The captive-reared broodstocks (G_N-1 and G_N-2) used in the present trial originated from a single stock of wild P. indicus brooders (n=75) shore seined at 30-90 m depth off the coasts of Chennai, Tamil Nadu (13°02′N and 80°10′E; 12° 46’ N and 80° 18’ E) during 2018. The present investigation was grouped into two trials, Trial I and II. In the first part of the investigation (Trial I), primary sexual maturity, broodstock development, and reproductive performance of captive-reared P. indicus (G_N-1 and G_N-2) were compared in relation to existing tropical environmental conditions. Further, the spawning performance of the captive-reared broodstocks was compared with wild P. indicus brooders. In the second part of the investigation (Trial II), moulting interval, mating frequency, spermatophore quality, and oocyte development of captive-reared broodstocks subjected to eyestalk ablation were investigated in indoor maturation units.

The second part of the investigation, study II was conducted to evaluate moulting frequency, mating percentage, sperm quality analysis, and the histological architecture of ovarian developmental stages of captive-reared broodstock subsequent to eye stalk ablation for a period of 21 days maturation cycle in the indoor maturation units. The captive-reared broodstocks, G_N-1 (28-35 g) of P. indicus, were stocked in the maturation units at 6 shrimp m^-2. The broodstocks were fed with fresh and pellet feed as per the details given in Trial I.

The survival and hatching rate data are presented as percentages, and values were arcsine transformed prior to further statistical analyses. The water quality and microbial community parameters among the broodstock ponds, indoor maturation units, and larval rearing units were expressed as mean with standard error. One-way ANOVA was carried out to compare reproductive performance among G₀, G_N-1, and G_N-2 broodstocks. All the analyses were carried out after examining the data for normality and homogeneity of variance, and the ANOVA was followed by Tukey’s post hoc means comparison. The level of statistical significance was set at P < 0.05. The type of spawning (completer or partial) among the broodstock was analyzed based on the chi-square analysis (X²). All the statistical analysis was performed using SPSS Version 20 (IBM SPSS Statistics for Windows, Version 20.0.). The correlation matrix was estimated using the corrplot, a multivariate correlation program to understand the correlation among fecundity, average body weight, eggs per gram body weight, total length, etc., using a statistical package in R4.0.2.

During trial, I, broodstock development and closing of the life cycle of P. indicus were attempted in captivity, and the reproduction protocol was optimized. The nursery-reared G_N-1 and G_N-2 were subsequently stocked in grow-out and broodstock ponds. Three successive generations of P. indicus (G_N-1 and G_N-2, G_N-3) were produced in captivity from a single stock of P. indicus.

To understand the interaction effect between the monthly variation in environmental factors and reproductive parameters such as mating and gonad development of captive-reared broodstocks, the principal component analysis was carried out. The analysis suggests that reproductive parameters such as mating, spermatophore, and ovary development, along with environmental parameters like months, maximum and minimum temperature, day-night cycle, salinity, and light intensity at noon and dusk were the major contributor to the first principal component and accounted for 40.2% variances (Figure 6C). The day-length period had a positive influence on the mating of the shrimp and gonad development (Figure 6). Further, the result suggests that the effect of environmental conditions was more pronounced on mating and spermatophore development than ovarian development. Among the light intensity measured at different time points, the light intensity at dusk had the maximum negative influence on the reproductive development of the shrimp. Annual temperature and salinity also tend to have a strong influence on the reproductive performance of the shrimp, as both these parameters follow the same directions. The second principal component was mainly contributed by month and light intensity at dawn (Figure 6).

The spawning population of wild broodstock (G₀) had a wide size range from 25 g to 90 g, whereas captive-reared broodstock size ranged from 25 to 45 g. The percentage frequency distribution of the spawners (G₀, G_N-1, and G_N-2) size groups (g) contributed to successful spawning depicted in Figures 7A, B. Among the wild spawners (G₀), 35 – 40 g size groups contributed to maximum (21.6%) spawning, followed by 40 – 45 g (17.1%) size groups, while the maximum size group contributed by captive-reared spawners were 30 – 35 g (32.0%) in G_N-1 groups and 35 – 40 g (53.8%) in G_N-2 groups. Partial spawning or complete spawning was estimated visually based on the ovarian colour through the dorsal exoskeleton after spawning. The type of spawning (complete or partial) among the broodstocks was calculated based on the chi-square analysis (X²) is given in Figure 8 While 80.0% of the spawning by the wild broodstock G₀ was complete, the percentage of partial spawning was higher in G_N-1 (62.9%) and G_N-2 (53.8%), and only 37.2 and 46.2% of GN-1 and G_N-2 broodstocks, respectively recorded complete spawning.

The spawning performance of wild (G₀) and captive-reared broodstock (G_N-1 and G_N-2) is presented in Table 3. There was a significant difference (p<0.01) in fecundity among the broodstocks, with the highest fecundity recorded in wild broodstock, G₀ (2,25,110 ± 11,173 eggs), followed by captive-reared broodstocks, G_N-2 (59,923 ± 6,512 eggs) and G_N-1 (52,612 ± 5,874 eggs). Similarly, the wild broodstock also had the highest (p<0.01) eggs per gram body weight (5,137 ± 303 eggs g^-1) compared to captive-reared broodstocks, G_N-2 (1,715 ± 162 eggs g^-1) and G_N-1 (1,476 ± 151 eggs g^-1). Among the captive-reared broodstocks, 2^nd generation lines (G_N-2) had 16.2 and 13.9% higher (p>0.05) eggs per gram body weight and fecundity compared to G_N-1. Based on the egg hatchability or fertilization rate, the spawners were divided into two groups; Group I, with above 60% egg hatchability, and Group II, with a reproductive cycle below 60% egg hatchability. In the first group (Group I), G₀ recorded the highest (p<0.01) hatchability (83 ± 0.7%), followed by G_N-2 (77 ± 2%) and G_N-1 (75 ± 1%) groups. In the second group (Group II), 18.3 ± 2.9, 16.5 ± 2.5, and 6.3 ± 1.3% spawn of G_N-1 G_N-2 and G_0, respectively, recorded unfertilized eggs or poor egg hatchability. Nevertheless, the captive-reared G_N-2 had the highest contribution of ablated broodstock to successful spawning (69%), followed by G_N-1 (47%). In contrast, only 18% of ablated wild broodstocks contributed to successful spawning, as the majority of the wild broodstock spawning was contributed from unablated wild gravid spawners. Similarly, after eyestalk ablation, captive-reared broodstocks recorded higher survival (89-92%) compared with wild broodstock (71 ± 0.8%).

The correlation matrix depicts the relationship between fecundity and spawner’s body weight, carapace length, eggs per gram body weight, and total nauplius counts are given in Supplementary Figures 1A–C. The fecundity had a significant positive (p<0.05) correlation with eggs per gram body weight in broodstocks G₀ (0.48, p<0.05), G_N-1 (0.94, p<0.05) and G_N-2 (0.89; p<0.01) lines. Similarly, the egg hatchability (G_N-1: 0.85, p<0.05 and G_N-2: 0.86, p<0.05) and total nauplius count (G₀: 0.64, p<0.05; G_N-1: 1, p<0.05 and G_N-2: 1, p<0.05) were also positively correlated with the fecundity of broodstocks. However, wild broodstock fecundity did not have any positive correlation with egg hatchability. Likewise, although non-significant (p>0.05), the body weight of all three lines of spawners (G₀, G_N-1, and G_N-1) had a positive correlation with fecundity.

The average length of the different larval stages produced from wild and captive-reared broodstocks during the study cycle is given in Supplementary Figure 2. The fertilized eggs had an average size of 256 ± 3 µm, 259 ± 8, and 240 ± 13 µm, respectively, in G_N-1, G_N-2, and G_N-3 lines, and nauplius stage 1 had a size of 347 – 402 µm length. The size of the zoea larval stages (Z1 to Z3) ranged from 840 – 2502 µm, and the mysis stages (M1 to M3) ranged from 2628 to 4300 µm. The average length of post-larvae 1 of G_N-1, G_N-2, and G_N-3 lines were 4.5 ± 0.6, 4.8 ± 0.1, and 4.8 ± 0.2 mm, respectively.

In trial II, a detailed investigation of the moulting frequency, mating percentage, male reproductive performance parameters, and histological architecture of different ovarian stages of captive-reared broodstock subjected to eyestalk ablation was evaluated in indoor maturation units for a period of 21 days. At the end of the reproductive performance trial, 79% of the impregnated broodstock recorded moulting subsequent to eyestalk ablation. This corresponds to spermatophore loss with a decrease in impregnated females and a subsequent reduction in mating efficacy to 29%. Similarly, 59% of the broodstock recorded gonad development, and 52% of the broodstock responded to spawning successfully (Figure 9). The daily percentage of spawning, mating, and moulting in indoor maturation units is given in Figure 9. The average daily moulting of the broodstock post-eyestalk ablation period varied from 11 to 22%. The highest spawning was recorded during 3^rd and 12^th day of the post-eyestalk ablation. The histological architecture of ovarian developmental stages of the ablated captive-reared broodstocks is given in Figure 10, and comparative gonad development of wild brooders, G₀ is also provided (Figure 10) to have an understanding of the variation in ovarian stages among wild gravid and eyestalk ablated captive brooders. Gravid ovaries with heterogeneous distribution of cortical oocytes along with previtellogenic oocytes were noticed in ablated captive-reared brooders in the final stage of maturity, while in wild gravid ovaries, a homogenous distribution of cortical oocytes surrounded by oogonia cells was only noticed.

The total sperm count and percentage of normal and abnormal sperm of the captive-reared male broodstock before and after 21 days maturation cycle in the indoor maturation unit are given in Table 4. The male broodstock had an initial total sperm count of 4 ± 1 × 10⁶ spermatophore ampoule^-1, which reduced (p>0.05) to 3.5 ± 0.5 × 10⁶ spermatophore ampoule^-1 at the end of the trial. The percentage of the sperm with normal morphology was 83.9 ± 1.6% at the start of the experiment, which reduced to 79.6 ± 2.8% at the end of 21 days maturation cycle.

The reproductive performance of the shrimp can be evaluated through mating efficiency, spawning frequency, fecundity, hatching rate, etc. (Browdy, 1992; Wouters et al., 2001; Hoa et al., 2009; Coman et al., 2011; Goodall et al., 2016). The size of the individual spawner also plays a vital role in shrimp reproduction (Ibarra et al., 2007), as it is positively correlated with fecundity (Menasveta et al., 1994; Arcos et al., 2004; Debnath et al., 2014) and spawning frequencies (Andriantahina et al., 2012; Wen et al., 2015). In the present trial too, although non-significant, a positive correlation between fecundity and body weight was noticed among wild (G₀) and captive-reared broodstocks (G_N-1 and G_N-2). In the present study, although the size ranges of the P. indicus broodstock were 25 – 90 g (wild) and 25 to 45 g (captive-reared) (Figure 7), the maximum spawning was contributed from 30 – 45 g size groups spawners. While the size-related maturity is species-specific, above 35 grams and below two years of age of Farfantepenaeus paulensis (Peixoto et al, 2005) and P. vannamei (Ren et al., 2020) are reported as ideal for seed production. Likewise, many commercial P. vannamei hatcheries prefer 30 – 45 g broodstock for nauplius production (Marsden et al., 2013; Wen et al., 2015; Ren et al., 2020). Ren et al. (2020) reported that females in different size classes employ different strategies when allocating energy to reproduction. The authors noted that medium-size group females generally invest more energy in reproduction than growth, resulting in a slower growth rate and multiple spawning than large brooders with a faster growth rate and less frequent spawning.

Hatchability is another important factor that limits the successful seed production of captive-reared shrimps. In the current study, wild broodstocks recorded higher hatchability (83 ± 1.7%) compared to captive-reared stocks G_N-1 (75 ± 3%) and G_N-2 (77 ± 3.7%) stocks. This matches with earlier reports where wild broodstocks were also reported to have better fecundity and hatchability compared to domesticated ones (Browdy et al., 1986; Keys & Crocos, 2006; Emerenciano et al., 2012; Arnold et al., 2013; Marsden et al., 2013; Wen et al., 2015). Likewise, Coman et al. (2006) reported below 40% and often below 10% hatching rate from domesticated P. monodon. Similar to this line, in the present trial, 16.5 to 18% of the captive-reared spawning population recorded hatchability below 60% compared to wild spawners (6 ± 1.2%). Further, the fecundity or eggs per gram body weight of captive-reared broodstocks (G_N-1 and G_N-2) were significantly lower (p<0.05) than the wild broodstocks (Table 3), which can be attributed to a higher percentage of partial spawning in captive-reared stocks (Figure 8) compared to wild gravid spawners. However, it is worth noting here that the reproductive performance of domesticated broodstocks increases over successive generations (Preston et al, 2010; Coman et al., 2013). For instance, the 5^th generation line (G5) of P. monodon recorded an increase in spawning percentage and hatching rate from 52.0 to 77.1% and 21.5 to 31.6%, respectively, compared to G1 stocks (Coman et al., 2013). In the present study, too, G_N-2 of P. indicus recorded 16.2 and 13.9% higher (P>0.05) eggs per gram body weight and fecundity, respectively, compared to first-generation (G_N-1). Besides, Coman et al. (2013) reported significant improvements in nauplius count from the first two generations to the later generation of P. monodon (G5 and G8) stocks. As wild broodstock of P. indicus is reported to be a highly fecund animal with a fecundity up to 4,00,000 (Jayawardane et al., 2002; Anand et al., 2019), it is reasonable to assume that domestication and genetic improvement of the species through the selective breeding program can definitely improve the fecundity, breeding efficacy and reproductive size group of the captive-reared broodstock.

It is also important to note that, unlike open thelycum shrimp, closed thelycum shrimp possesses inherent reproductive constraints in captivity (Browdy et al., 1986; Hetzel et al., 2000; Coman et al., 2007; Wilder, 2019). Lack of proper mating in the confined indoor system ultimately results in poor fertilization rates. This is supported by the frequent moulting (79%) of broodstock subsequent to eyestalk ablation in the indoor maturation units (Trial II), which resulted in a loss of sperm pack resulting in poor impregnation percentage (29%). Further, a decline in total sperm count and an increase in the rate of abnormal sperm (Table 4) also point toward the fact that sperm quality deterioration in indoor captive maturation units can also lead to poor mating percentage or fertilization rate. Although it is reported that spermatophores from single mating can fertilize up to three successive spawns (Primavera, 1985), sperm quality deterioration and blackening of sperm pack in thelycum are also reported from closed thelycum shrimps (Pratoomchaat et al., 1993; Jiang et al., 2009; Alfaro-Montoya et al., 2018).

Furthermore, the delay between impregnation and gonad development of closed thelycum shrimps, unlike open thelycum shrimp, where shrimp freshly mate just before spawning (Palacios and Racotta, 2003; Arcos et al., 2004; Ibarra et al., 2007) also remain another challenge in successful spawning performance of closed thelycum shrimp in captivity. These remain major reasons behind the fact that despite the advancement in reproduction, domestication, and selective breeding of closed thelycum shrimp lag far behind the selective breeding of open thelycum shrimps such as P. vannamei. Even though artificial inseminations are successful in open thelycum shrimp, a few with limited success are reported in closed thelycum shrimps (Alfaro-Montoya et al., 2018). In fact, there is a dearth of scientific data on how to improve the mating efficacy of closed thelycum shrimps in captive systems. Recently, Santander-Avancena et al. (2022) reported monosex rearing of P. indicus (5-7 g) potentially increases mating success up to 80% compared to 20% mating recorded in communal rearing. This suggests the scope for exploring a single-pair mating system in P. indicus indoor captive systems. Further, as shrimps reared in outdoor broodstock ponds recorded 100% impregnation or mating, a multipronged strategy with manipulation of conducive environmental cues can enhance the mating efficacy in the indoor confined maturation units and is worth investigating.

The wild broodstocks used in the trial had a higher latency period and lower post-eye stalk ablation survival (71%) compared to domesticated broodstocks (89 – 92%). This matches with a recent historical review on the reproductive performance of wild P. monodon brooders in Bangladesh shrimp hatcheries, where authors reported wild broodstock quality deterioration and decline in survival after eyestalk ablation (Debnath et al., 2015). According to the authors, higher viral load in coastal water, poor post-capture handling practices, and transportation stress might be plausible explanations for the low survival of wild broodstocks in captivity (Debnath et al., 2014; Debnath et al., 2015). In fact, the latency period ranges from three days up to two months depending on the age and source of broodstock, stage of the moult cycle, etc. (Primavera, 1985; Palacios and Racotta, 2003; Arcos et al., 2004). It is worth noting that a lower latency period and higher survival recorded in the captive-reared broodstocks in the present study emphasize better broodstock condition and reproductive readiness of the captive-reared shrimp.

The oocyte size groups in different ovarian developmental stages of captive-reared broodstock in the present trial (Trial II) match the ovarian histological architecture of wild P. indicus broodstocks (Mohamed and Diwan, 1994; Anand et al., 2019). However, matured ovaries of the captive-reared broodstock recorded asynchronous ovarian development with cortical oocyte and previtellogenic oocytes, unlike gravid ovaries of wild broodstock with homogenous development of cortical oocytes throughout the ripe ovaries (Figure 10). In contrast, Palacios et al. (1999c) could not find any differences in relation to the proportion of late vitellogenic or cortical oocyte stages in pond-reared P. vannamei, unlike differences recorded in previtellogenic and early-vitellogenic oocyte distribution. A heterogeneous distribution of vitellogenin in the ovaries of ablated pond-reared broodstocks may be a plausible reason for the lower proportion of cortical oocytes in captive broodstock compared to wild gravid ovaries. Therefore, further investigation is needed to understand the vitellogenin pathways in ablated captive stocks compared to wild gravid spawners. The lower viability of the spawners immediately after eyestalk ablation was recorded (Primavera, 1985) and is reported to be due to due to the stress associated with the eyestalk ablation procedure (Palacios et al., 1999b).

The differential oocyte development in gravid ovaries of captive-reared brooders might have resulted in the release of only matured ova and a subsequent higher percentage of partial spawning in captive-reared stocks compared to wild gravid spawners. Further, unlike earlier reports where a lack of cortical rods was reported in gravid oocytes of pond-reared P. japonicus broodstock (Yano, 1998), well-distinguished cortical oocytes were recorded in the gravid ovaries of captive-reared P. indicus. The complete development of oocytes into its final cortical oocyte stages suggests the broodstock development potential of P. indicus captivity. The size of the fertilized eggs and larval stages (Supplementary Figure 2) also matches with earlier observations in wild P. indicus (Muthu et al., 1979; Anand et al., 2019).

Spawning frequency or multiple spawning is another crucial factor in breeding programs (Ibarra et al., 2007). It is reported that the majority of the nauplius production is contributed by a smaller population of multiple spawners (Arcos et al., 2003; Racotta et al., 2003). For instance, a 36-day study using P. vannamei broodstocks revealed only 19% population had multiple spawning with 18, and 15% population spawned once or twice, whereas 48% of females did not spawn at all (Arcos et al., 2003). Equally, a 29 days broodstock rearing study revealed only 14% of the population recorded multiple spawning (Arcos et al., 2004). Our results are consistent with earlier reports where only 20% of the captive-reared broodstock reported multiple spawning. However, selectively bred P. vannamei are reported to spawn up to 10 to 12 times per cycle in commercial shrimp hatcheries (Kannan et al., 2015; Ren et al., 2020). This indicates apart from phenotypic parameters (Hoang et al., 2002; Arcos et al., 2003; Palacios and Racotta, 2003), genetic (Ibarra et al., 2007) or endocrine parameters (Tsutsui et al., 2005; Treerattrakool et al., 2014; Huerlimann et al., 2018) also can play a significant role in shrimp broodstock performance. This further suggests that elucidating the endocrine mechanism behind the productive and non-productive domesticated shrimp is worth exploring. Additionally, though the broodstock performance of P. indicus cannot be compared with the reproductive performance of selectively bred P. vannamei, the reproductive potential of Indian white shrimp is on par with the optimal ranges reported for other closed thelycum shrimps (Keys et al., 2004; Preston et al., 2010). As the reproductive output of the species increases over successive generations, genetic improvement of spawning frequency, which is reported to have moderate heritability, h2 = 0.15 (Ren et al., 2020), may also be aimed at genetic improvement and selective breeding programs of Indian white shrimp.

Length-weight relationship and the condition factor of the shrimp is influenced by both biotic and abiotic parameters (Gopalakrishnan et al., 2014). It is reported that crustaceans exhibit a bimodal growth pattern where one sex grows larger in size than the other (Hartnoll, 1982). This sigmoidal growth pattern becomes more conspicuous as the shrimp attain sexual maturity (Dall et al., 1990). In the present study, females were bigger than males as they grew faster toward maturity (Hoang et al., 2003). The L-W of P. indicus during the grow-out, and reproductive phase recorded negative allometric growth. It is recorded that body weight in proportion to length is higher in older shrimps. This is in accordance with earlier reports on P. monodon (Primavera et al., 1998); P. vannamei (Chow and Sandifer, 1991), and M. rosenbergii (Kurup et al., 2000; Lalrinsanga et al., 2012). This can be attributed to the fact that as shrimp grows towards maturity, the weight gain is not proportional to the length of the shrimp, unlike nursery stages where a positive allometric growth is observed due to increases in length with proportional body weight gain. Similarly, a negative allometric was also reported in pond-reared (b = 2.72) and wild shrimp (b = 2.49) by Gopalakrishnan et al. (2014). Though it is reported that wild P. monodon from natural water was heavier than the shrimps from culture ponds (Cheng and Chen, 1990), similar trends were not applicable in the present studies as only broodstock from wild or cultured ponds were available.

Relative condition index, a ratio of body weight to length, suggests that shrimp in good condition has a factor 1 or above (high body weight/length ratio), compared to poor condition factor (below 1) with a low body weight to length ratio (Emmerson, 1980; Palacios et al, 1999a; Arcos et al., 2003). In the present study, all the shrimps at different growth stages had a relative condition index (Kn) above 1, indicating shrimps were in good rearing condition. In contrast, a lower Fulton condition factor, K value, is reported in the early life stages (Kunda et al., 2008; Lalrinsanga et al, 2012; Mane et al., 2019). This matches with present reports where nursery-reared shrimps had b value closer to 3, so the Fulton condition factor (K) was calculated (Lalrinsanga et al, 2012), which ranged from 0.65-0.67, and was lower compared to the relative condition factor (Kn) calculated for grow out or adult brooders (1.02-1.04).This can be attributed to the fact that in early life stages, shrimp grew faster in length than weight (Mane et al., 2019). Similarly, a lower K value (0.70 ± 0.1) was recorded during the early culture phase of M. rosenbergii compared to a higher K value (1.05 ± 0.1) towards the end of the culture (Kunda et al., 2008; Lalrinsanga et al, 2012) Among the broodstocks, the condition factors of the female broodstocks were higher than the male irrespective of the generation (Udoinyang et al., 2016). On the contrary, Kumari et al. (2013) recorded higher K values for males. However, in the present study, no sex-wise variation in the relative condition index was recorded, which is in line with Choe (1971), who observed no difference in the body length and weight between sexes in P. japonicus. It is also important to note that environmental or cultural conditions (stocking density, feed management) can also play a direct role in the growth and condition factor of the penaeid shrimp.

Although the reproductive performance of the shrimp is under the control of additive genetic composition, farm management regime and culture environment also play a crucial role in shrimp reproduction (Ibarra et al., 2007). Environmental cues are critical modulators of gonad or vitellogenic stimulating hormones in penaeids shrimps. The PCA analysis revealed a strong influence of monthly variation in environmental parameters such as temperature, day-night cycle, and salinity in mating and gonad development of captive-reared Indian white shrimp (Figure 6). Abiotic parameters such as salinity, temperature, tank size and depth, light intensity and photoperiod, and lunar periodicity are also reported to play a prominent role in the reproductive performance of the shrimp (Hoang et al., 2002). In the present trial, although mating and maturation were not major constraints in outdoor broodstock ponds, a lack of predictability over critical parameters such as salinity and temperature was recorded due to seasonal fluctuations. For example, although gonad development of captive-reared P. indicus was recorded at 28 – 36 ppt salinity, 100% gonad absorption was recorded when salinity was reduced to below 25 ppt during the rainy season. Similarly, the earlier reports on the resorption of ovaries below the optimum level of salinity (Ituarte et al., 2010) also suggest the vital role of salinity in shrimp reproduction. Photoperiod (light and dark cycle) and light intensity were also reported to play a critical role in ovarian maturation and spawning of P. monodon (Primavera, 1985), P. stylirostris, P. vannamei (Chamberlain and Lawrence, 1987, Wyban et al., 1987), P. indicus (Emmerson et al., 1983), and P. merguiensis (Hoang et al., 2003). For example, in the present trial, successful mating was recorded in broodstocks reared under natural photoperiod compared to indoor maturation units with limited light intensity (Figures 2A, B). However, the gonad development was also not hampered even at higher light intensity during the natural photoperiod regime (Figure 2B) in broodstock ponds. In contrast, it is reported that dim light is an important prerequisite for maturation (Primavera, 1985), and bright light (above 1100 lux) or continuous dark conditions can inhibit the ovarian maturation of brooders (Hoang et al., 2003). This suggests light intensity requirement of the species varies depending on the mating or maturation stages of the shrimp. Natural ecosystems where shrimp mature and spawn in deep waters have weak light intensity, unlike complete darkness (Hoang et al., 2002). Unlike 100% mating or impregnation recorded in outdoor rearing units, the lack of subdued dawn and dusk effect and insufficient light intensity gradation over time might have resulted in poor mating efficiency in indoor maturation units (Trial II). The PCA analysis also revealed a strong influence of water temperature on the gonad development of captive-reared P.indicus (Figure 6). The present findings are in agreement with earlier studies where temperature is reported as one of the vital abiotic factors in shrimp reproduction (Primavera, 1985; Staples and Heales, 1991). For example, an increase in temperature from winter to spring was reported to enhance spawning in natural conditions (Rothlisberg et al., 1987; Dall et al., 1990; Yano, 1993). Further, a combined effect of day length and temperature is also found to stimulate ovarian maturation and spawning in P. duorarum (Cripe, 1994) and P. stylirostris (Robertson et al., 1991). These reinforce the importance of environmental cue manipulation to achieve better mating efficiency and reproductive performance of closed thelycum shrimps in captivity.

The rearing system is also reported to play a vital role in shrimp broodstock development. It is reported that the reproductive performance of domesticated P. vannamei was much better when raised in earthen ponds compared to the clear water recirculatory system (Ren et al., 2020). This finding corroborates our personal observation that mating efficiency and body weight gain were much inferior in indoor systems compared to outdoor broodstock ponds. Likewise, various studies also report the poor performance of indoor tank-raised broodstock compared to pond-reared broodstock irrespective of similar husbandry techniques (Preston et al., 2004; Coman et al., 2006; Andriantahina et al., 2012; Arnold et al., 2013). This urges many hatchery operators to depend on broodstocks raised in earthen broodstock ponds compared to SPF stocks reared in indoor maturation tanks (Ren et al., 2020). Conversely, outdoor ponds are subjected to seasonal fluctuation, natural calamities, biosecurity breaches, and subsequent pathogen entries. These necessitate the development of controlled semi-outdoor broodstock rearing units to achieve optimum positive conducive cues for shrimp gonad maturation under stringent biosecurity.

Further, suboptimal maturation conditions can also definitely affect the performance of broodstocks. Hence, it is also important to develop strategies to compensate for nutritional loss through well-managed nutritional supplements (Bray et al., 1990; Palacios et al., 2001; Wyban, 2007; Goodall et al., 2016). Broodstock ponds are conducive ecosystems that can enhance broodstock performance by harnessing natural productivity, whereas tank-based confined systems need natural microbial community manipulation. For example, a biofloc-based tank system reared P. duorarum broodstock recorded better reproductive performance compared to clear water system (Emerenciano et al., 2013). However, most hatchery owners prefer clear water indoor broodstock rearing systems for better broodstock visibility. The development of outdoor mixotrophic broodstock rearing units based on the integration of autotrophic and heterotrophic communities can have dual advantages as it can mimic the natural ecosystem through conducive environmental cues and can provide the scope for nutritional intervention based on microbial enrichment under stringent biosecurity in confinement.

The domesticated shrimps are also reported to have better feed utilization efficiency (Glencross et al., 2013) and disease resistance (Knibb et al., 2015), and enhanced farm productivity (Preston et al., 2010; Norman-Lόpez et al, 2016) over successive generations. Further, improvement in growth and metabolic advantages of closed thelycum P. monodon (Preston et al., 2010; Glencross et al., 2013), P. japonicus (Preston et al., 2004), P. esculentus (Keys and Crocos, 2006) were also recorded over generations. Further, it is reported that significant improvement in growth can be achieved from the 4^th or 5^th generation onwards (Norman-Lόpez et al, 2016). Although no genetic selection procedures were followed in the present study, the improved growth rate or reproductive performance of G_N-2 compared with G_N-1 might be due to unintended selection that happened during the domestication procedure (Figure 4). These further indicate the scope for potential growth improvement during the genetic selection-based breeding program of Indian white shrimp.