A total of 201 C. pardalis and 60 A. scopas were collected between January 2016 and February 2017 from the reefs of Vavanga Village, located on the southwest coast of Kolombangara Island, Western Province, Solomon Islands (8°03’41” S, 156°58’05” E). Samples were collected from artisanal indigenous spearfishers with no control over the selectivity of size or sex, therefore the age structure presented here is likely to reflect the population in the area sampled. Total length (TL), standard length (SL) and body depth (BD) were recorded to the nearest mm, and total weight (TW) and gutted weight (WG) were measured to the nearest mg. The sex of each specimen was determined during dissection based on external characteristics and macroscopic examination of the gonadal tissue. Both sagittal otoliths were extracted from each specimen ( Figures 1A, C ), rinsed in 70% ethanol and stored dry in 96-well plates.

For each fish, one otolith from the paired sagittal set was randomly selected and used to prepare a thin transverse section, following the methodology described by Visconti et al. (2018b). Otoliths were ground from the posterior and anterior margins to obtain a thin transverse section through the nucleus of approximately 200μm using a combination of Carbimet Silicon carbide P2500 and FiberMet 0.3 μm (Buehler, http://www.buehler.com). Otolith thin sections were subsequently covered with a clear low viscosity epoxy resin (RT310, Resintech, http://www.resintech.co.uk). Otolith sections were viewed under reflected light against a black background with a Leica DM2000 compound microscope (www.leica.com) at ×100 magnification, and pictures were taken for each sample with a GT Vision GXCAM-U3PRO camera (www.gtvision.co.uk) for subsequent analyses. Age determination was conducted using ObjectJ software (www.simon.bio.uva.nl/object) by examining the alternation of opaque and translucent zones along a linear path from the otolith core to its distal edge. The assumption of annual periodicity in the formation of these zones in monacanthid otoliths has been previously proposed (Rogers et al., 2001; Visconti et al., 2020). Each otolith section was read twice by the expert reader (V.V.), with the number of opaque zones counted to estimate age in years. If the two independent counts were identical, the reading was accepted as the final age estimate; otherwise, the individual was excluded from subsequent age-based analyses. A standard birthday of 1 July was assigned to all individuals, based on the assumption of year-round spawning and following the approach of Proctor et al. (2021). In order to examine the precision of the reading, the coefficient of variation (CV) and the average percentage error (APE) were calculated amongst readings (Beamish & Fournier 1981; Campana, 2001). A total of 140 C. pardalis and 42 A. scopas individuals were successfully aged ( Figure 1 ).

All specimens were eviscerated upon capture, and each pair of gonads was photographed ( Supplementary Figure S1 ), weighed to the nearest 0.01 g, stored in 10% buffered formalin, and histologically processed within six months. All the tissues were processed following Visconti et al. (2018a) where one of the two gonad lobes for each gonad was embedded in paraffin wax, sectioned at 7 μm, and stained using a combination of Ehrlich’s and Gill’s haematoxylin and eosin (H&E). All the gonad sections were observed using a Leica DMRE upright microscope equipped with a colour camera (Leica DC500), and digital images were captured using AnalySIS LifeScience software. Macroscopic and microscopic developmental stages were identified and assigned according to Trip et al. (2011a) and Visconti et al. (2018a) ( Supplementary Figure S2 ).

The distribution of size and age class frequency was analysed to understand the growth and longevity of both species. Instantaneous total mortality rates (Z) were derived from age-based catch curves. A linear regression was applied to the natural logarithm of the descending frequency of individuals per age class, starting from the modal age class (i.e., the age with the highest abundance) and extending to the oldest age class showing a consistent decline in abundance. This included ages 3–9 for C. pardalis and 6–12 for A. scopas.

The relationship between weight and length was described following equation (Ricker, 1973):

where W denotes total weight (g), L represents standard length (mm), ‘a’ is the y-intercept or initial growth coefficient, and ‘b’ is the slope or growth coefficient. For most fish species, growth is isometric, with ‘b’ typically around 3.0. A ‘b’ value less than 3.0 indicates negative allometric growth, while a value greater than 3.0 signifies positive allometric growth.

The relationship between size and age was modelled using the von Bertalanffy growth function (VBGF), fitted for combined sexes, and for males and females separately for both species, following the equation:

where L_t is the estimated mean size-at-age t, and L_∞ represents the mean asymptotic size or the maximum average length a fish would theoretically reach if it lived indefinitely. K is the growth coefficient, reflecting the steepness of the ascending portion of the growth curve and indicating how rapidly the fish approaches its asymptotic length, and t₀ denotes the theoretical age at which the fish’s length would be zero. t₀ is not always an actual observable point but rather a parameter that shapes the growth curve, especially if sufficient young and smaller samples are available. The VBGF was fitted by constraining the curve to a length-at-settlement of 10 mm TL according to previous studies on the Monacanthidae family (Kingsford and Milicich, 1987; Visconti et al., 2020). VBGF growth trajectories were compared between the two species using 95% confidence ellipses surrounding the traditional VBGF estimates of parameters L_∞ and K (Kimura, 1980). Mean maximum age T_max and mean maximum body size L_max were calculated for each species as the average age (in years) of the 15% oldest individuals, and as the average body size (total length, in mm) of the 15% largest individuals found within each species (Beverton, 1992; Trip et al., 2008).

Age and size at sexual maturity were determined for C. pardalis using immature and mature females collected between 2016 and 2017 ( Supplementary Figure S2 ). Size at maturity was estimated from the size at which 50% of females were sexually mature (L₅₀ ) and the size at which 95% of females were sexually mature (L₉₅ ) (Trip et al., 2011b; Visconti et al., 2018a). Age at maturity was estimated from the age at which 50% and 95% of females were sexually mature (T₅₀ and T₉₅ , respectively) following Trip et al. (2011b) and Visconti et al. (2018a). The logistic function (the maturity ogive) was fitted to the proportion of mature fish in each year class that were sampled during the spawning season and the ogive was fitted to the mid-point of each age class. The best-fit logistic function was estimated by minimizing the negative log₁₀ of the likelihood based on a probability density function with a binomial distribution (Haddon, 2001). No immature females were present among the A. scopas sampled, which prevented us from calculating the logistic functions and estimating size- and age-at-maturity for this species.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The transverse sectioned otoliths showed clear alternating patterns of opaque and translucent zones, with C. pardalis exhibiting better reading patterns ( Figure 1B ) than A. scopas ( Figure 1D ). C. pardalis ranged in total length (TL) from 104 mm to a maximum of 163 mm, while A. scopas exhibited a broader size range, from 110 mm up to 188 mm TL ( Figure 2A ). In terms of age distribution, C. pardalis reached a maximum age of 9 years while A. scopas attained 12 years ( Figure 2B ). The ageing precision for both species resulted similar to those published for other monacanthids (Visconti et al., 2018a; Miller et al., 2010), with C. pardalis (n=140) scoring a CV=6.7% and APE=8.12, and A. scopas (n=42), a CV=2.5%, APE=20.3.

Mortality estimates for C. pardalis and A. scopas based on age-based catch curves revealed annual mortality rate of 39.25% (yr^-1) and 47.74% (yr^-1), respectively, reflecting their relatively short life spans ( Table 1 ).

Analysis of the length (L) and weight (W) relationship for C. pardalis revealed a negatively allometric L-W growth coefficient of 2.75, which indicates a species that prioritises growth over condition ( Table 1 ; Supplementary Figure S3 ). When partitioned by sex, females exhibited a stronger negative growth coefficient (b=2.52) compared to males (b=2.80), suggesting a behavioural dominance of males over females. The negative allometric relationship between length and weight was the same for A. scopas (b=2.79), however, when partitioned by sex the growth coefficient was similar between males (b=2.85) and females (b=2.90).

Given the 1:1 sex ratio for both species, the length-weight relationships provide an interesting insight into the growth vs condition strategy of this species, which coincides with the results of the VBGF analysis. C. pardalis ( Figure 3 ; Table 1 ) exhibited fast early growth (K=1.06) in the first two to three years of life, before reaching an asymptotic length of 145.68 mm (TL). In contrast, A. scopas showed comparatively slower and more gradual growth (K=0.63), reaching an asymptotic length of 159.02mm (TL) later in life ( Figure 3 ; Table 1 ). The comparison of combined growth curves of both species showed C. pardalis to grow faster in their first 3 years of life while A. scopas grow relatively slower but attain bigger asymptotic lengths and older ages. These differences were confirmed by plotting the 95% confidence ellipses surrounding the VBGF estimates of parameters L_∞ and K ( Figure 3 ). When partitioned by sex, C. pardalis males grew slower early on but reached a larger asymptotic length (151.60 mm TL) compared to females (135.32 mm TL) ( Supplementary Figure S4 ; Table 1 ). When partitioned by sex, A. scopas males grew faster than females, and males reached a comparatively larger asymptotic length (159.37 mm TL) than females (149.38 mm TL) ( Supplementary Figure S4 ; Table 1 ).

Histological analysis identified six reproductive stages that were assigned to females to the gonad histological sections of both species, following the criteria published in Visconti et al. (2018a) ( Supplementary Figure S2 ). Inactive ovaries showed immature characteristics with chromatin nucleolar stage of primary growth cells (pre-vitellogenic oocytes) organised in a compact grid of ovigerous lamellae surrounded by a thin ovarian wall. Ripening ovaries exhibited the developing features during the secondary growth and were divided into two sub-stages: the first (ripening1) displaying early-stage cortical alveolar and primary vitellogenic oocytes, and the second (ripening2) displaying secondary (Vtg2) and tertiary (Vtg3) vitellogenic oocytes. Spawning ovaries contained several full yolk granule oocytes, with most undergoing germinal vesicle migration and breakdown, fully hydrated oocytes were abundant in the central lumen, and postovulatory follicles (POFs) and atretic oocytes were observed in the ovarian tissue. Post-spawning ovaries showed atretic hydrated oocytes and signs of previous spawning (post-ovulatory follicles, intra-lamellar muscle bundles). Resting ovaries (reproductively inactive) had numerous primary growth cells kept together in the lamellae by thick interlamellar muscle bundles and a thick ovary wall ( Supplementary Figure S2 ).

The logistic function for size-at-maturity for C. pardalis showed that maturity is reached at 110mm of FL (L₅₀ ) with the 95% of the individuals being mature at 123.5mm (L₉₅ ). The logistic function for age-at-maturity for C. pardalis revealed an early maturation in females at just 1 year of life, with 50% of females becoming mature and active spawners by 0.92 years, 95% of females becoming mature by 1.09 years, and 100% of the individuals becoming mature and active spawners by the second year of life ( Supplementary Figure S5 ). The youngest and smallest A. scopas female (1 year old and 110 mm TL) was diagnosed as being sexually mature with ripened ovaries, suggesting early sexual maturation (in age and body size) in A. scopas, which was also observed in C. pardalis.