean peak GSI of 3% or less, which is considerably low among harvested species in coral reef environments [29]. A low rate of reproductive output aligns with a life-history strategy that emphasizes enhanced survival and population stability rather than rapid generational turnover.

The fishery-dependent nature of the samples presented challenges for calculating all maturity metrics. The truncated size distribution led to a near-complete overlap of immature and mature individuals in the smallest size classes, which limited our ability to precisely estimate the size and age at 50% maturity for several species. However, the female estimate of L₅₀ for M. amaena (14.9 cm) from this study was at the bottom of the range of maturity from Johnston Atoll and Hawaii of 14.9–16 cm, indicating our initial maturity estimates are within a reasonable range [7]. Despite these limitations, the available age-at-maturity estimates validate a delayed investment in reproduction compared with many similarly sized reef-associated fish species.

The Holocentridae family, characterized by long lifespans, protracted maturity, low investment in reproductive output, and low total mortality (Z ≤ 0.20 yr⁻¹) appears to occupy a unique life-history trade-off in the coral reef environment. This slow-turnover assembly is likely maintained by their specific ecological niche of being a predominantly nocturnal and cryptic species. Their cryptic behavior of hiding within reef structures during the day significantly reduces their exposure to diurnal visual predators. By spending most of their time in shelters, they limit their energy expenditure and associated risk of predation [30,31]. Their nocturnal activity allows Holocentrids to exploit a temporal niche that may further contribute to low total mortality, indicating that these populations naturally rely on a high survival rate of older individuals to create a stable biomass that can buffer against periods of poor environmental conditions or intermittent recruitment failure.

These specific traits (longevity, slow growth, and delayed maturity) create a high vulnerability to fishing pressure. Even a slight increase in fishing mortality could cause declines in the spawning stock biomass. Holocentrids are comparatively more vulnerable than other small-bodied reef taxa, which often possess faster growth rates (e.g., Acanthuridae), making Holocentrids less resilient to the loss of older, highly fecund individuals. Fishing can lead to age-truncation of the stock, which makes stocks more vulnerable to environmental changes [32,33]. Significantly, four out of the five species had over a quarter of the samples aged over 10 years old, with the only exception being M. murdjan, which had 15% of the samples greater than 10 years. The presence of so many older individuals in the sampled population indicates that, at the time of sampling, the stock had not yet been heavily truncated due to fishing pressure. These important life-history characteristics have important implications for understanding their ecology and developing management strategies.

5. Conclusions

This study definitively characterizes the Holocentridae species in American Samoa as a unique example of long-lived, slow-turnover assemblage. The combination of long lifespans (over 17 years), slow growth rates, delayed maturity, and low total mortality rates (Z ≤ 0.20 yr⁻¹) creates a highly vulnerable biological profile. This is the first study to characterize the age, growth, and maturity of the Holocentridae species throughout the Indo-Pacific using direct-aging methods. The robust information presented here will help fishery managers to accurately determine stock status and develop appropriate fishery management strategies for this unique fishery assemblage.